Abstract
In this paper I will present a puzzle about epistemic akrasia, and I will use that puzzle to motivate accepting some non-standard views about the nature of epistemological judgment. The puzzle is that while it seems obvious that epistemic akrasia must be irrational, the claim that epistemic akrasia is always irrational amounts to the claim that a certain sort of justified false belief—a justified false belief about what one ought to believe—is impossible. But justified false beliefs seem to be possible in any domain, and it’s hard to see why beliefs about what one ought to believe should be an exception. I will argue that when we get clearer about what sort of psychological state epistemic akrasia is, we can resolve the puzzle in favor of the intuitive view that epistemic akrasia is always irrational.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I’ll alternate freely between the locutions “one ought to believe that P,” “it is reasonable to believe that P,” “one has justification to believe that P,” “it is rational for one to believe that P,” and “one’s evidence supports believing that P.” I don’t deny that these locutions can fruitfully be given distinct senses, but I do think that this is a matter of stipulation; I don’t think ordinary language already distinguishes them.
These claims are all reminiscent of Moore’s paradox. I believe there are close connections between Moore’s paradox and the phenomenon of epistemic akrasia, but discussing Moore’s paradox in any depth is beyond the scope of this paper.
Some philosophers have argued that epistemic akrasia is impossible (see Hurley (1989), Pettit and Smith (1996), Adler (1999, 2002), and Raz (2007)), so there are no cases that can be completely uncontroversially described as ones in which someone holds a belief while taking that belief to be irrational. For instance, rather than describing this case in which Matt straightforwardly believes that flying is dangerous, and straightforwardly believes that he oughtn’t believe this, we might instead describe it as a case of “in-between believing”; one in which Matt neither straightforwardly believes nor straightforwardly disbelieves that flying is dangerous. See Schwitzgebel (2001, 2002). While this isn’t how I put things, I suspect that such a position would ultimately dovetail nicely with the theses defended in this paper.
Though I’m certainly not the first to discuss such states, or to use the phrase “epistemic akrasia.” See, e.g., Owens (2002).
Of course, depending on the specific nature of the case, it might be reasonable to revise one’s beliefs in either direction—paradigm cases are ones in which the first-order belief or intention ought to be revised, but there will be others in which the first-order belief or intention is fine, and it is the normative judgment about that belief/intention that should be revised. For instance, we might say that Huck Finn is akratic with respect to the question of whether he should turn Jim in, but that he ought to revise his belief about the moral status of slavery, rather than changing his intention not to turn Jim in. See Arpaly (2003).
Though see Littlejohn (2012) for a dissenting note.
This isn’t to say that there are no propositions such that justifiably believing those propositions entails truly believing them. Necessarily true propositions would provide counterexamples to that claim; if one justifiably believes that 2 + 2 = 4, then one truly believes that 2 + 2 = 4. Rather, the claim relies on a the vague (but intuitive) idea that propositions can be grouped into domains, and that for no domain D is it the case that for all propositions P in domain D, justifiably believing P entails that P is true.
I am ignoring the possibility that the student might believe both halves of the conjunction without believing the conjunction. There may very well be cases in which beliefs fail to be closed under conjunction, and in which such failures are rational—see Kyburg (1970) for a canonical argument for such a view. But I take it that such cases, if they exist, are sufficiently different from the case of the skeptic’s student that I don’t beg any important questions by setting them aside.
Weatherson (2008) offers a similar argument, though he focuses on the general possibility of justified false beliefs about epistemology, and less on the specific case of testimony as a method for generating such beliefs.
See Weatherson (2008) and Coates (2012), and arguably Wedgwood (2011, 2012), Williamson (2000) and Lasonen Aarnio (Forthcoming), for endorsements of the view that epistemic akrasia is sometimes rational. Williamson and Lasonen Aarnio are tricky cases, since both are sympathetic to the idea that one ought to believe that P only if one knows that P. Given this view, it can never be the case that one ought to believe that one ought to believe that P, while one oughtn’t believe that P—knowledge is factive, so if one knows that one knows that P, then one knows that P. But because they allow for cases in which one ought to be confident that one ought to be confident that P even though one oughtn’t be confident that P, I think they deserve to be included as defenders of the rationality of (something like) epistemic akrasia. Also, while Wedgwood doesn’t quite endorse the view that epistemic akrasia is sometimes rational, he comes very close—in Wedgwood (2012) he defends the idea that one can be rational in drawing an inference while taking that inference to be an irrational one.
The only alternative attempt I know of is provided by Smithies (2012). His strategy is quite different from mine; he defends a strong form of internalism on which what one has justification to believe supervenes on one’s phenomenology, and tries to ground our distinctive access to facts about justification in our distinctive access to facts about our own phenomenology. While I’m sympathetic to many of Smithies’ conclusions, there are reasons to doubt whether his strategy for explaining the irrationality of epistemic akrasia can ultimately succeed. For one, Smithies’ strategy requires that we have a degree of epistemic access to facts about our own phenomenology that we may not, in fact, have: see Schwitzgebel (2008).
John Stuart Mill was an early defender of such a strategy. He describes a case of akrasia as follows:
But it is obvious that “I” am both parties in the contest; the conflict is between me and myself; between (for instance) me desiring a pleasure, and me dreading self-reproach. What causes Me, or, if you please, my Will, to be identified with one side rather than the other, is that one of the Me’s represents a more permanent state of my feelings than the other does. After the temptation has been yielded to, the desiring “I” will come to an end, but the conscience-stricken “I” may endure to the end of life. (Quoted in Dennett 1984, pp. 90–91.)
See esp. Blackburn (1998, Sect. 3.3).
See, e.g., Haidt (2003).
This isn’t to say that sentiments are never directed at beliefs. The following possibilities are completely familiar: I might feel guilty for doubting that my friend will keep her promise; I might be angry at you for believing that I won’t keep my promise; a fundamentalist or ideologue might feel blame or contempt for others who fail to share her beliefs. These examples suggest that, insofar as we are metaethical sentimentalists, we should allow that it is possible to make moral judgments not only about actions, but also about beliefs. That is, because the range of negative(positive) sentiments that are associated with moral disapproval(approval) can be directed at beliefs as well as actions, metaethical sentimentalists should allow that just as one can judge that it is morally wrong(right) to act in certain ways, one can also judge that it is morally wrong(right) to hold certain beliefs.
None of this, however, should encourage somebody hoping to give a a sentimentalist account of distinctively epistemological judgments. In fact, it just makes the difficulty of that project more apparent—it’s hard to see how a sentimentalist account of normative judgments about beliefs won’t end up just telling us what it is to judge that beliefs are morally (as opposed to epistemically) permitted, required, or forbidden.
I intend “masked” to be understood in the sense in which Johnston (1992) raises the possibility of dispositions being masked.
Well, a relatively straightforward case, subject to the qualification mentioned in footnote 3.
This account of epistemic akrasia isn’t completely neutral with respect to accounts of ordinary, non-epistemological belief. In particular, views of belief on which all beliefs essentially involve language, and in which animals properly speaking don’t have beliefs, have a more difficult time allowing for a distinction between a linguistically infused belief system and a non-linguistic belief system. See, e.g., Davidson (1982). Views on which there are constitutive connections between belief and action generally, rather than belief and specifically linguistic action, have an easier time drawing such a distinction. See, e.g., Stalnaker (1984) and Dennett (1987). I set this worry aside in what follows.
See Blackburn (1993, 1998), Gibbard (1990), (2003). I don’t mean to imply that Blackburn and Gibbard take expressivism to be plausible only as an account of moral judgment; both of them are quite explicit in hoping to give broadly expressivist accounts of epistemological judgment as well, even if they devote more attention to the moral case. I also don’t mean to hold that expressivists must deny that we in fact have moral beliefs—many expressivists (including Blackburn, and more recently Gibbard) are happy to say (roughly) that any sentence in the indicative mood can express a belief, though they go on to hold that belief is a less psychologically unified category than we might have thought, and that normative beliefs differ in significant ways from non-normative beliefs.
I say merely “structurally” similar because the strategy does not involve treating epistemological beliefs as non-cognitive, or desire-like.
In recent years, many writers have explored expressivist strategies for understanding epistemological judgments, especially when “expressivist” is read in the broad way that I suggest above. See, Chrisman (2012) for a helpful survey.
Of course Schroeder considers this sort of position. He rejects it on the grounds that it requires appealing to what he calls “B-type” inconsistency, whereas he favors a methodological principle on which we should only appeal to “A-type” inconsistency. Ultimately, I think the distinction between A-type and B-type inconsistency is not as important as Schroeder thinks it is, and that it is not a significant cost of a view that it must appeal to B-type inconsistency. But I cannot defend that view here.
Assuming, at least, that you had no reason to expect me to steal your car.
Which is not to say that she might not also agree that justified false beliefs are possible. It’s just that in doing so, she’ll be making substantive epistemological claims, rather than sticking to metaepistemology.
I don’t mean to be proposing any sharp first person/third person asymmetry in the nature of beliefs about what subjects ought to believe. To unify the two cases, we might understand beliefs about what I now ought to believe as a kind of degenerate case of conditional beliefs—maybe judging that I now ought to believe that P is a kind of conditional belief l such that I now also believe l that the condition is met, and so also believe l that P.
For classic reasons to think that this is a tricky question, see Lewis (1976).
I say “almost” because of the following possibility. If one both (a) opts for a tight connection between conditional beliefs and conditional probability, and (b) endorses a ratio analysis of conditional probability (of the sort Hájek (2003) argues against), then one will have difficulty accepting my account of epistemic ought judgments. The reason is that the ratio analysis of conditional probability entails that there cannot be well-defined conditional probabilities whose conditions have zero probability. But we can make judgments about what ought to be believed given bodies of information we are sure are false—e.g., I can judge that, in a situation in which frogs have rained from the sky for 100 days in a row, one ought to believe that they’ll rain on the next day, even though I am certain that frogs have in fact not rained from the sky 100 days in a row.
The idea that we can have contingency plans for contingencies we know will not arise may seem unusual—I’m sure that I will at no point in my life be Caesar—but I don’t think we should ultimately find it so puzzling. After all, it’s utterly uncontroversial that we can have preferences that we know we will never be in a position to act upon.
For instance, it would be hasty to simply identify one’s beliefs about what one ought to believe with one’s beliefs l . We might, for instance, allow that one can believe l that P without believing that one ought to believe that P, if one thinks that one’s case is a permissive one—that is, if one thinks that one’s evidence permits one either to believe P, or to believe ∼ P. White (2005) argues against the possibility of such cases, but even if we agree with him, we might want to have some sense of what it is to think that some case is a permissive one. I take the question of what’s involved in taking a permissive attitude towards one’s beliefs to be an interesting, tricky one. I suspect that answering it will require saying something about how one interacts with others.
For the former, see Christensen (2007). For the latter, see Egan (2008). Though importantly, neither argues that being rationally ideal in all respects is compatible with inconsistent or fragmented belief. Christensen argues that rational ideals sometimes conflict, and that respecting certain ideals will sometimes require having inconsistent beliefs. Egan argues that for agents with unreliable belief-forming modules, being cognitively fragmented may be an effective form of “damage control,”—one that will stop the beliefs formed by the unreliable modules from spreading to the rest of the agent’s cognitive system.
A number of writers have discussed the principle that ought implies can specifically in the context of epistemology—some endorsing it, some attacking it. Dretske (2000, p. 598) endorses a strong epistemological ought implies can principle, and uses it to argue that we have a right to hold perceptual beliefs because we are psychologically incapable of failing to form them. William Lycan (1985, p. 146) suggests that ought implies can is less plausible in epistemology than in ethics, and expresses doubt that beliefs can be warranted in virtue of our being psychologically incapable of abandoning them. My sympathies are closer to Lycan’s than to Dretske’s. See Greco (2012, pp. 348–352).
Or perhaps less plausible, depending on our views about the putative ideal of logical consistency.
The notion of a category mistake comes from Ryle (1949).
Though in light of the observations at the end of the considerations at the end of Sect. 3, we might in some cases stop short of calling epistemic akrasia irrational, and instead hold that it is sometimes merely arational.
References
Aarnio, M. L. (Forthcoming). Higher-order evidencse and the limits of defeat. Philosophy and Phenomenological Research.
Adler, J. E. (1999). The ethics of belief: Off the wrong track. Midwest Studies in Philosophy, 23(1), 267–285.
Adler, J. E. (2002). Akratic believing? Philosophical Studies, 110(1), 1–27.
Arpaly, N. (2003). Unprincipled virtue: An inquiry into moral agency. New York: Oxford University Press.
Blackburn, S. (1993). Essays in quasi-realism. New York: Oxford University Press.
Blackburn, S. (1998). Ruling passions. Oxford: Oxford University Press.
Chrisman, M. (2012). Epistemic expressivism. Philosophy Compass, 7(2), 118–126.
Christensen, D. (2007). Does Murphy’s law apply in epistemology? self-doubt and rational ideals. In T. S. Gendler, J. Hawthorne, (Eds.), Oxford studies in epistemology, Vol. I (pp. 3–31). New York : Oxford University Press.
Coates, A. (2012). Rational epistemic akrasia. American Philosophical Quarterly, 49, 113–124.
Davidson, D. (1982). Rational animals. Dialectica, 36, 317–328.
Dennett, D. (1978). Brainstorms. Cambridge: MIT Press.
Dennett, D. (1984). Elbow room: The varieties of free will worth wanting. Oxford: Oxford University Press.
Dennett, D. (1987). The intentional stance. Cambridge: The MIT Press.
Dretske, F. (2000). Entitlement: Epistemic rights without epistemic duties?. Philosophy and Phenomenological Research, 60(3), 591–606.
Egan, A. (2008). Seeing and believing: Perception, belief formation and the divided mind. Philosophical Studies, 140(1), 47–63.
Feldman, R. (2005). Respecting the evidence. Philosophical Perspectives, 19(1), 95–119.
Field, H. (2005). Recent debates about the a priori. In T. S. Gendler & J. Hawthorne (Eds.), Oxford studies in epistemology, (Vol. I). New York: Oxford University Press.
Gendler, T. S. (2008). Alief and belief. Journal of Philosophy, 105(10), 634–663.
Gibbard, A. (1990). Wise choices, apt feelings. Cambridge: Harvard University Press.
Gibbard, A. (2003). Thinking how to live. Cambridge: Harvard University Press.
Greco, D. (2012). The impossibility of skepticism. Philosophical Review, 121(3), 317–358.
Haidt, J. (2003). The moral emotions. In R. J. Davidson, K. R. Scherer, H. H. Goldsmith, (Eds.), Handbook of affective sciences. New York: Oxford University Press.
Hájek, A. (2003). What conditional probability could not be. Synthese, 137(3), 273–323.
Hume, D. (2007). Treatise of Human Nature. Oxford: Clarendon Press.
Hurley, S. (1989). Natural reasons. Oxford: Oxford University Press.
Johnston, M. (1992). How to speak of the colors. Philosophical Studies, 68, 221–263.
Kyburg, H. (1970). Conjunctivitis. In M. Swain (Ed.), Induction, acceptance, and rational belief (pp. 55–82). Dordrecht: D. Reidel.
Lewis, D. (1976). Probabilities of conditionals and conditional probabilities. Philosophical Review, 85(3), 297–315.
Lewis, D. (1982). Logic for equivocators. Noûs, 16(3), 431–441.
Littlejohn, C. (2012). Justification and the truth-connection. New York: Cambridge University Press.
Lycan, W. G. (1985). Epistemic value. Synthese, 64(2), 137–164.
Owens, D. (2002). Epistemic akrasia. The Monist, 85(3), 381–397.
Pettit, P., Smith, M. (1996). Freedom in belief and desire. Journal of Philosophy, 93(9), 429–449.
Prinz, J. (2007). The emotional construction of morals. Oxford: Oxford University Press.
Raz, J. (2007). Reasons: practical and adaptive. SSRN eLibrary.
Ryle, G. (1949). The concept of mind. New York: Barnes & Noble, Inc.
Scanlon, T. (1998). What we owe to each other. Cambridge: Belknap Press.
Schroeder, M. (2010). Being for. Oxford: Oxford University Press.
Schwitzgebel, E. (2001). In-between believing. Philosophical Quarterly, 51(202), 76–82.
Schwitzgebel, E. (2002). A phenomenal, dispositional account of belief. Nous, 36(2), 249–275.
Schwitzgebel, E. (2008). The unreliability of naive introspection Philosophical Review, 117(2), 245–273.
Smithies, D. (2012). Moore’s paradox and the accessibility of justification. Philosophy and Phenomenological Research, 85(2), 273–300.
Stalnaker, R. (1984). Inquiry. Cambridge: The MIT Press.
Stalnaker, R. (1991). The problem of logical omniscience, I. Synthese, 89(3), 425–440.
Stalnaker, R. (1999). Context and content. Oxford: Oxford University Press.
Stanovich, K. E., West, R. F. (2000). Individual differences in reasoning: Implications for the rationality debate?. Behavioral and Brain Sciences, 23(5), 645–665.
Weatherson, B. (2008). Deontology and descartes’s demon. Journal of Philosophy, 105(9), 540–569.
Wedgwood, R. (2011). Justified inference. Synthese, 189(2),1–23.
Wedgwood, R. (2012). Justified inference. Synthese, 189, 273–295.
White, R. (2005). Epistemic \permissiveness. Philosophical Perspectives, 19(1), 445–459.
Williamson, T. (2000). Knowledge and its limits. Oxford: Oxford University Press.
Acknowledgments
Thanks to Stewart Cohen, Hartry Field, Tristram McPherson, Alejandro Pérez Carballo, Alex Silk, Agustín Rayo, Eric Schwitzgebel, Robert Stalnaker, Mike Titelbaum, Roger White, and audiences at the ANU and MIT.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Greco, D. A puzzle about epistemic akrasia. Philos Stud 167, 201–219 (2014). https://doi.org/10.1007/s11098-012-0085-3
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-012-0085-3