This is a preview of subscription content, log in via an institution to check access.
Notes
A further, very important question is what significance this challenge has for the question of whether skepticism is true. I will not broach that further question here.
The name was first used, I believe, in (DeRose 1995).
This presentation will idealize away from possible concerns about degrees of belief or credence and work with an on/off notion of belief.
In (Leite forthcoming1) I offer a full-dress discussion of this form of response to the dream-argument for external world skepticism.
I’m grateful to Katy Abramson for highlighting this way of seeing my project.
In Nozick’s formulation of the Sensitivity Requirement (Nozick 1981, p. 179), the way or method by which the belief was formed must be held fixed across the actual and counterfactual cases. It might be objected that in my example the method of belief formation isn’t held fixed, because one’s perceptual experiences would be quite different in the two cases. However, there are three points to be made here. First, there is a straightforward (not ad hoc or jury-rigged) characterization on which the method of belief formation is the same in the two cases. Second, Nozick’s own examples and his explanation of why, e.g., one’s current belief that one has hands meets the Sensitivity Requirement, allow for differences of this sort between the cases. Third, if we required that the relevant counterfactual case be one in which one’s current sensory experience is introspectively indistinguishable from the actual case but the relevant proposition is false, then many of our ordinary beliefs that we take to constitute knowledge wouldn’t meet the Sensitivity Requirement and so wouldn’t count as knowledge.
If there is a general principle at play in this argument, it is something along the following lines: if p is false, then p isn’t evidence for anything. A principle along such lines appears to me to accord with our ordinary practice. However, Branden Fitelson has offered a putative counterexample. Suppose that a reliable source tells me that p, and I thereby (reasonably) come to believe that p. Recognizing that p entails some q, I then reasonably infer q from p and thereby take p to be evidence for q in this case. Branden comments, “I’m inclined to say that p could be evidence for q (for me) in some such contexts—even if it turns out that p is false (especially, if I’m—in principle—never in a position to know that p is false)” (personal communication; Jon Kvanvig has suggested something similar to me as well). However, there is a slide away from ordinary ways of thinking if one takes such examples in the suggested way. We should grant that one could reasonably take p to be evidence in this sort of case. But note the shift in terminology—the relativization to a person—involved in the suggestion that p could “be evidence for q (for me)” in such a case. This may well just mean the same thing as “S reasonably takes p to be evidence for q” or “S would be reasonable to treat p as evidence for q”—which is compatible with p’s not being evidence for q. But even if that is not what’s meant, such formulations are not equivalent to saying that p is evidence for q simpliciter. And we wouldn’t ordinarily say things like, “Well, p is evidence for S, but not for me, since (I know) p is false.” Rather, we’d say, “S reasonably thinks p is evidence for q, but he’s wrong.” I conclude that evidence-for-a-person is a technical, theoretical notion aimed at helping us explain the sense in which someone can be reasonable in certain sorts of cases. Confusion between this technical notion and our ordinary notion of evidence is facilitated, I think, by a commitment to the thought that if S’s belief is reasonable, then it must be based on good evidence, or at least S must have good evidence for it. But the judgments involved in our ordinary practice do not themselves show a commitment to that thought.
This is the core suggestion behind Ram Neta’s rule “R!” in (Neta 2002, p. 674). Neta’s own formulation is considerably more complicated, for reasons arising from the details of his theory of evidence.
As is familiar from the literature, a really plausible closure principle would need to be refined in various ways. Such refinements are irrelevant to my discussion, since the issue that I will focus on equally arises for any more refined principle, so long as it is simply a closure principle—not a transmission principle to the effect that the knowledge appearing on the consequent side derives from and is inferentially based upon the knowledge appearing on the antecedent side. (David and Warfield (2008) provide additional reasons why it is doubtful that the skeptical argument can be successfully underwritten by a more refined closure principle.).
In Stroud’s version of the skeptical argument, the closure principle is replaced by this: if you know that q is incompatible with your knowing that p, then in order to know that p, you must know that not-q (Stroud 1984, pp. 29–30). (This principle is designed to allow for the fact that the skeptical hypothesis apparently need not be incompatible with the truth of all of the relevant beliefs.) This principle likewise fails to yield an unsatisfiable requirement when applied to the case of one’s knowledge that one is not the victim of an evil demon, and for the same reason.
Stroud’s principle may have been motivated by the thought that in order to know that p, one must be able to rule out all alternatives that one knows to be incompatible with knowing that p—where ruling out is understood to mean being able to determine that those alternatives do not obtain only on the basis of evidence that one would have even if they did obtain. This thought would do the needed work. However, it is not true that in order to know that a possibility doesn’t obtain one has to be able to rule it out in this sense, as is shown by the example regarding the children of brunettes discussed in Sect. 4.
Obviously, what is needed here needn’t be a temporal requirement, but a requirement stating a precondition or some other relation of asymmetric epistemic dependence. To put it formally, the requirement will have to be formulable using an irreflexive relational predicate such that the two demands
aRb and bRa
are not co-satisfiable when a is not identical with b. I use the phrase, “antecedently know,” to formulate this logical or structural feature of the requirement. For my purposes here and in what follows, we don’t need to focus on a particular substantive theory of epistemic priority, but only on this formal structure (though I will discuss some factors that might be thought to impose epistemic priority relations).
This thought will be unavailable if one understands every episode of critical reflection or deliberation about a possibility on the model of performing an explicit inference. I’m trying to bring out that that is a mistake.
Or a priority version of Stroud’s variant (see fn. 13).
Here’s a simple example which provides a template applicable to any proposition you like. Suppose that you are looking straight at a cow in normal conditions. You recognize that it’s being a cow implies that it is not a chicken. So according to this principle, to know that it is a cow, you must antecedently know that it is not a chicken. Now consider your knowledge that it is not a chicken. You recognize that its not being a chicken implies that it is not a chicken. So according to this principle, to know that it is not a chicken, you must antecedently know that it is not a chicken. But that’s impossible. So you can’t know it’s not a chicken. And so you can’t know it’s a cow.
The implausibility of this principle has also been noted by James Pryor (2000) and Byrne (2004). I first articulated the problem in my undergraduate senior honors thesis (UC Berkeley, 1992).
It should be obvious that the same points apply to Stroud’s variant of the familiar closure principle as well.
In this paragraph and what follows, I mean the term “ground” to include anything that one might want to allow as an epistemically acceptable basis for a belief. (I do assume, however, that practical or prudential reasons won’t count.) I likewise mean the phrase “reliability, competence, or authority” to be understood in a pretheoretical way.
The requirement formulated in the main text is vulnerable to facile counterexamples: it has the consequence that one ought not believe that one is not dead, for instance. (I am indebted here to Kurt Ludwig.) Such counterexamples do not cut to the heart of the matter, however; nothing of importance will turn on tidying up the requirement in this respect, so long as it is formulated in a way that would enable it to do the work needed for the skeptical argument. The same goes for refining the requirement to focus upon hypotheses that one recognizes to be such that their truth would undermine one’s reliability, competence, or authority regarding some domain D.
There is now an extensive literature concerning what exactly goes wrong in instances of reasoning like this one. On my view, the problem arises because of the commitments that are undertaken when one forms a belief by consciously and explicitly inferring from some evidence (Leite 2008 and forthcoming2). For some other approaches, see Crispin Wright (2003, 2004, 2007) and the literature regarding Stewart Cohen’s so-called “problem of easy knowledge” (Cohen 2002, 2005). For a dissenting voice, see Pryor (2004, forthcoming).
This is roughly the view proposed in (Foley 1987).
It should be noted that to be relevant at all, the examples must be understood in such a way that the person can’t even provide track-record considerations in favor of believing as she does.
To get the example exactly right, we’d have to fiddle to adjust for indexicals, any externally determined contents, etc. I assume for the sake of argument that this can be made to work.
Stroud focuses on the question, “How does anyone know anything about the world at all?”, whereas Bonjour focuses on the question, “Do we have any good reasons for thinking that our beliefs about the world … are true?” Another question that would have much the same effect is, “What, if anything, am I to believe about the world?”.
For a parallel, suppose that someone asked you to find out, by sight, how many items of furniture are in a completely dark room. When you fail, it would be a mistake to conclude that there is no furniture in the room or even that you can’t find out whether there is. To find out whether there is, you’d either need to turn on the lights or use something other than your eyes. As with the analogy, so too regarding the question about reasons.
Notably, Bonjour (1985 pp. 179–188 and 2003 pp. 92–96).
Compatible, that is, so long as one doesn’t follow the “Moorean Dogmatist” in assuming that there would be nothing epistemically objectionable about a fully rational and reflective deliberator consciously and explicitly inferring from her sensory experiences to such conclusions as that she has hands and from there that she is not a disembodied spirit deceived by an evil demon—thus arriving at a belief on the matter for the first time. (I argued above that parallel reasoning, exhibiting the same structure, would be unacceptable in the George W. Bush case (see Sect. 4 above).) However, the doctrine of immediate experiential justification does not force the additional claim about acceptable inference or reasoning. See my “Immediate Warrant, Epistemic Responsibility, and Moorean Dogmatism” (forthcoming).
I would like to thank Gary Ebbs, Branden Fitelson, Peter Graham, Mark Kaplan, Kirk Ludwig, Penelope Maddy, Berislav Marusic, Wallace Matson, Ram Neta, Fred Schmitt, Matthias Steup, and Jonathan Weinberg for comments and conversations relating to this paper. I am grateful, too, to the audience at an invited paper session at the 2009 Pacific Division American Philosophical Association meeting and to my commentators, Al Casullo and Jon Kvanvig. My greatest debt is to Katy Abramson, without whom these thoughts—some 15 years in gestation—would never have made it onto paper.
References
Bonjour, L. (1985). The structure of epistemic justification. Cambridge, MA: Harvard University Press.
Bonjour, L., & Sosa, E. (2003). Epistemic justification: Internalism vs. externalism, foundations vs. virtues. Oxford: Blackwell.
Byrne, A. (2004). How hard are the sceptical paradoxes? Noûs, 38, 299–325.
Chisholm, R. M. (1989). Theory of knowledge (3rd ed.). Englewood Cliffs: Prentice-Hall.
Cohen, S. (1984). Justification and truth. Philosophical Studies, 46(1984), 279–296.
Cohen, S. (2002). Basic knowledge and the problem of easy knowledge. Philosophy and Phenomenological Research, 65(2), 309–329.
Cohen, S. (2005). Why basic knowledge is easy knowledge. Philosophy and Phenomenological Research, 70(2), 417–430.
Cohen, S., & Lehrer, K. (1983). Justification, truth, and knowledge. Synthese, 55, 191–207.
David, M., & Warfield, T. (2008). Knowledge-closure and skepticism. In Q. Smith (Ed.), Epistemology: New essays. New York, NY: Oxford University Press.
DeRose, K. (1995). Solving the skeptical problem. The Philosophical Review, 104, 1–52.
Foley, R. (1987). The theory of epistemic rationality. Cambridge, MA: Harvard University Press.
Goldman, A. (1999). Internalism exposed. Journal of Philosophy, 96(6), 271–293.
Leite, A. (2008). Believing one’s reasons are good. Synthese, 161(3), 419–441.
Leite, A. (forthcoming1). Austin, dreams, and skepticism. In: R. Sorli & M. Gustafsson (Eds.), New essays on the philosophy of J. L. Austin. New York, NY: Oxford University Press.
Leite, A. (forthcoming2). Immediate warrant, epistemic responsibility, and moorean dogmatism. In A. Reisner & A. Steglich-Petersen (Eds.), Reasons for belief.
Maddy, P. (2007). Second philosophy. New York, NY: Oxford University Press.
Moore, G. E. (1993). Proof of an external world. In T. Baldwin (Ed.), G. E. Moore: Selected writings. London: Routledge.
Neta, R. (2002). S knows that p. Noûs, 36, 4, 663–4, 681.
Nozick, R. (1981). Philosophical explanations. Cambridge, MA: Harvard University Press.
Pritchard, D. (2005). Epistemic luck. New York, NY: Oxford University Press.
Pryor, J. (2000). The skeptic and the dogmatist. Noûs, 34, 517–549.
Pryor, J. (2004). What’s wrong with Moore’s argument. Philosophical Issues (a supplement to Nous), 14, 349–378.
Pryor, J. (forthcoming). When warrant transmits. In: A. Coliva (Ed.), Wittgenstein, epistemology and mind: Themes from the philosophy of Crispin Wright. New York, NY: Oxford University Press.
Sosa, E. (2007). A virtue epistemology. New York, NY: Oxford University Press.
Sosa, E. (2009). Reflective knowledge. New York, NY: Oxford University Press.
Stroud, B. (1984). Significance of philosophical scepticism. New York, NY: Oxford University Press.
Stroud, B. (2002). Understanding human knowledge in general. In: B. Stroud (Ed.), Understanding human knowledge: Philosophical essays (pp. 99–121). Oxford: Oxford University Press.
Williams, B. (1978). Descartes: The project of pure enquiry. Harmondsworth: Pelican.
Williamson, T. (2000). Knowledge and its limits. New York, NY: Oxford University Press.
Wright, C. (2003). Some reflections on the acquisition of warrant by inference. In S. Nuccetelli (Ed.), New essays on semantic externalism and self-knowledge (pp. 57–77). Cambridge, MA: MIT Press.
Wright, C. (2004). Warrant for nothing (and foundations for free?). Aristotelian Society Supplement, 78, 167–212.
Wright, C. (2007). The perils of dogmatism. In S. Nuccetelli & G. Seay (Eds.), Themes from G. E. Moore: New essays in epistemology and ethics. New York, NY: Oxford University Press.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Leite, A. How to take skepticism seriously. Philos Stud 148, 39–60 (2010). https://doi.org/10.1007/s11098-010-9502-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-010-9502-7