Destructive defeat and justificational force: the dialectic of dogmatism, conservatism, and meta-evidentialism


Defeaters can prevent a perceptual belief from being justified. For example, when you know that red light is shining at the table before you, you would typically not be justified in believing that the table is red. However, can defeaters also destroy a perceptual experience as a source of justification? If the answer is ‘no’, the red light defeater blocks doxastic justification without destroying propositional justification. You have some-things-considered, but not all-things-considered, justification for believing that the table is red. If the answer is ‘yes’, the red light defeater blocks doxastic justification by destroying propositional justification. You have neither all-things-considered nor some-things-considered justification for believing that the table is red. According to dogmatism, the justificational force of perceptual experiences is indestructible. According to conservatism about sense experience, a perceptual experience ceases to have justificational force if there is evidence against its reliability. Finally, according to meta-evidentialism, a perceptual experience is blocked from being a source of justification is there is no evidence of its reliability. I argue that, of these three theories, meta-evidentialism is the most plausible.

This is a preview of subscription content, access via your institution.


  1. 1.

    Roderick Chisholm and John Pollock were the first to highlight the importance of defeaters. See Chisholm (1966, p. 48), Pollock (1974, p. 40ff; 1986, p. 38f).

  2. 2.

    See Pollock (1974, p. 40).

  3. 3.

    Consider Pollock’s definition of a defeater: “If P is a reason for S to believe Q, R is a defeater for this reason if and only if R is logically consistent with P and (P&R) is not a reason for S to believe Q” (Pollock 1986, p. 38). According to this definition, R’s being a defeater entails that the conjunction of P and R fails to be a reason for S to believe Q. This leaves open whether P by itself continues to be a reason for S to believe Q. However, on p. 40 in Pollock (1974), we find the following passage: “An inductive reason for accepting a generalization can be defeated on at least two grounds. First, no matter how strong the initial inductive evidence for the generalization, if further investigation reveals a counterexample then the original reason ceases to be a good reason. Second, if it is discovered that the sample on which the original generalization was based was not a fair sample, this will make the initial reason no longer a good reason even though it was a good reason until this was discovered.” [My italics.] If we assume that a good reason is a source of justification and a bad reason is not, then this passage clearly suggests that defeated reasons no longer have justificational force.

  4. 4.

    For an example of a hybrid theory of perceptual justification, see Brogaard (2013).

  5. 5.

    See Alston (1999, p. 235).

  6. 6.

    Huemer (2001, p. 99).

  7. 7.

    Pryor (2000).

  8. 8.

    Here are two passages that define dogmatism in terms of Sufficiency. The first is by Elia Zardini: “Dogmatism about perceptual justification holds ...  that one’s having a (perceptual) experience as though P suffices to provide one with an at least prima facie justification for believing that P” (Zardini 2014, p. 35). The second is by Chris Tucker: “Necessarily, if it seems to S that P, then S thereby has prima facie (non-inferential) justification for P” (Tucker 2010, p. 529).

  9. 9.

    See Foley (1983) and Fumerton (2007).

  10. 10.

    The difference between dogmatism and \(\hbox {conservatism}_\mathrm{SE}\) is subtle and not always fully appreciated. For example, Huemer’s phenomenal conservatism asserts that a seeming as if p is sufficient for having defeasible justification for believing p. If we are strict about identifying conservatism with a theory that treats justificational force as a defeasible (i.e. not indestructible) default status, then Huemer’s view is an instance of dogmatism, not conservatism. Or consider the following characterization of dogmatism by Tucker: “If it perceptually seems to S that P, then, in the absence of defeaters, S thereby has justification to believe P” (Tucker 2013, p. 2). Although Tucker means to define dogmatism, what his definition expresses is \(\hbox {conservatism}_\mathrm{SE}\). According to the view Tucker defines, justificational force is conditional upon the absence of defeaters. If defeaters are present, S does not have justification for believing that P. This is \(\hbox {conservatism}_\mathrm{SE}\), not dogmatism. It is crucial, then, to distinguish between (A) if S has an experience as of p, then S has defeasible justification for believing p, and (B) if S has an experience as of p and (destructive) defeaters are absent, then S has justification for believing p that is defeasible via undermining or rebutting. (A) is dogmatism, (B) is \(\hbox {conservatism}_\mathrm{SE}\).

  11. 11.

    Alston takes Chisholm to be a dogmatist about perceptual justification (Alston 1999, p. 235). The view he ascribes to Chisholm is that perceptually taking there to be a tree is sufficient for having prima facie justification for believing there to be a tree. However, on p. 48 of Chisholm (1989), we find the following passage: “The assumption is that, occasionally at least, the senses provide us with evidence pertaining to the existence of such things as trees, ships, and houses. The best answer to the question, “What is the nature this evidence?” seems to be this: the fact that we are appeared to in certain ways tends to make it evident that there is an external thing that is appearing to us in those ways. And the fact that we take there to be a tree tends to make it evident for us that there is a tree that we perceive.” (Italics by Chisholm.) Two points deserve emphasis: First, Chisholm asserts that the senses give us evidence only occasionally. Second, he emphasizes that the senses merely tend to give us evidence about external objects. This suggests Chisholm might have agreed that the senses do not always give us evidence. He might actually have been a conservative rather than a dogmatist about perceptual experiences as a source of justification.

  12. 12.

    In Tucker’s 2013 collection Seemings and justification, ‘conservatism’ typically refers to the dogmatic view Huemer and Pryor have defended. In contrast, in the collection Sceptism & Perceptual Justification (Dodd and Zardini 2014), the term ‘conservatism’ refers to views opposed to dogmatism. For an influential version of the latter type of conservatism, see Wright (2004, (2014).

  13. 13.

    See Feldman (2003, p. 144).

  14. 14.

    Two points of clarification: First, I’m going to use the phrase ‘has evidence of x’s reliability’ as a short for saying ‘has a body of evidence that in its totality justifies the ascription of reliability to x’. Reference to the subject’s total evidence is necessary because defeated evidence of reliability does nothing towards making a perceptual experience a source of justification. Second, like conservatism, meta-evidentialism requires the absence of evidence of unreliability. This is to be understood as the condition that the subject’s experience is not unreliable in light of the subject’s total evidence. Since having a body of evidence that in its totality justifies the ascription of reliability entails this condition, it need not be stated separately.

  15. 15.

    See BonJour (1985) and Lehrer (1990).

  16. 16.

    Moreover, since meta-evidentialism allows for a perceptual belief to be basic in the sense that it can be justified without receiving support from any beliefs, meta-evidentialism qualifies as a version of foundationalism. At the same time, though, since meta-evidentialism makes immediate justification impossible, there is also a sense in which the view qualifies as a (non-doxastic) version of coherentism.

  17. 17.

    Given these stipulations, the following is possible: although you have a little bit of evidence that your sense experience as of p is unreliable, your experience nevertheless gives you stc justification for believing p.

  18. 18.

    Huemer and Pryor say things suggesting that they might actually not be opposed to \(\hbox {conservatism}_\mathrm{SE}\). For example, on p. 100 in his 2001, Huemer explicitly endorses the spirit of conservatism. This suggests to me he might actually accept that a perceptual experience is a source of justification unless one has reason to think otherwise. On p. 354 of his 2004, Pryor suggests that, according to a liberal treatment of a visual experiences of your car, for such experiences to give you justification for believing your car is present, “it’s enough that you lack reason to believe your experiences are unreliable.” If that’s liberalism, then liberalism is to be identified with not dogmatism but \(\hbox {conservatism}_\mathrm{SE}\) instead. Finally, on p. 96 in his 2013, Pryor says: “When I began several years ago to use the term dogmatism, I meant the view that ...  justification is sometimes both immediate and underminable.” In this passage, Pryor identifies dogmatism with the view that defeasible justification can be immediate. Thus understood, dogmatism is not committed to Sufficiency and therefore consistent with \(\hbox {conservatism}_\mathrm{SE}\).

  19. 19.

    See Markie (2013, p. 257). The case is meant to be a counterexample to Huemer’s phenomenal conservatism. For Huemer’s reply, see Huemer (2013, p. 343f).

  20. 20.

    In describing the case as involving a defeater, I deviate from the way Markie construes the case. According to Markie, if defeaters are taken to be mental states, it remains unclear whether Virgil has a defeater. See Markie (2013, p. 258). We could of course imagine Virgil to be suffering from the delusion of being expert, where this delusion is realistic enough to justify him in this belief. Huemer replies (correctly, I think) that, if interpreted in such a way, the case does not challenge phenomenal conservatism (or, for that matter, dogmatism), just as BIV scenarios do not. For the case to be an effective problem case for phenomenal conservatism, we must image Virgil to have a defeater. If we imagine him to be an unreliable novice, it is hard to see how he can fail to have a defeater. And if we imagine him to be aware of his strong desire to find gold, then he would have a second defeater.

  21. 21.

    How might you acquire knowledge of being envatted? Imagine, for example, that on occasion you are having perceptual experiences completely incongruous with those you have most of the time. In these experiences, while feeling strangely disembodied and helpless, you see a large lab with technicians in white coats attending to multiple brains floating in their individual vats.

  22. 22.

    As an anonymous referee suggested, it might be objected that Joe’s testimony provides at least some tiny bit of justification for believing p because, given Joe’s assertion, it is now just a tiny bit more probable that p is true. I discuss this type of objection in the next section.

  23. 23.

    An anonymous referee suggested that my argument here counts as well against meta-evidentialism. According to the referee, I claim it’s arbitrary to endorse dogmatism about sense experience while rejecting dogmatism about witnesses and gauges. At the same time, the position I defend, meta-evidentialism, forces me to say it’s not arbitrary to endorse self-support coming from sense experience and memory while rejecting self-support provided by witnesses and gauges. In reply, I deny that meta-evidentialism cannot allow for witnesses and gauges to provide self-support. When interviewing a witness and asking the right kind of questions in the right way, it is possible to acquire from the witness evidence supporting the witness’s reliability. Likewise, it is possible for gauges to provide self-support. Think of a gas gauge equipped with a device to monitor its reliability, consisting of sophisticated sensors and electronics. This device monitors every way in which a gas gauge might fail in relevantly close scenarios. As long as the gas gauge functions reliably, a green light is on. In case of malfunction, the light turns red. Arguably, a sophisticated gauge like that provides its user with (epistemically valuable) self-support.

  24. 24.

    This reply on behalf of dogmatism was suggested by an anonymous referee.

  25. 25.

    For a clear articulation of the path-to-truth conception of justification, see BonJour (1985, p. 8f). See also what Alston says about the ‘epistemic point of view’: Alston (1989, p. 83f).

  26. 26.

    This reply on behalf of dogmatism was also suggested by an anonymous referee.

  27. 27.

    De Rose applied the term ‘abominable conjunction’ to statements such as “I know I have hands but I don’t know I’m not a BIV.” See his 1995.

  28. 28.

    An anonymous referee suggested that there are two ways to defend the Never Zero Argument. First, it might be argued that fallibilism mandates the Never Zero view. For if fallibilism is true, it’s possible to know p even though one has a little bit of evidence that p is false. For example, I might know that p even though a witness I know to be lying asserts \({\sim }p\). In response, I do not agree that fallibilism mandates the Never Zero view. According to fallibilism, knowledge does not require truth-entailing evidence. Endorsing fallibilism in this sense is consistent with denying that, in the case in question, knowledge of p comes together with some justification for believing \({\sim }p\). According to the second way of defending the Never Zero Argument, the assertions I claim are abominable are in fact true but not felicitously assertible because they generate false implicatures. But what might the false implicatures be? Unless a convincing story is told about which false implicatures the assertions in question generate, the true but unassertible maneuver remains implausible.

  29. 29.

    The argument presented here suggests a way to identify when a defeater destroys perceptual justification and when a defeater merely undermines perceptual justification without destroying it. Whether a defeater destroys justificational force or merely undermines might depend on whether the defeater is known. For example, when you know that you are hallucinating, perceptual justification is destroyed. But when you have only weak evidence that you are hallucinating, evidence falling short of giving you knowledge, then the defeater does not destroy the justificational force of your experience but merely undermines it, thus preventing you from having all things considered justification.

  30. 30.

    An anonymous referee suggested that there is a Moorean response to this. It goes as follows. Both dogmatism and \(\hbox {conservatism}_\mathrm{SE}\) allow for a cognitive structure that includes plenty of evidence of perceptual reliability. Both views deny that the justificational force of sense experience comes from such evidence. Instead, it has its sole origin immediately justified perceptual beliefs. Via inferences from such beliefs, evidence for perceptual reliability can be generated via inference. Therefore, both views are capable of satisfying the path-to-truth requirement. I offer two brief replies. First, the proposed inference from immediately justified perceptual beliefs to justified belief in perceptual reliability runs into one of the main liabilities of both dogmatism and \(\hbox {conservatism}_\mathrm{SE}\): the problem of easy knowledge, as highlighted in Cohen (2002). Second, if dogmatists and SE-conservatives accept justification closure, they must hold, given awareness of the entailment from, say, hand possession to non-envatment, that justification for believing I’m not a BIV (non-envatment-J) is necessary for justification for believing I have hands (hands-J). The claim in question is that hands-J does not come from or rest on non-envatment-J although non-envatment-J is necessary for hands-J. According to this line of reasoning, there is a coming-from or resting-on relation over and above the necessity relation at work in justification-closure. It is unclear what this additional relation is supposed to be. Advocates of dogmatism or SE-conservatism need to tell a convincing story of why we should think that the following is true: non-envatment J is necessary for hands-J but hands-J does not rest on or come from non-envatment-J. Without such a story, the claim is a verbal maneuver without a substantive core. On behalf of the claim, it might be suggested that, for example, blood-supply to the brain is necessary for justified belief but surely a belief’s justification doesn’t come from blood that’s circulated through the brain. The idea is that an analogous point can be made for Hands-J and non-envatment-J. Although the point about blood-supply to the brain is well taken, in this example we have a contrast between an epistemic property and a non-epistemic condition, whereas when we consider hands-J and non-envatment-J, we are looking at two different elements within an overall epistemic structure. I do not see, therefore, that pointing to non-epistemic necessity relations helps to render the claim in question plausible.

  31. 31.

    Thanks to Paul Draper for suggesting the lottery-type aspect of this case. For a similar example, see the echolocation case in Steup (2004).

  32. 32.

    Suppose, upon experience the trumpet, Ted forms the belief that there is an actual, external trumpet sound he is hearing. What should dogmatists and SE-conservatives say about the justificational status of Ted’s belief? It would be tempting for them to say that Ted’s belief is not justified because it is defeated. But what’s the defeater? Clearly there is no rebutting defeater. If Ted has as defeater, it would have to be an underminer. Now, undermining defeaters attack the reliability of the belief source. Since Ted has no evidence to think that his experience is unreliable, it would seem he lacks an undermining defeater. But if he has neither a rebutting nor an undermining defeater, it follows that his belief is justified. Given the details of the case, this does not look like a plausible outcome.

  33. 33.

    See BonJour (1985, p. 117ff).

  34. 34.

    In Steup (2013), I argue that epistemic circularity can be benign. See also Vogel (2014). He argues that it’s possible for evidence E to be evidence for the proposition that E is not misleading.

  35. 35.

    See Alston (1993).


  1. Alston, W. (1989). Epistemic justification. Essays in the theory of knowledge. Ithaca, NY: Cornell University Press.

    Google Scholar 

  2. Alston, W. (1993). The reliability of sense perception. Ithaca, NY: Cornell University Press.

    Google Scholar 

  3. Alston, W. (1999). Perceptual knowledge. In J. Greco & E. Sosa (Eds.), The Blackwell guide to epistemology (pp. 223–242). Malden, MA: Blackwell.

    Google Scholar 

  4. BonJour, L. (1985). The structure of empirical knowledge. Cambridge: Harvard University Press.

    Google Scholar 

  5. Brogaard, B. (2013). Phenomenal seemings and sensible dogmatism. In C. Tucker (Ed.), Seemings and justification (pp. 270–289). New York: Oxford University Press.

    Chapter  Google Scholar 

  6. Chisholm, R. (1966). Theory of knowledge (1st ed.). Englewood Cliffs, NJ: Prentice Hall.

    Google Scholar 

  7. Chisholm, R. (1989). Theory of knowledge (3rd ed.). Englewood Cliffs, NJ: Prentice Hall.

    Google Scholar 

  8. Cohen, S. (2002). Basic knowledge and the problem of easy knowledge. Philosophy and Phenomenological Research, 65, 309–329.

    Article  Google Scholar 

  9. DeRose, K. (1995). Solving the skeptical problem. The Philosophical Review, 104, 1–52.

    Article  Google Scholar 

  10. Dodd, D., & Zardini, E. (Eds.). (2014). Scepticism & perceptual justification. Oxford: Oxford University Press.

    Google Scholar 

  11. Feldman, R. (2003). Epistemology. Upper Saddle River, NJ: Pearson.

    Google Scholar 

  12. Foley, R. (1983). Epistemic conservatism. Philosophical Studies, 443, 165–182.

    Article  Google Scholar 

  13. Fumerton, R. (2007). Epistemic conservatism: Theft or honest toil? In T. S. Gendler & J. Hawthorne (Eds.), Oxford studies in epistemology (pp. 63–86). Oxford: Oxford University Press.

    Google Scholar 

  14. Huemer, M. (2001). Skepticism and the veil of perception. Lanham, MD: Roman and Littlefield.

    Google Scholar 

  15. Huemer, M. (2013). Phenomenal conservatism über alles. In C. Tucker (Ed.), Seemings and justification (pp. 328–349). New York: Oxford University Press.

    Chapter  Google Scholar 

  16. Lehrer, K. (1990). Theory of knowledge. Boulder: Westview Press.

    Google Scholar 

  17. Markie, P. (2013). Searching for true dogmatism. In C. Tucker (Ed.), Seemings and justification (pp. 248–269). New York: Oxford University Press.

    Chapter  Google Scholar 

  18. Pollock, J. (1974). Knowledge and justification. Princeton, NJ: Princeton University Press.

    Google Scholar 

  19. Pollock, J. L. (1986). Contemporary theories of knowledge. Totowa, NJ: Rowman and Littlefield.

    Google Scholar 

  20. Pryor, J. (2000). The skeptic and the dogmatist. Nous, 34, 517–549.

    Article  Google Scholar 

  21. Pryor, J. (2004). What’s wrong with Moore’s argument? Philosophical Issues, 14, 349–378.

    Article  Google Scholar 

  22. Pryor, J. (2013). Problems for credulism. In C. Tucker (Ed.), Seemings and justification (pp. 89–131). New York: Oxford University Press.

    Chapter  Google Scholar 

  23. Steup, M. (2004). Internalist reliabilism. Philosophical Issues, 14, 404–425.

    Article  Google Scholar 

  24. Steup, M. (2013). Is epistemic circularity bad? Res Philosophica, 90, 215–235.

    Article  Google Scholar 

  25. Tucker, C. (2010). Why open-minded people should endorse dogmatism. Philosophical Perspectives, 24, 529–545.

    Article  Google Scholar 

  26. Tucker, C. (Ed.). (2013). Seemings and justification. New York: Oxford University Press.

    Google Scholar 

  27. Vogel, J. (2014). E & \({\lnot }\)H. In D. Dodd & E. Zaardini (Eds.), Scepticism and perceptual justification (pp. 87–107). Oxford: Oxford University Press.

    Chapter  Google Scholar 

  28. Wright, C. (2004). On epistemic entitlement: Warrant for nothing (and foundations for free?). Aristotelian Society Supplement, 78, 167–212.

    Article  Google Scholar 

  29. Wright, C. (2014). On epistemic entitlement (2): Welfare state epistemology. In D. Dodd & E. Zardini (Eds.), Scepticism & perceptual justification (pp. 213–247). Oxford: Oxford University Press.

    Chapter  Google Scholar 

  30. Zardini, E. (2014). Confirming the less likely, discovering the unknown: Dogmatisms—Surd and doubly surd, natural, flat and doubly flat. In D. Dodd & E. Zardini (Eds.), Scepticism & perceptual justification (pp. 33–70). Oxford: Oxford University Press.

    Chapter  Google Scholar 

Download references


I wish to thank two anonymous referees for their excellent feedback on this paper.

Author information



Corresponding author

Correspondence to Matthias Steup.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Steup, M. Destructive defeat and justificational force: the dialectic of dogmatism, conservatism, and meta-evidentialism. Synthese 195, 2907–2933 (2018).

Download citation


  • Justificational force
  • Perceptual justification
  • Rebutting defeaters
  • Undermining defeaters
  • Destructive defeaters
  • Conservatism
  • Dogmatism
  • Meta-evidentialism