Skip to main content

Evidentialism, circularity, and grounding


This paper explores what happens if we construe evidentialism as a thesis about the metaphysical grounds of justification. According to grounding evidentialism, facts about what a subject is justified in believing are grounded in facts about that subject’s evidence. At first blush, grounding evidentialism appears to enjoy advantages over a more traditional construal of evidentialism as a piece of conceptual analysis. However, appearances are deceiving. I argue that grounding evidentialists are unable to provide a satisfactory story about what grounds the evidential facts, and that this provides good reason to reject grounding evidentialism.

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


  1. Typically, grounding is taken to be a form of a metaphysical dependence corresponding to the “in virtue of” relation: a fact F1 grounds a fact F2 iff F2 obtains in virtue of F1. For recent work on grounding, see Fine (2001, 2012), Schaffer (2009, 2012), Rosen (2010), Raven (2012), Clark and Liggins (2012), (deRosset 2013), and Trogdon (2013a).

  2. See e.g. (Fine 2001), Rosen (2010), and Schaffer (2009).

  3. For the purposes of this paper, I’ll follow Conee and Feldman in taking sentences of the form, “S is justified in believing p” to ascribe propositional rather than doxastic justification. However, I’ll remain noncommittal on how exactly to understand the relation between the two.

  4. Other epistemologists formulate evidentialism differently; however, many alternative statements of evidentialism also take the form of necessary biconditionals. For instance, Fantl and McGrath (2002) and Weatherson (2005) characterize evidentialism as the view that, for any two subjects S and S′, necessarily, if S and S′ have the same evidence for/against p, then S is justified in believing that p iff S′ is, too.

  5. Other authors also talk of “determination” when characterizing evidentialism. See, for instance, Ganson’s characterization of evidentialism as “the view that facts about whether or not an agent is justified in having a particular belief are entirely determined by facts about an agent’s evidence.” (2008, p. 441).

  6. Goldman relies here on Kelly 2006 overview of various conceptions of evidence.

  7. It’s worth noting that the primary defender of E = K denies that the equation needs to be understood as a conceptual analysis. (Williamson 2000, p. 186).

  8. Of course, evidentialists might follow Williamson in denying that the concept knowledge is analyzable at all, let alone analyzable in terms of the concept justification. In a footnote, Goldman contends that “A commitment to defining ‘knowledge’ in terms of ‘justification’ is clear at least for Feldman, who writes: ‘knowledge requires justified true belief that does not essentially depend upon a falsehood.’ (Feldman 2003, p. 37)” (Goldman 2011, p. 394). However, I don’t think that this quotation—taken by itself—warrants Goldman’s attribution to Feldman of the position that “knowledge” is only definable in terms of “justification.” After all, one could hold that “knowledge” is undefinable, but that we can nonetheless formulate at least some necessary conditions for knowledge—conditions that include truth and justification. (Indeed, this is probably Williamson’s view on the matter).

  9. Conee and Feldman’s reply to Goldman strikes me as puzzling for a couple of reasons. First, not all of their definitions are equivalent. Hence it’s not clear whether Conee and Feldman think that something is evidence iff it satisfies all of these definitions, or if they think a sufficient condition for something to be evidence is for it to satisfy one of these definitions. Second, some of their definitions (e.g. “an indication, mark, sign”) seem closely akin to Goldman’s proposal that we analyze evidence in terms of reliable indication—a proposal that they explicitly reject. (Perhaps the difference lies in the fact that the definitions they provide only mention indication, not reliable indication. However, it’s not at all obvious to me that an entity e can be an indication of a truth t without being a reliable indication of t.) In light of these perplexing features, I won’t commit myself to any particular interpretation of Conee and Feldman’s response.

  10. In a similar vein, Weatherson considers the possibility that the concept evidence can only be spelled out in a “massively disjunctive” clause of the form: “p is part of our evidence iff we see that p or hear that p or smell that p, or, etc.” (2009, p. 7) Weatherson complains that such an account would be “unfortunate”, since (i) it would possess the ugliness of all disjunctive theories, (ii) it would be hard to tell where, exactly, to stop the list of disjuncts. I’m sympathetic to Weatherson’s complaints; indeed, I think they apply not just to any disjunctive conceptual analysis of evidence, but also to any disjunctive answer to EGQ.

    (Of course, there’s probably a level at which the ultimate grounds for E-facts are highly disjunctive. If all the fundamental facts concern physical fields, forces, and particles, this will almost certainly be the case. But we need not deny this to resist our hypothetical epistemologist’s suggestion; we need only insist that there’s another level at which there’s a non-disjunctive answer to EGQ).

  11. In other words, whenever it’s a fact that S has e as evidence, this fact is partially or wholly grounded in facts of the form: [S is (propositionally) justified in believing p].

  12. Williamson denies there are any such cases (Williamson 2000, p. chp.9); Maher (1996) and Hyman (2006) seem to agree. For dissent, see Goldman (2009), Littlejohn (2011), and Arnold (2013).

  13. See Williamson (2000, p. chp.9) and Kelly (2006) for development of the idea that one important role of evidence is to justify the formation of further beliefs. (Note that this is compatible with the view that often when we come to know (and hence justifiably believe) p, we don’t do so on the basis of any evidence that’s independent of our knowledge that p. On this point, see e.g. Williamson 2005, 2009).

  14. Schroeder (2011) endorses a “low bar” picture of evidence, according to which S can have p as evidence even if p isn’t propositionally justified for S. In such cases p will be defeated evidence; on Schroeder’s view, if p is part of S’s defeated evidence, it won’t justify S in believing p’s consequences. Thus Schroeder would endorse Justificatory Role* rather than Justificatory Role.

  15. What’s more, if p could be part of S’s evidence even though S isn’t justified in believing p, we’d expect that there would be lots of cases where we could truly say, “Of course, Fred’s evidence entails that q is true; nonetheless, Fred has no good reason to believe q.” But such sentences sound (to my ears at least) like contradictions; at the very least I think we’d raise an eyebrow at any such utterance.

  16. For a defense of the view that evidence is knowledge, see Williamson (2000, p. chp.9). (As far as I can tell, Williamson never takes a stand on whether facts about evidence are grounded in facts about knowledge).

  17. Since Further Epistemic Grounding is noncommittal on exactly what R is, Further Epistemic Grounding is best viewed as a family of answers to the Propositional Evidential Grounding Question. In what follows, I argue for the disjunction of the instances of Further Epistemic Grounding, rather than for any particular instance.

  18. Of course, I haven’t shown it’s the only answer to the Propositional Evidential Grounding Question that satisfies our three desiderata. In Sect. 7, I’ll consider various rival answers that may appeal to evidentialists and argue that none of them constitutes a promising alternative.

  19. This is, of course, compatible with the view that the concept knowledge is unanalyzable.

  20. Consider: I know that Kerala is in India. Intuitively, this fact is partially explained by the fact that I’m justified in believing Kerala is in India. What’s more, the explanation in question seems to be metaphysical (rather than, say, causal)—exactly the sort of explanation that grounding is supposed to provide.

  21. Certainly, reflection on individual cases makes the assumption that partial grounding is transitive look eminently plausible. The fact that Moore knows <Here’s a hand, and here is another> is partially grounded in the fact that he knows <Here’s a hand>; this fact, in turn, is partially grounded in the fact, [Here’s a hand]. And, intuitively, the fact Moore knows <Here’s a hand, and here is another> is partially grounded in the fact, [Here’s a hand]. In light of cases like this, the assumption that partial grounding is transitive has won widespread support among metaphysicians; see Schaffer (2009), Rosen (2010), Fine (2010, 2012), and Whitcomb (2012).

    However, it should be noted that there are some voices of dissent: recently, Schaffer has revised his view on this front, offering putative counterexamples to the transitivity of partial grounding (2012). A full assessment of Schaffer’s alleged counterexamples is outside the scope of this paper; suffice to say that even if Schaffer’s counterexamples are successful, it’s far from clear that this would necessitate abandoning Knowledge is Partially Grounded in Justification. (Even if there are some cases where fact F1 is partially grounded in fact F2, which is in turn partially grounded in fact F3, even though F1 isn’t partially grounded in F3, this wouldn’t, by itself, show that there are any cases where [S has p as evidence] is partially grounded in [S knows p], and [S knows p] is partially grounded in [S is justified in believing p], but [S knows p] isn’t partially grounded in [S is justified in believing p]).

  22. For defense of the thesis that all evidence is propositional, see Williamson (2000, pp. 194–200) and Neta (2008).

  23. See Williamson (2000, p. 195) for a similar observation.

  24. As I’m using the phrase, a relation is asymmetric iff it’s antisymmetric and irreflexive. So if fact F1 is partially grounded in fact F2, then it’s not the case that fact F2 is partially grounded in fact F1. For a defense of the claim that partial grounding is asymmetric, see Rosen (2010, p. 116).

  25. See, for instance, Conee and Feldman (2004, p. 56).

  26. The “non-factive” qualifier is inserted to keep views according to which (i) knowledge is a mental state, (ii) E-facts are grounded in facts about a subject’s knowledge, from counting as forms of Grounding Mentalism. (Since any such view would be a form of Further Epistemic Grounding, it clearly wouldn’t be a rival answer to EGQ).

  27. This isn’t entirely obvious. After all, a philosopher might claim that justifiably believing is a mental state. Or she might claim that whenever S has p as evidence, this fact is grounded in the fact that S is in some particular non-factive mental state M whose content is p—where M is a state that involves having justification for believing its content. However, I’ll set these possibilities aside.

  28. See Feldman (1988) for the view S has p as evidence at t iff S is thinking of p at t.

  29. Huemer (2007) argues that seemings play an indispensable justificatory role (though he doesn’t, as far as I can tell, commit himself to the view that seemings wholly ground E-facts).

  30. For a classic account of evidence in terms of perceptual experiences and apparent memories, see Lewis (1996).

  31. One could also try to combine these proposals in various ways. For instance, Schroeder (2011) proposes that S has p as evidence iff S has a “presentational attitude” towards p, where a “presentational attitude” is any attitude that presents its content as true. Schroeder’s discussion makes it clear that he takes both belief and perceptual experience to be presentational attitudes.

  32. See Fine (2012) and Trogdon (2013b) for extended defenses of Grounding Necessitarianism.

  33. For the distinction, see Pollock (1986).

  34. It’s worth noting that Conee and Feldman (2005) explicitly endorse an account of defeat in terms of evidence.

  35. As an anonymous referee pointed out, grounding evidentialists might deny one of my desiderata on an answer to EGQ in order to save their view. While I agree that this is an option, it doesn’t seem particularly attractive. Each of the desiderata strikes me as well-motivated (see the discussion in Sect. 4); ceteris paribus, a view that satisfies all of them will be preferable to a view that doesn’t. If this is right, then Further Epistemic Grounding has a leg up on grounding evidentialism.

  36. For discussion of epistemic competences, see e.g. Sosa (2007, 2011), Mantel (2013), Sylvan and Sosa (2014).

  37. See Field (2009) for an account along these lines.

  38. Sylvan and Sosa (2014) might be sympathetic to an account along these lines. According to their view, facts about justification are grounded in facts about epistemic reasons, which are grounded in facts about epistemic competences. However, Sylvan and Sosa are officially noncommittal on whether epistemic reasons should be identified with E-facts.

  39. E.g. Conee and Feldman’s (E)—see Sect. 2.

  40. See e.g. Fantl and McGrath (2002).

  41. I use “C” as a placeholder for Comesaña’s reliabilist account of what it is for S’s belief that p to “fit” evidence e. See Comesaña 2010, p. 581–584 for details.


  • Arnold, A. (2013). Some evidence is false. Australasian Journal of Philosophy, 91(1), 165–172.

    Article  Google Scholar 

  • Clark, M. J., & Liggins, D. (2012). Recent work on grounding. Analysis, 72(4), 812–823.

    Article  Google Scholar 

  • Comesaña, J. (2010). Evidentialist reliabilism. Noûs, 44(4), 571–600.

    Article  Google Scholar 

  • Conee, E., & Feldman, R. (2004). Evidentialism: Essays in epistemology. Oxford: Oxford University Press.

    Book  Google Scholar 

  • Conee, E., & Feldman, R. (2005). Some virtues of evidentialism. Veritas, 50, 95–108.

    Google Scholar 

  • Conee, E., & Feldman, R. (2008). Evidence. In Smith (Ed.), Epistemology: New essays. Oxford: Oxford University Press.

    Google Scholar 

  • Conee, E., & Feldman, R. (2011). Response to goldman. In Dougherty (Ed.), Evidentialism and its discontents. Oxford: Oxford University Press.

    Google Scholar 

  • deRosset, L. (2013). Grounding explanations. Philosophers’ Imprint, 13(7), 1–26.

  • Fantl, J., & McGrath, M. (2002). Evidence, pragmatics, and justification. The Philosophical Review, 111(1), 67–94.

    Article  Google Scholar 

  • Feldman, R. (1988). Having evidence. In Austin (Ed.), Essays presented to Edmund Gettier (pp. 88–104). Dordrecht: Kluwer Academic Publishers.

    Google Scholar 

  • Feldman, R. (2003). Epistemology. Upper Saddle River, NJ: Prentice Hall.

    Google Scholar 

  • Field, H. (2009). Epistemology Without Metaphysics. Philosophical Studies, 143(2), 249–290.

    Article  Google Scholar 

  • Fine, K. (2001). The question of realism. Philosophers’ Imprint, 1(1), 1–30.

    Google Scholar 

  • Fine, K. (2010). Some puzzles of ground. Notre Dame Journal of Formal Logic., 51(1), 97–118.

    Article  Google Scholar 

  • Fine, K. (2012). Guide to Ground. In Correia & Schneider (Eds.), Metaphysical grounding. Cambridge: Cambridge University Press.

    Google Scholar 

  • Ganson, D. (2008). Evidentialism and pragmatic constraints on outright belief. Philosophical Studies, 139(3), 441–458.

  • Goldman, A. (2009). Williamson on knowledge and evidence. In Pritchard & L. Greenough (Eds.), Williamson on knowledge. Oxford: Oxford University Press.

    Google Scholar 

  • Goldman, A. (2011). Toward a synthesis of reliabilism and evidentialism. In Dougherty (Ed.), Evidentialism and its discontents. Oxford: Oxford University Press.

    Google Scholar 

  • Huemer, M. (2007). Compassionate phenomenal conservatism. Philosophy and Phenomenological Research, 74, 30–55.

    Article  Google Scholar 

  • Hyman, J. (2006). Knowledge and evidence. Mind, 115(460), 891–916.

    Article  Google Scholar 

  • Kelly, T. (2006). Evidence. In: Zalta (Ed.) Stanford encyclopedia of philosophy.

  • Kim, J. (1988). What is ‘naturalized epistemology’? Philosophical Perspectives, 2, 381–405.

    Article  Google Scholar 

  • Lewis, D. (1996). Elusive knowledge. Australasian Journal of Philosophy, 74(4), 549–567.

    Article  Google Scholar 

  • Littlejohn, C. (2011). Evidence and knowledge. Erkenntnis, 74(2), 241–262.

    Article  Google Scholar 

  • Maher, P. (1996). Subjective and objective confirmation. Philosophy of Science, 63(2), 149–174.

    Article  Google Scholar 

  • Mantel, S. (2013). Acting for reasons, apt action, and knowledge. Synthese, 190(17), 3865–3888.

    Article  Google Scholar 

  • Neta, R. (2008). What evidence do you have? British Journal for the Philosophy of Science, 59(1), 89–119.

    Article  Google Scholar 

  • Pollock, J. (1986). Contemporary theories of knowledge. Savage, MD: Rowman and Littlefield.

    Google Scholar 

  • Raven, M. (2012). In defence of ground. Australasian Journal of Philosophy, 90(4), 687–701.

    Article  Google Scholar 

  • Rosen, G. (2010). Metaphysical dependence: Grounding and reduction. In Hale & L. Hoffman (Eds.), Modality: Metaphysics, logic, and epistemology. Oxford: Oxford University Press.

    Google Scholar 

  • Schaffer, J. (2009). On what grounds what. In M. Chalmers & Wasserman (Eds.), Metametaphysics. Oxford: Oxford University Press.

    Google Scholar 

  • Schaffer, J. (2012). Grounding, transitivity, and contrastivity. In Correia & Schneider (Eds.), Grounding and explanation. New York: Cambridge University Press.

    Google Scholar 

  • Schroeder, M. (2011). What does it take to ‘have’ a reason? In Reisner & Steglich-Peterson (Eds.), Reasons for belief (pp. 201–222). New York: Cambridge University Press.

    Chapter  Google Scholar 

  • Sosa, E. (2007). Apt belief and reflective knowledge (Vol. 1). Oxford: Oxford University Press.

    Google Scholar 

  • Sosa, E. (2011). Knowing full well. Princeton: Princeton University Press.

    Book  Google Scholar 

  • Sylvan, K. & Sosa, E. (2014). The place of reasons in epistemology. In Star (Ed.), The oxford handbook of reasons and normativity. Oxford: Oxford University Press. (forthcoming).

  • Trogdon, K. (2013a). An introduction to grounding. In Hoeltje, S, & Steinberg (eds.), Varieties of dependence (pp. 97–122). Basic Philosophical Concepts, Munich: Philosophia Verlag.

  • Trogdon, K. (2013b). Grounding, necessary or contingent? Pacific Philosophical Quarterly, 94(4), 465–485.

    Article  Google Scholar 

  • Weatherson, Brian. (2005). Can we do without pragmatic encroachment? Philosophical Perspectives, 19, 417–443.

    Article  Google Scholar 

  • Weatherson, B. (2009). Evidence ≠ knowledge.

  • Whitcomb, D. (2012). Grounding and omniscience. In Kvanvig (Ed.), Oxford studies in philosophy of religion (Vol. 4). Oxford: Oxford University Press.

    Google Scholar 

  • Williamson, T. (2000). Knowledge and its limits. Oxford: Oxford University Press.

    Google Scholar 

  • Williamson, T. (2005). Replies to commentators. Philosophy and Phenomenological Research, 70(2), 468–491.

    Article  Google Scholar 

  • Williamson, T. (2009). Replies to critics. In Greenough & Pritchard (Eds.), Williamson on knowledge. Oxford: Oxford University Press.

    Google Scholar 

Download references


Special thanks to Alvin Goldman, for extensive feedback and consistent encouragement. Thanks also to Marco Dees, Andy Egan, Simon Goldstein, Lisa Mirrachi, Carlotta Pavese, Jonathan Schaffer, Susanna Schellenberg, Alex Skiles, Ernest Sosa, Kurt Sylvan, Tobias Wilsch, an anonymous referee, and participants in the Third International Summer School in Cognitive Science and Semantics for helpful comments and conversations.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Bob Beddor.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Beddor, B. Evidentialism, circularity, and grounding. Philos Stud 172, 1847–1868 (2015).

Download citation

  • Published:

  • Issue Date:

  • DOI:


  • Evidentialism
  • Grounding
  • Circularity
  • Justification
  • Evidence
  • Reliabilism