Epistemic dependence and cognitive ability


In a series of papers, Jesper Kallestrup and Duncan Pritchard argue that the thesis that knowledge is a cognitive success because of cognitive ability (robust virtue epistemology) is incompatible with the idea that whether or not an agent’s true belief amounts to knowledge can significantly depend upon factors beyond her cognitive agency (epistemic dependence). In particular, certain purely modal facts seem to preclude knowledge, while the contribution of other agents’ cognitive abilities seems to enable it. Kallestrup and Pritchard’s arguments are targeted against views that hold that all it takes to manifest one’s cognitive agency is to properly exercise one’s belief-forming abilities. I offer an account of the notion of cognitive ability according to which our epistemic resources are not exhausted by abilities to produce true beliefs as outputs, but also include dispositions to stop belief-formation when actual or modal circumstances are not suitable for it (precautionary cognitive abilities). Knowledge, I argue, can be accordingly conceived as a cognitive success that is also due to the latter. The resulting version of robust virtue epistemology helps explain how purely modal facts as well as other agents’ cognitive abilities may have a bearing on the manifestation of one’s cognitive agency, which shows in turn that robust virtue epistemology and epistemic dependence are not incompatible after all.

This is a preview of subscription content, log in to check access.


  1. 1.

    For two influential versions of robust virtue epistemology, see Greco (2010, 2012) and Sosa (2007, 2015). Greco opts for a causal explanatory reading of the ‘because of’ relation and accordingly understands knowledge as a cognitive success that is explained by one’s cognitive abilities. Sosa prefers a metaphysical reading according to which knowledge is a cognitive success that manifests one’s cognitive abilities. The version of robust virtue epistemology that I will give in Sect. 3 endorses the metaphysical reading.

  2. 2.

    The idea of epistemic dependence is discussed at length in Pritchard (2015). In their 2014 paper, Kallestrup and Pritchard introduce their key case of knowledge-precluding epistemic dependence: the epistemic twin earth case (see below) and try to block possible replies by robust virtue epistemologists. Their 2012 paper is mainly concerned with knowledge-enabling epistemic dependence in testimony cases. Their 2013a and 2013b papers represent good summaries of their objections to robust virtue epistemology. Finally, in Kallestrup and Pritchard (2016) they focus on versions of robust virtue epistemology that account for knowledge in terms of the notion of manifestation of cognitive ability. All objections, arguments and cases by Kallestrup and Pritchard that I will discuss in this paper are spread over these works. To avoid constant repetition of these references, I omit them in what follows (for specific points, however, I do refer the reader to the papers where they are discussed in-depth).

  3. 3.

    See especially Kallestrup and Pritchard (2013a) for this point.

  4. 4.

    See Goldberg (2011). For relevant discussion of third-party epistemic dependence, see especially Kallestrup and Pritchard (2012).

  5. 5.

    See especially Kallestrup and Pritchard (2014) for this case.

  6. 6.

    See Putnam (1973) for the original example.

  7. 7.

    In Broncano-Berrocal (2017), I discuss this way to individuate abilities at length.

  8. 8.

    Of course, this disposition must be adequately integrated within one’s cognitive system. See Breyer and Greco (2008) for relevant discussion on the notion of cognitive integration.

  9. 9.

    This way to understand reliability is a generalization from the way the term ’reliability’ is typically used in epistemology. In quantitative research (e.g., in experimental psychology), reliability is understood more simply as consistency or repeatability of a certain type of outcome (e.g., in an experiment). It is the term ‘validity’ that is used instead to refer to the tendency to produce outcomes of a certain type such that they meet the requirements of a standard of assessment (e.g., being successful, accurate, correct, and so on).

  10. 10.

    What counts as ’sufficiently high’ depends on the domain of the relevant ability.

  11. 11.

    Another important remark about the specific reliability of precautionary abilities: if a precautionary ability is broad or narrow so will be its reliability. A broad disposition to stop visual belief formation when the light conditions are bad is broad enough to be reliable in any situation in which the light conditions are bad. A narrow disposition to stop belief formation on one occasion based on trustworthy testimonial information that some sort of deception is going on is reliable only relative to that situation. Suppose that one erroneously believes such information to be true of a large number of situations (e.g., think of believers in conspiracy theories). Stopping belief formation in those situations is reckless, because they are beyond the range of reliability of the locally acquired disposition to stop belief formation in the specific circumstances the information is about.

  12. 12.

    This case is analogous to the much-discussed fake barn case (Goldman 1979).

  13. 13.

    Belief-forming abilities are first-order because they are dispositions to form true beliefs. Precautionary cognitive abilities are second-order because they are dispositions to prevent the manifestation of other dispositions. However, it is important to keep in mind that the fact that the former are first-order and the latter second-order is not what distinguishes them. What distinguishes them is the fact that they play distinctfunctional roles.

  14. 14.

    For extensive discussion on the notion of monitoring see Nelson and Narens (1990).

  15. 15.

    A note of clarification is in order. My version of robust virtue epistemology doesn’t follow the simple reliabilist template according to which S knows p if and only if S’s true belief is produced by a reliable belief-forming process. In this way, my view is not that knowledge is belief that is produced by a two-level belief-forming system whose overall reliability is sufficiently high. Instead, it follows the standard template of virtue accounts according to which an agent S knows p if and only if S believes p truly because of her reliable cognitive abilities. The main difference with standard virtue-theoretic views is that the agent must get things right because of reliably exercising not only her first-order belief-forming abilities but also her precautionary cognitive abilities: in particular, the correctness of the agent’s belief must manifest the fact that her relevant belief-forming abilities reliably produce a true belief in appropriate circumstances and the fact that her precautionary cognitive abilities reliably avoid stopping belief formation precisely because the circumstances are adequate. Unlike simple reliabilist views, the ‘because of’ clause (which applies to both types of abilities) serves, as in standard virtue accounts, as a measure against Gettierization. In particular, Gettierized beliefs are not only true but typically reliably formed; however, their correctness does not manifest (i.e., it is not because of) the exercise of ability (namely, of belief-forming and precautionary ability). In Broncano-Berrocal (2017), I develop this account in more detail. See Broncano-Berrocal (forthcoming) for how this view fares better than rivals. A more sophisticated but significantly different version of this view would define knowledge as true belief that manifests overall reliability, i.e., as belief that is true because of the overall reliable belief-forming process that results from the interaction of the agent’s first-order belief-forming abilities and second-order precautionary abilities. As Michaelian insightfully explains, the reliability of an overall reliable process allows for different degrees of reliability of the former and the latter in such a way that knowledgeable beliefs can be attained, e.g., with a reckless precautionary policy if the relevant first-order abilities enjoy a greater degree of reliability than usual. For present purposes, I will stick with the first (and arguably simpler) version of the view, keeping in mind that it could be enriched with Michaelian’s careful examination of how the reliability of first-order and second-order cognitive processes interact in two-level belief-forming systems.

  16. 16.

    See Mackie (1974) for a similar distinction between causes and background conditions.

  17. 17.

    See Heil (2005), Martin (2008) and Mumford and Anjum (2011) for some representative defenses of this idea.

  18. 18.

    See Michaelian (2010) for relevant discussion of empirical findings concerning the claims that our capacities for monitoring deception are not reliable, that most people tend to be honest, and that we are truth-biased, as well as of the implications that these have for the epistemology of testimony.

  19. 19.

    This is particularly true of the kind of cooperative environments described by Goldberg (2011) in which one’s community or a third party epistemic agent monitors and polices testimonial exchanges in such a way that one is mostly exposed to reliable speakers (see Sect. 1).


  1. Breyer, D., & Greco, J. (2008). Cognitive integration and the ownership of belief: Response to Bernecker. Philosophy and Phenomenological Research, 76, 173–184.

    Article  Google Scholar 

  2. Broncano-Berrocal, F. (2017). A robust enough virtue epistemology. Synthese, 194, 2147–2174.

    Article  Google Scholar 

  3. Broncano-Berrocal, F. (forthcoming). Purifying impure virtue epistemology. Philosophical Studies. doi:10.1007/s11098-017-0873-x.

  4. Goldberg, S. (2011). The division of epistemic labour. Episteme, 8, 112–25.

    Article  Google Scholar 

  5. Goldman, A. I. (1979). What is justified belief? In E. Sosa & J. Kim (Eds.), Epistemology. An anthology (pp. 340–353). Malden: Blackwell.

    Google Scholar 

  6. Greco, J. (2010). Achieving knowledge. Cambridge: Cambridge University Press.

    Google Scholar 

  7. Greco, J. (2012). A (Different) virtue epistemology. Philosophy and Phenomenological Research, 85, 1–26.

    Article  Google Scholar 

  8. Heil, J. (2005). Dispositions. Synthese, 144, 343–356.

    Article  Google Scholar 

  9. Kallestrup, J., & Pritchard, D. H. (2012). Robust virtue epistemology and epistemic anti-individualism. Pacific Philosophical Quarterly, 93, 84–103.

    Article  Google Scholar 

  10. Kallestrup , J., & Pritchard , D. H. (2013a). Robust virtue epistemology and epistemic dependence. In T. Henning & D. Schweikard (Eds.), Knowledge, virtue and action: Putting epistemic virtues to work, ch. 11, Routledge.

  11. Kallestrup , J., & Pritchard , D. H. (2013b). The power, and limitations, of virtue epistemology. In J. Greco & R. Groff (Eds.), Powers and capacities in philosophy: The new aristotelianism, ch. 13, Routledge.

  12. Kallestrup, J., & Pritchard, D. H. (2014). Virtue epistemology and epistemic twin earth. European Journal of Philosophy, 22, 335–357.

    Article  Google Scholar 

  13. Kallestrup , J., & Pritchard , D. H. (2016). Dispositional robust virtue epistemology versus anti-luck virtue epistemology. In M. Fernández (Ed.), Performance epistemology: Foundations and applications, ch. 2. Oxford: Oxford University Press.

  14. Mackie, J. L. (1974). The cement of the universe. Oxford: Clarendon Press.

    Google Scholar 

  15. Martin, C. B. (2008). The mind in nature. Oxford: Oxford University Press.

    Google Scholar 

  16. Michaelian, K. (2010). In defence of gullibility: The epistemology of testimony and the psychology of deception detection. Synthese, 176, 399–427.

    Article  Google Scholar 

  17. Michaelian, K. (2012). Metacognition and endorsement. Mind & Language, 27, 284–307.

    Article  Google Scholar 

  18. Michaelian, K. (2016). Mental time travel: Episodic memory and our knowledge of the personal past. Cambridge: MIT Press.

    Google Scholar 

  19. Mumford, S., & Anjum, R. (2011). Getting causes from powers. Oxford: Oxford University Press.

    Google Scholar 

  20. Nelson, T. O., & Narens, L. (1990). Metamemory: A theoretical framework and new findings. In G. Bower (Ed.), The psychology of learning and motivation: Advances in research and theory (pp. 125–173). San Diego: Academic Press.

    Google Scholar 

  21. Pritchard, D. H. (2015). Epistemic dependence. Philosophical Perspectives, 29, 305–324.

    Article  Google Scholar 

  22. Putnam, H. (1973). Meaning and reference. Journal of Philosophy, 70, 699–711.

    Article  Google Scholar 

  23. Sosa, E. (2007). A virtue epistemology: apt belief and reflective knowledge. Oxford: Oxford University Press.

    Google Scholar 

  24. Sosa, E. (2015). Judgment and agency. Oxford: Oxford University Press.

    Google Scholar 

Download references


I am grateful to an anonymous reviewer for Synthese for helpful suggestions. This work was funded by a Marie Skłodowska-Curie Individual Fellowship (Grant agreement: No. 656082).

Author information



Corresponding author

Correspondence to Fernando Broncano-Berrocal.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Broncano-Berrocal, F. Epistemic dependence and cognitive ability. Synthese 197, 2895–2912 (2020). https://doi.org/10.1007/s11229-017-1559-9

Download citation


  • Epistemic dependence
  • Cognitive ability
  • Precautionary ability
  • Virtue epistemology
  • Epistemic twin earth
  • Testimonial knowledge
  • Metacognition