Springer Nature is making Coronavirus research free. View research | View latest news | Sign up for updates

Mistakes as revealing and as manifestations of competence

  • 102 Accesses


The final chapter of Elgin’s (True enough, The MIT Press, Cambridge, 2017) defends the claim that some mistakes mark significant epistemic achievements. Here, I extend Elgin’s analysis of the informativeness of mistakes for epistemic policing. I also examine the type of theory of competence that Elgin’s view requires, and suggest some directions in which this can be taken.

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


  1. 1.

    A previous version of the material appears in Elgin (2012).

  2. 2.

    This rough characterization coincides in the main with Whitcomb et al. (2015), who say that intellectual humility is ‘having the right stance towards one’s intellectual limitations’ (p. 516), where ‘the right stance’ is ‘to be appropriately attentive to [those limitations] and to own them’. Church (2016) argues that epistemic humility should also be sensitive to one’s strengths, and proposes that intellectual humility is ‘the virtue of accurately tracking what one could non-culpably take to be the positive epistemic status of one’s own beliefs’ (p. 427). Similar points are made by Kallestrup and Pritchard (2016), and by Tanesini (2018), who argues that epistemic humility is a complex attitude involving modesty (about one’s epistemic successes) and self-acceptance (of one’s epistemic failures).

  3. 3.

    The disjunction prescriptions/dispositions signals two variant reconstructions of the fallibilist stance. In one, fallibilism is characterized by the belief in certain prescriptions on what to do if one falls into error; in this case fallibilism can be identified with a doctrine or higher-order belief about one’s understanding. In the other, fallibilism is rather characterized by the possession of some dispositions to act in certain ways relevant to the management of error. Elgin herself says that her argument shows that fallibilism ‘is not a higher-order stance toward our understanding of a topic’, but that it ‘is, or should be, woven into the fabric of understanding’ (p. 309), which suggests she prefers the dispositional variant. However, for every dispositional account there will be a doctrine that asserts that understanding should be as if it was determined by the disposition, and vice versa. What favours the dispositional view is that it seems to provide a natural solution to the issue of motivating the agent to act as the norms prescribe: the disposition would presumably be expected to intrinsically lead the agent to act in the required way given certain triggers. A purely dispositional account, however, seems problematic: shouldn’t one be further disposed to be so disposed (and so on), falling into a regress? How can one even adopt a stance, if so? I would suggest formulating fallibilism in such a way that it entailed the acceptance of the doctrine that one should ameliorate one’s understanding’s management of error, including the disposition to ameliorate it. Then, we could appeal to more general rational competences for the uptake and application of the relevant norms (this does not necessarily break the potential regress of dispositions, but the fact that the move doesn’t require the immediate supposition of ad hoc capacities could make it more palatable). It should be noted that the strategy of appealing to more general capacities can be pursued in such a way that no specific dispositions are involved in the description of a fallibilist stance: instead, general dispositions could drive acting according to the fallibilist prescriptions.

  4. 4.

    In the ex post case, the distinction between objective and epistemic policies is difficult to sustain, because while there could be norms that prescribe what to do once one has fallen into error, these cannot be applied unless one also recognizes or comes to believe that one has indeed fallen into error. So, at best, objective ex post policies could coincide with epistemic ex ante ones.

  5. 5.

    Pritchard (2017) points out that risk assessment is forwards looking, whereas luck assessment seems to be backwards-looking. Elgin observes that being merely able to assess some unlucky event as such is less good than also being able to identify the factors of risk involved, and is the sign of a worse understanding (p. 298). This emphasizes the importance of adopting certain ex ante policies (although in the example Elgin discusses the relevant policies are not concerned with error, but failures of other kinds, like the manifestation of side effects).

  6. 6.

    In mathematics, it is common to try to prove theorems indifferent ways. Polya (1957) emphasizes this a crucial heuristic for discovery and understanding. In these cases the mathematician is often sure that his initial proof is correct. Nevertheless, they may still be tracking some notion of epistemic failure, as they have come to believe that their commitments are somehow incomplete or otherwise inadequate (in some cases by aesthetic criteria). On this point, see also Elgin (2002).

  7. 7.

    Cf. Elgin (2017, p. 306): ‘... sometimes, we get things wrong in epistemically fruitful ways. Once discovered, such errors provide not just incentives but resources for serious, focused, effective inquiry. By revealing not only that but also where we have got things wrong, they point us in the direction of advancing our understanding. We are lucky we are disposed to make such mistakes.’ If I’m right, the appeal to luck here should be treated at least in part as rhetorical.

  8. 8.

    Write ‘Mst’ for ‘s makes a mistake about t’, ‘Bst’ for ‘s holds beliefs about t’, ‘NsC’ for ‘s has a network of commitments which satisfy condition C’, and ‘As’ for ‘s has an epistemic achievement’. Then, (1) is \(\lnot \Diamond (Mst\wedge \lnot Bst)\), (2) is \(\lnot \Diamond (Bst\wedge \lnot NsC)\), and (3) is \(NsC\rightarrow As\). From (1), it follows that \(Mst\rightarrow Bst\), and from (2) it follows that \(Bst\rightarrow Nsc\). Then, by iterated application of transitivity of \(\rightarrow \), \(Mst\rightarrow Asc\).

  9. 9.

    Whether this is acceptable will depend on how we view the structure of epistemic concepts. Other frameworks will have different concepts (like knowledge) as central. However, this does not necessarily undermine the heuristic. If the notions of belief and acceptance exist at all in a framework, the heuristic could still be robust if their place was at least relatively central. A different option would be to distinguish between centrality and ‘fundamentality’: then, we could claim that, while belief or acceptance are strictly central, other notions are more fundamental. Yet another option would be to limit the scope of the heuristic.

  10. 10.

    Sosa’s (2011) conception of levels of epistemic evaluation could offer a complementary explanation. If mistakes are conceived as a form of performance, the evaluative language associated to performances can be transposed to their case, just like Sosa does in the case of belief. For example, we can consider whether mistaken beliefs or acceptances are inaccurate (as opposed to accurate), incompetent (as opposed to adroit) or inept (as opposed to apt); note that in this case, the severity of the mistakes decreases as we go from inaccuracy to incompetence, and then to ineptitude.

  11. 11.

    This is closely connected to the Quine-Duhem problem of underdetermination. Unrestricted closure endangers the possibility of determining where errors lie. Cf. Duhem (1954, p. 187): “[...] what the experiment declares stained with error is the whole group of propositions accepted by Newton, and after him by Laplace and Biot, that is, the whole theory from which we deduce the relation between the index of refraction and the velocity of light in various media. But in condemning this system as a whole by declaring it stained with error, the experiment does not tell us where the error lies.”

  12. 12.

    For our purposes, \(\left[ _M\phi \right] \psi \) could be defined in terms of the \(\left[ \phi !\right] \psi \) operator of public announcement logic (PAL), as the special case ‘after a public announcement that \(\phi \) is mistaken, it should be that...’. In a more general setting, we need to distinguish between this and the case where a mistake has been detected but no announcement has been made. An overview of PAL and related formalisms can be found in van Ditmarsch et al. (2007).

  13. 13.

    An alternative way to develop the semantics for \(\left[ _M\phi \right] U_\varGamma (\psi )\) could be to use some form of exact truthmaker semantics, in the way that Fine (2017) does in its treatment of imperatives: the update action \(U_\varGamma (\psi )\) should be in exact compliance with the announcement that ‘\(\phi \) has been found as a mistake’. Note that the compositional story is more complex than in the imperative case: for example, an update in compliance with \([_M\phi \wedge \psi ]\) is not equivalent to an update in compliance with both \([_M\phi ]\) and \([_M\psi ]\): that a conjunction is mistaken does not entail that the conjuncts are independently mistaken.

  14. 14.

    The prescription would be impossible for us to apply fully, so it can be tempting to judge it is poor or deficient in some sense as an action-guiding principle (if it is one at all). Even so, it still seems possible to try to follow it. A more difficult situation would arise if even that was prevented; in that case the prescription would not provide any sort of action-guidance.

  15. 15.

    Cf. also Hawke (2017).

  16. 16.

    The theme is explored further in Friedman (2017).

  17. 17.

    On the notion of effort, see Von Kriegstein (2016).

  18. 18.

    This liberalization of the conditions that could constitute achievement does not prevent having a consistent account of the magnitude of epistemic achievements that fits adequately with the ordinary intuition that achievements are in some sense uncommon: one could say that while getting any achievement is easy, great achievements are hard to come by. A view like this has to assessed in the balance of conservativeness/fruitfulness; in this case, it is unclear that the revisionary view has more expressive power than the ordinary one (although it should not have less expressive power).

  19. 19.

    Rohrbaugh (2015) develops a notion of ‘inner’ achievement which can be useful here: such achievements apply to ‘inner’ projects like health, or friendship, where ‘our activities do not aim at bringing about some further, as yet unachieved goal, but instead constitute, realize, or prolong the valuable state itself.’ Engaging in such projects is their own achievement. Rohrbaugh suggests that this model of achievement also applies to the epistemic domain, where it pertains to the adequacy of our reasons to believe (p. 1200); instead of ‘believe’, Elgin could say ‘accept’(see (2017), pp. 97–99).

  20. 20.

    The traditional view is that dispositions are individuated by a pair of sets of trigger conditions and manifestations. Some, like Vetter (2014), disagree (she argues that the set of manifestations is sufficient). Here, I only need to assume that manifestation conditions are necessary, and remain neutral about sufficiency.

  21. 21.

    Someone could bite the bullet and accept that the effective completion of a procedure is enough for a minimal sense of success, but I think this trivializes the notion of success.

  22. 22.

    I thank an anonymous reviewer for pressing me on the reach of the argument that follows.

  23. 23.

    I will remain neutral on whether we should make use of the notion of competence or on whether competence-based accounts are the best explanations for epistemic phenomena. My point is merely that there are constraints on what notion of competence can be used.

  24. 24.

    A different worry is that Turri’s idea was meant to explain successes which nonetheless failed to constitute a higher form of achievement, so it might not be apt for the case of failures.

  25. 25.

    Cf. Elgin’s (2017, pp. 59–61) criticism of Kvanvig’s (2003) idea that in some cases (like with incompletely educated children) we attribute understanding only in an honorific sense. If the children’s understanding is epistemically significant (qua understanding), then their competences should also be epistemically significant.

  26. 26.

    Miracchi (2014, pp. 31–33) criticizes an argument by Burge (2011) that every case of success could have been a failure. If we allow the possibility of neutral exercises, a similar argument can be rejected for the claim that for every case of failure, there could have been a success.

  27. 27.

    It will not necessarily or even inescapably propagate as a mistake. This will depend on what kind of cognitive safeguards are in place; we have seen how we can have epistemic policies to minimize risks, and these could minimize the consequences of faulty procedures.

  28. 28.

    Someone could object that the kind of systematicity requirement that COP endorses (that the exercise of a competence fits in a larger process) should be itself understood as a kind of achievement (for example, in analogy to Elgin’s view on understanding as a form of achievement, which also relies on structural properties of the cognitive states). If we accepted that, COP would collapse into a version of PSRT. However, structure is ubiquitous, so there is no reason to think that it is universally or inescapably valuable as a success. We need stronger reasons to think that in this case the required structure is an achievement. Consider, however, that some intuitively epistemically/cognitively significant processes (significant because we would have to appeal to them to explain our epistemic/cognitive states) could be automatic, and consequently not personally attributable as an achievement. Elgin’s argument for the idea that understanding is an achievement seems to require that acceptance is in some sense under personal control (2017, pp. 95–99). But this is not so plausible for the present case. On the other hand, the overall view could be adjusted so that the claim that understanding is an achievement was replaced by the claim that understanding is a manifestation of competence, without the assumption that the relevant manifestations of competence are achievements.

  29. 29.

    In some cases, like when the requirements for trying are very involved, coming to be able to try should count as an achievement, but in general coming to try is a neutral condition. To count trying as a form of success is to trivialize success (cf. also note 21 above). Offering a defense of this idea is outside of the scope of this paper. If it is not granted, the conclusion we arrive at is simply a statement of the lower bound of success that can be epistemically meaningful, as suggested above. That is success enough.


  1. Berto, F., & Hawke, P. (2018). Knowability relative to information. Mind,.

  2. Burge, T. (2011). Disjunctivism again. Philosophical Explorations, 14(1), 43–80.

  3. Church, I. M. (2016). The doxastic account of intellectual humility. Logos & Episteme VI, I(4), 413–433.

  4. Cohen, L. J. (1992). An essay on belief and acceptance. Oxford: Oxford University Press.

  5. Duhem, P. (1954). The aim and structure of physical theory. Princeton: Princeton University Press.

  6. Elgin, C. Z. (2002). Art in the advancement of understanding. American Philosophical Quarterly, 39, 1–12.

  7. Elgin, C. Z. (2012). Ignorancia, error y el avance de la comprensión. In V. Campos, A. Estany, & M. Izquierdo (Eds.), Error y Conocimiento: La Gestación de la Ignorancia desde la Didáctia, la Éthca y la Filosofía. Grenada: Comares.

  8. Elgin, C. Z. (2017). True enough. Cambridge: The MIT Press.

  9. Fine, K. (2017). Truthmaker semantics. In B. Hale, C. Wright, & A. Miller (Eds.), A companion to the philosophy of language (pp. 556–577). Oxford: Blackwell.

  10. Friedman, J. (2017). Junk beliefs and interest driven epistemology. Philosophy and Phenomenological Research,.

  11. Harman, G. (1986). Change in view: Principles of reasoning. Cambridge, Massachusetts: The MIT Press.

  12. Hawke, P. (2017). The problem of epistemic relevance. Ph.D. thesis, Department of Philosophy, Stanford University.

  13. Kallestrup, J., & Pritchard, D. (2016). From epistemic anti-individualism to intellectual humility. Res Philosophica, 93(3), 533–552.

  14. Kvanvig, J. (2003). The value of knowledge and the pursuit of understanding. Cambridge: Cambridge University Press.

  15. Lewis, D. (1988). Relevant implication. Theoria, 54, 161–174.

  16. Mach, E. (1976). Knowledge and error: Sketches on the psychology of enquiry. Dordrecht: D. Reidel Publishing Company.

  17. Millar, A. (2008). Perceptual-recognitional abilities and perceptual knowledge. In A. Haddock & F. Macpherson (Eds.), Disjunctivism: Perception, action, knowledge. Oxford: Oxford University Press.

  18. Miracchi, L. (2014). Getting things done. Ph.D. thesis.

  19. Miracchi, L. (2015). Competence to know. Philosophical Studies, 172, 29–56.

  20. Polya, G. (1957). How to solve it: A new aspect of mathematical method (2nd ed.). Princeton: Princeton University Press.

  21. Pritchard, D. (2017). Anti-risk epistemology and negative epistemic dependence. Synthese,.

  22. Rescorla, M. (2009). Assertion and its constitutive norms. Philosophy and Phenomenological Research, 79(1), 98–130.

  23. Rohrbaugh, G. (2015). Inner achievement. Erkenntnis, 80, 1191–1204.

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

  25. Sosa, E. (2010). How competence matters in epistemology. Philosophical Perspectives, 24, 465–475.

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

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

  28. Tanesini, A. (2018). Intellectual humility as attitude. Philosophy and Phenomenological Research, 96(2), 399–420.

  29. Turri, J. (2011). Manifest failure: The Gettier problem solved. Philosopher’s Imprint, 11(8), 1–11.

  30. van Ditmarsch, H., van der Hoek, W., & Kooi, B. (2007). Dynamic epistemic logic. Berlin: Springer.

  31. Vetter, B. (2014). Dispositions without conditionals. Mind, 123(489), 129–156.

  32. Von Kriegstein, H. (2016). Effort and achievement. Utilitas, 29(1), 27–51.

  33. Whitcomb, D., Battaly, H., Baehr, J., & Howard-Snyder, D. (2015). Intellectual humility: Owning our limitations. Philosophy and Phenomenological Research, XCIV(3), 509–539.

  34. Yablo, S. (2014). Aboutness. Princeton: Princeton University Press.

Download references


This research was supported by CONICYT’s (Chile) Beca de Doctorado para Estudios en el Extranjero Nr. 72180096. I want to thank my colleagues at KU Leuven, and in particular, I would like to thank Lars Tump and Jan Heylen for their feedback. I would also like to thank two anonymous reviewers for their helpful suggestions.

Author information

Correspondence to Felipe Morales Carbonell.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Morales Carbonell, F. Mistakes as revealing and as manifestations of competence. Synthese (2019).

Download citation


  • Understanding
  • Mistakes
  • Informativeness of mistakes
  • Epistemic policies
  • Competence