Epistemic modals and credal disagreement


Considerations involving disagreement, as well as related considerations involving correction and retraction, have played an important role in recent debates about epistemic modals. For instance, it has been argued that contextualist views about epistemic modals have problems when it comes to explaining cases of disagreement. In response to these challenges, I explore the idea that the relevant cases of disagreement may involve credal disagreement. In a case of credal disagreement, the parties have different degrees of belief or credences. There does not have to be a difference in outright beliefs in order for the parties to disagree. I argue that the idea of credal disagreement allows us to make sense of otherwise problematic cases of disagreement involving epistemic modals. I also discuss how these ideas can be extended to cases of correction and retraction.

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


  1. 1.

    A standard approach is to treat the modals in question as quantifiers over a contextually restricted set of possible worlds. This allows us to implement the idea that the relevant sentences are true just in case the embedded proposition is compatible with a body of knowledge or evidence. In the case of epistemic modals, the relevant worlds are the worlds that are compatible with the relevant knowledge or evidence. For some important work on modality within the possible worlds framework, see Kratzer (1977, 1981, 1991b). See Portner (2009) for an introduction to this kind of framework and linguistic theories of modality in general.

  2. 2.

    This is not the only version of contextualism and it is probably not even the most plausible. It could be that what matters is the combined knowledge of all the conversational participants or perhaps even what the conversational participants could come to know through some further, but presumably limited, investigation. See e.g. Hacking (1967) and DeRose (1991) for relevant discussion. It is could also be that is not knowledge, but some other epistemic notion that is relevant. See e.g. Dougherty and Rysiew (2009), Dever (2011), and Hawthorne (2012) for relevant discussion. However, for the purpose of the present discussion, these complications do not matter and I will stick with a simple version of contextualism.

  3. 3.

    See e.g. Egan et al. (2005), Egan (2007), Stephenson (2007), and MacFarlane (2011) who have argued that we should adopt a relativist semantics for epistemic modals, as opposed to a contextualist semantics. However, considerations involving disagreement can also be used to motivate a non-truth conditional approach. See e.g. Price (1983). For various attempts to defend a more traditional contextualist approach, see e.g. Papafragou (2006), Hawthorne (2007), Wright (2007), Schaffer (2011), von Fintel and Gillies (2008, 2011), and Dowell (2011). For present purposes, it does not matter what we take the main alternative to contextualism to be.

  4. 4.

    It is beyond the scope of this paper to attempt to solve all the problems involving contextualist treatments of epistemic modals. For instance, Yalcin (2007, 2011) has identified interesting problems involving embedded epistemic modals that I will not attempt to address.

  5. 5.

    It is worth noting that Alex’ attempt to stand her ground raises difficult problems concerning the interaction between tense and modals. These problems remain even if we adopt a contextualist treatment.

  6. 6.

    See e.g. von Fintel and Gillies (2008, p. 92) and Schaffer (2011, p. 214) for a discussion of similar cases.

  7. 7.

    The idea that two parties can disagree in virtue of having conflicting credences is not uncontroversial. See e.g. Cappelen and Hawthorne (2011b) who seem reluctant to talk about disagreement when it comes to credences. They note that it is more natural to talk about disagreement as applying to beliefs rather than weighted credences, but beyond that they do not offer any further arguments. However, in ordinary speech it is rare to find speakers talking explicitly about their credences, so it is not clear that this is a major concern.

    It is also worth noting that I am not alone in thinking that there are genuine cases of credal disagreement. In the literature on so-called ‘peer disagreement’ in epistemology, it is common to talk about cases of disagreement involving conflicting credences. See e.g. Jehle and Fitelson (2009, p. 280) and Goldman (2010, p. 190) who are explicit about thinking about disagreement in his way.

  8. 8.

    I am assuming that having a relatively high credence, such as 0.7, is compatible with not having an outright belief in the relevant proposition. For the purpose of the following discussion, I am also assuming that credences are precise, though as far as I can see, nothing here turns on whether we allow credences to be imprecise. See e.g. Joyce (2010) for a recent defence of imprecise credences.

  9. 9.

    In a realistic scenario, it will be hard to make sure that the two parties have different credences in a proposition, without also having different outright beliefs regarding related propositions. This admittedly makes it more difficult to identify the cases in which the disagreement is a result of different credences as opposed to a different outright beliefs.

  10. 10.

    Even if we accept Williamson’s (2000) views about evidence, the connection between what an agent takes herself to know and her credences will not be completely straightforward. For instance, let us follow Hawthorne (2012) and suppose that someone throws an infinitely thin dart at random at dart board with continuum many points. In that case, it looks like the chance of the dart hitting the bulls eye is going to be zero. But then there is also a temptation to assign a credence of zero to that proposition. However, we do not want to say that we know that it will not hit the bulls eye and it is still fine to say ‘It might hit the bulls eye’.

  11. 11.

    This simple expressivist view might be too simple. As pointed out earlier, a worry is that there are cases in which it seems fine to say ‘It might hit the bulls eye’ even if one has a credence of zero in the proposition that it will hit the bulls eye. For a more sophisticated discussion of how we can motivate and make sense of a non-factualist or expressivist view about epistemic modality, see e.g. Yalcin (2007, 2010, 2011, 2012). Blackburn (1984, 1998) and Gibbard (1990, 2003) are often mentioned as prominent contemporary defenders of expressivism in moral philosophy.

  12. 12.

    Dreier (1999) makes a similar point with respect to expressivism in moral philosophy. Insofar as an expressivist is in a position to tell a plausible story about disagreement, it is not clear why a contextualist cannot tell more or less the same story. See e.g. also Jackson and Pettit (1998). For an application of this strategy in defence of a contextualist treatment of predicates of taste, see Huvenes (2012).

  13. 13.

    Similar points have been made by e.g. Papafragou (2006) and Wright (2007).

  14. 14.

    There is a lot more to be said about the notion of doxastic advice. Portner (2009, p. 157) worries that Swanson (2006) needs to provide theory of the speech act of giving advice. This is not a trivial task. Another worry is that the claim that that the speaker is advising her interlocutor to have a certain minimal credence in the embedded proposition is too strong. The proposal would make certain uses of ‘might’ inappropriate. However, in many cases in which the speaker knows that her interlocutor is better informed than she is, it is not clear that is appropriate to use an epistemic modal like ‘might’. For instance, let us suppose that I have a heard a rumour that you are leaving California. I take it to be compatible with what I know that the rumour is true. However, it is not clear that it would be appropriate for me to say ‘You might be leaving California’ in a conversation with you. This is something we can explain if epistemic modals are used to give doxastic advice. It would not make sense for me to give you doxastic advice when you are better informed than me.

  15. 15.

    See e.g. also Cappelen and Hawthorne (2011a) and Weatherson (2011) for relevant discussion.

  16. 16.

    I want to thank an anonymous referee for Philosophical Studies for encouraging me to address these questions.

  17. 17.

    I am assuming that ‘could’ can be used as an epistemic modal, at least when it occurs inside the scope of a negation. If we wanted a case in which the negation is scoping over ‘might’, we could have used the slightly more cumbersome sentence ‘It is not the case that Harry might be in the office’. However, this sentence is less natural than the sentence involving ‘could’. It therefore makes more sense to use ‘could’ for these purposes.

  18. 18.

    There is a debate about how strong the semantics for ‘must’ ought to be. See e.g. von Fintel and Gillies (2010) for relevant discussion. For the purpose of the present discussion, I am assuming a simple semantics for ‘might’ and ‘must’.

  19. 19.

    It is important to remember that it is not a part of the proposal under discussion that correction and retraction always have to target the embedded proposition. Furthermore, in a case like this, when the epistemic possibility modal occurs inside the scope of a negation, there does not seem to be any obvious reason for thinking that the embedded proposition would be available as a target of correction. It makes little sense to think of John as advising his audience to have a certain positive credence in the proposition that Harry is in the office.

  20. 20.

    See e.g. Stalnaker (1980) for a suggestion along these lines.

  21. 21.

    According to the proposal under discussion, the embedded proposition is the conditional proposition that if Harry is in the office, he is sleeping. The proposal does not require us to adopt a particular theory of conditionals or conditional propositions. However, one might question whether there is even such an embedded conditional proposition. For instance, according to Kratzer (1981, 1991a), ‘if’-clauses restrict other operators. If we treat the ‘if’-clause as restricting the epistemic modal ‘might’, there does not seem to be any embedded conditional proposition that the correction could be targeting. More generally, if it turns out that there are no conditional propositions of the right sort, we would have to look for a different way to account for cases of correction involving conditionals. Seeing as a thorough treatment of different theories of conditionals is beyond the scope of the present discussion, I will continue to assume that there are conditional propositions of the right sort. However, there are lot of questions here that merit further investigation.

  22. 22.

    For the purpose of the present discussion, I am assuming that the relevant cases involve epistemic modality. See e.g. Swanson (2010, p. 531) for relevant discussion.

  23. 23.

    There are other views according to which many propositions are expressed or asserted. See e.g. Cappelen and Lepore (2005, ch. 13) and Dorr (2014) for relevant discussion. In this context, it is worth pointing out that we do not have to think of the conversational participants as having all of these propositions in mind, at least not in a strong sense of ‘having in mind’. For instance, there is no obvious reason why the conversational participants would need to have occurrent mental states involving all of the propositions.

  24. 24.

    It is important to distinguish the idea that there is a potentially infinite number of relevant propositions from the idea that the relevant speech act involves a single proposition which is the conjunction of these propositions. The speaker might know that this proposition is false and it makes little sense to think that this is the proposition that is at issue. However, it would surprising if the speaker knew that all of the individual propositions were false.

  25. 25.

    In the case of non-universal quantifiers such as ‘most’ and ‘almost every’, the situation is a bit more complicated. However, I do not see any insurmountable obstacles to extending the proposed account to non-universal quantifiers. In order to simplify the discussion, let us consider the sentence ‘Most of the square inches of the mural might have paint on them’. As before, we have the proposition that square inch one of the mural has paint on it, the proposition that square inch two of the mural has paint on it, and so forth. The idea is that the sentence can be understood as conveying the advice to have a certain positive credence in most of these propositions. In order for someone to reject this advice and be in a position to correct the speaker, it must be the case that they do not have a sufficiently high credence in most of these propositions.

    One might worry that this amounts to a disjunctive speech act and that this is somehow problematic. See e.g. Krifka (2001) and Swanson (2010, p. 532) for relevant discussion. However, in this particular case, it looks like there is nothing obviously wrong with the relevant advice. It is reasonable clear what it would take to follow the advice and what it would take to reject the advice.

  26. 26.

    A worry is that the proposal presupposes that the speaker has a credence in each of the relevant propositions. Insofar as this is taken to be problematic, it is worth pointing out that this can still be understood in dispositional terms or in terms of the speaker have a commitment to having a certain positive credence in each of the relevant propositions.


  1. Blackburn, S. (1984). Spreading the word. Oxford: Clarendon Press.

    Google Scholar 

  2. Blackburn, S. (1998). Ruling passions. Oxford: Clarendon Press.

    Google Scholar 

  3. Cappelen, H., & Hawthorne, J. (2009). Relativism and monadic truth. Oxford: Oxford University Press.

    Google Scholar 

  4. Cappelen, H., & Hawthorne, J. (2011a). Reply to Glanzberg, Soames, and Weatherson. Analysis, 71, 143–156.

    Article  Google Scholar 

  5. Cappelen, H., & Hawthorne, J. (2011b). Reply to Lasersohn, Macfarlane, and Richard. Philosophical Studies, 156, 449–466.

    Article  Google Scholar 

  6. Cappelen, H., & Lepore, E. (2005). Insensitive semantics. Oxford: Blackwell.

    Google Scholar 

  7. DeRose, K. (1991). Epistemic possibilities. Philosophical Review, 100, 581–605.

    Article  Google Scholar 

  8. Dever, J. (2011). Epistemic modals. In S. Bernecker & D. Pritchard (Eds.), Routledge companion to epistemology (pp. 545–557). London: Routledge.

    Google Scholar 

  9. Dorr, C. (2014). Transparency and the context-sensitivity of attitude reports. In M. García-Carpintero & G. Martí (Eds.), Empty representations. Oxford: Oxford University Press.

    Google Scholar 

  10. Dougherty, T., & Rysiew, P. (2009). Fallibilism, epistemic possibility, and concessive knowledge attributions. Philosophy and Phenomenological Research, 78, 123–132.

    Article  Google Scholar 

  11. Dowell, J. (2011). A flexible contextualist account of epistemic modals. Philosopher’s Imprint, 11, 1–25.

    Google Scholar 

  12. Dreier, J. (1999). Transforming expressivism. Noûs, 33, 558–572.

    Article  Google Scholar 

  13. Egan, A. (2007). Epistemic modals, relativism and assertion. Philosophical Studies, 133, 1–22.

    Article  Google Scholar 

  14. Egan, A., Hawthorne, J., & Weatherson, B. (2005). Epistemic modals in context. In G. Preyer & G. Peter (Eds.), Contextualism in philosophy (pp. 131–168). Oxford: Clarendon Press.

    Google Scholar 

  15. Gibbard, A. (1990). Wise choices, apt feelings. Oxford: Clarendon Press.

    Google Scholar 

  16. Gibbard, A. (2003). Thinking how to live. Cambridge, MA: Harvard University Press.

    Google Scholar 

  17. Goldman, A. (2010). Epistemic relativism and reasonable disagreement. In R. Feldman & T. Warfield (Eds.), Disagreement (pp. 187–215). Oxford: Oxford University Press.

    Google Scholar 

  18. Hacking, I. (1967). Possibility. Philosophical Review, 76, 143–168.

    Article  Google Scholar 

  19. Hawthorne, J. (2007). Eavesdroppers and epistemic modals. Philosophical Issues, 17, 92–101.

    Article  Google Scholar 

  20. Hawthorne, J. (2012). Knowledge and epistemic necessity. Philosophical Studies, 158, 493–501.

    Article  Google Scholar 

  21. Huvenes, T. (2012). Varieties of disagreement and predicates of taste. Australasian Journal of Philosophy, 90, 167–181.

    Article  Google Scholar 

  22. Jackson, F., & Pettit, P. (1998). A problem for expressivism. Analysis, 58, 239–251.

    Article  Google Scholar 

  23. Jehle, D., & Fitelson, B. (2009). What is the ”equal weight view”? Episteme, 12, 280–293.

    Article  Google Scholar 

  24. Joyce, J. (2010). A defense of imprecise credences in inference and decision making. Philosophical Perspectives, 24, 281–323.

    Article  Google Scholar 

  25. Kratzer, A. (1977). What Must and Can must and can mean. Linguistics and Philosophy, 1, 337–355.

    Article  Google Scholar 

  26. Kratzer, A. (1981). The notional category of modality. In H.-J. Eikmeyer & H. Rieser (Eds.), Words, worlds, and contexts (pp. 38–74). Berlin: de Gruyter.

    Google Scholar 

  27. Kratzer, A. (1991a). Conditionals. In A. von Stechow & D. Wünderlich (Eds.), Semantik/semantics: International handbook of contemporary research (pp. 651–656). Berlin: de Gruyter.

    Google Scholar 

  28. Kratzer, A. (1991b). Modality. In A. von Stechow & D. Wünderlich (Eds.), Semantik/semantics: International handbook of contemporary research (pp. 639–650). Berlin: de Gruyter.

    Google Scholar 

  29. Krifka, M. (2001). Quantifying into question acts. Natural Language Semantics, 9, 1–40.

    Article  Google Scholar 

  30. MacFarlane, J. (2011). Epistemic modals are assessment-sensitive. In A. Egan & B. Weatherson (Eds.), Epistemic modality (pp. 144–178). Oxford: Oxford University Press.

    Google Scholar 

  31. Papafragou, A. (2006). Epistemic modality and truth conditions. Lingua, 116, 1688–1702.

    Article  Google Scholar 

  32. Portner, P. (2009). Modality. Oxford: Oxford University Press.

    Google Scholar 

  33. Price, H. (1983). Does ’probably’ modify sense? Australasian Journal of Philosophy, 61, 396–408.

    Article  Google Scholar 

  34. Schaffer, J. (2011). Perspective in taste predicates and epistemic modals. In A. Egan & B. Weatherson (Eds.), Epistemic modality (pp. 179–226). Oxford: Oxford University Press.

    Google Scholar 

  35. Simons, M. (2007). Observations on embedding verbs, evidentiality, and presupposition. Lingua, 117, 1034–1056.

    Article  Google Scholar 

  36. Stalnaker, R. (1980). A defense of conditional excluded middle. In W. Harper, R. Stalnaker, & G. Pearce (Eds.), Ifs: Conditionals, belief, decision, chance, and time (pp. 87–104). Dordrecht: Reidel.

    Google Scholar 

  37. Stephenson, T. (2007). Judge dependence, epistemic modals, and predicates of personal taste. Linguistics and Philosophy, 30, 487–525.

    Article  Google Scholar 

  38. Swanson, E. (2006). Interactions with context. Unpublished Doctoral Dissertation, MIT.

  39. Swanson, E. (2010). On scope relations between quantifiers and epistemic modals. Journal of Semantics, 27, 529–540.

    Article  Google Scholar 

  40. von Fintel, K., & Gillies, A. (2007). An opinionated guide to epistemic modality. In T. S. Gendler & J. Hawthorne (Eds.), Oxford studies in epistemology (Vol. 2, pp. 32–62). Oxford: Oxford University Press.

    Google Scholar 

  41. von Fintel, K., & Gillies, A. (2008). CIA leaks. Philosophical Review, 117, 77–98.

    Article  Google Scholar 

  42. von Fintel, K., & Gillies, A. (2010). Must..stay..strong!. Natural Language Semantics, 18, 351–383.

    Article  Google Scholar 

  43. von Fintel, K., & Gillies, A. (2011). Might made right. In A. Egan & B. Weatherson (Eds.), Epistemic modality (pp. 108–130). Oxford: Oxford University Press.

    Google Scholar 

  44. Weatherson, B. (2011). No royal road to relativism. Analysis, 71, 133–143.

    Article  Google Scholar 

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

    Google Scholar 

  46. Wright, C. (2007). New age relativism and epistemic possibility: The question of evidence. Philosophical Issues, 17, 262–283.

    Article  Google Scholar 

  47. Yalcin, S. (2007). Epistemic modals. Mind, 116, 983–1026.

    Article  Google Scholar 

  48. Yalcin, S. (2010). Probability operators. Philosophy Compass, 5, 916–937.

    Article  Google Scholar 

  49. Yalcin, S. (2011). Nonfactualism about epistemic modality. In A. Egan & B. Weatherson (Eds.), Epistemic modality (pp. 295–332). Oxford: Oxford University Press.

    Google Scholar 

  50. Yalcin, S. (2012). Bayesian expressivism. Proceedings of the Aristotelian Society, 112, 123–160.

    Google Scholar 

Download references


Earlier versions of this paper were presented at the Arché Contextualism and Relativism Seminar (University of St Andrews, March 2009) and the Arché Language, Context, and Truth Seminar (University of St Andrews, July 2010). While the paper has changed substantially since then, I am still grateful to audiences on those occasions for useful comments and criticisms. I would also like to thank Derek Ball, Herman Cappelen, Josh Dever, Jon Litland, Ofra Magidor, Dilip Ninan, Jonathan Schaffer, Anders Schoubye, Andreas Stokke, Brian Weatherson, and an anonymous referee for helpful discussion and comments.

Author information



Corresponding author

Correspondence to Torfinn Thomesen Huvenes.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Huvenes, T.T. Epistemic modals and credal disagreement. Philos Stud 172, 987–1011 (2015). https://doi.org/10.1007/s11098-014-0334-8

Download citation


  • Context-dependence
  • Contextualism
  • Disagreement
  • Epistemic modals