The State of the Debate on Epistemic Circularity

Following Michael Bergmann, let’s say that S’s belief that X is a reliable source of true beliefs is epistemically circular if S used X in order to arrive at this belief.  A person who used memory to form the belief that my memory is reliable would have an epistemically circular belief.

I’m trying to understand the state of the debate in the epistemic circularity literature.  From what I can tell, Richard Fumerton (his 1995 book), Jonathan Vogel (2000, “Reliabilism Leveled”), and Michael Huemer (2011, his recent PPR paper on his metacoherence problem) all seem to think that justified, epistemically circular beliefs are impossible.  On the other hand, James van Cleve, William Alston, Michael Bergmann, James Pryor, Peter Markie, Ernest Sosa, and many others all think that it is possible for there to be justified, epistemically circular beliefs.

Three questions:

1) Am I correct to say that there is a growing consensus among nonskeptical epistemologists that it is possible to have justified, epistemically circular beliefs?

2) Are there others that I failed to mention who think that it is impossible for there to be justified, epistemically circular beliefs?

3) Certainly there are a great many cases in which epistemically circular beliefs are not justified.  For those who think it is possible to have justified, epistemically circular beliefs, who among them actually gives conditions for when they are justified and when they are not?  I know Bergmann does in his (2006) book.  I hear that van Cleve does in his 1984 induction paper and that Pryor does in his 2004 “What’s Wrong with Moore’s Argument?” paper.  (I still have to read the latter two papers.)  Anybody else?


The State of the Debate on Epistemic Circularity — 18 Comments

  1. Stewart Cohen in “Basic Knowledge and the Problem of Easy Knowledge” (2002) should probably be classed as another philosopher who denies the possibility of justified, epistemically circular beliefs (though I’m not completely sure about that).

    • Hi Daniel,
      Thanks for the reply. Cohen does think that certain types of epistemically circular belief are unjustified, specifically, beliefs that arise by way of the sort of closure or bootstrapping reasoning that he talks about; but he does not seem to think that it is impossible for them to be justified. He writes, “But it does not follow from the illegitimacy of bootstrapping, that any any way of coming to know a belief source is reliable by in part relying on that very source is objectionable… I shall argue that there are ways of [having justified epistemically circular beliefs] that do not allow for bootstrapping” (2002, 319). Cohen then presents his coherentist view to do just that.

  2. 1) Yes.

    2) Tim and Lydia McGrew have a chapter devoted to epistemic circularity in their book Internalism and Epistemology (2006). They take the externalists’ tu quoque head on and argue that it is impossible to have justified epistemically circular beliefs.

    [For what its worth, I’m about 2 months away from defending my dissertation entitled “Internalism, Externalism, and Epistemic Circularity”. Richard Fumerton is my advisor. In my third chapter I argue that the McGrews’ proposal fails on two accounts. If anyone is interested in talking about these issues more, feel free to send me an email: my name with a dash between]

    3) Not off the top of my head. That said, I discuss the issue in my dissertation and hope to publish on it soon.

  3. The McGrews are not skeptics; justified beliefs and knowledge are possible. The book is very tightly argued and their discussion of epistemic circularity actually runs throughout most of it in some way or other. Its worth looking at.

        Michael P. Lynch

        “.. Nonetheless, I also think that there is something right about our misgivings about the externalist answer to the criterion argument. What is right about those misgivings is that the externalist response has no traction against a different problem, distinct from scepticism.
        Like the sceptical argument above, this other problem is rooted in part in the issue
        of epistemic circularity. But unlike its cousin, it is not a problem about whether we in fact have knowledge or are justified in our opinions. It is about rationally resolving explicit disagreement over the reliability of our most basic methods for forming beliefs.
        Although I will not try to show it here, this latter problem, what I will call the problem of epistemic incommensurability, is arguably the root worry behind the criterion argument, and, in my view, the reason why Chisholm was right to say that the latter was ‘one of the most important and difficult of all problems of philosophy’ (1982: 61).”

        Stephen: Chisholm, referred to in the last sentence above mentions Peter Coffey, Cardinal Mercier,Kant and Hume in his discussion.

        Roderick Chisholm, The Foundations of Knowing (1982)
        “The problem of the criterion” seems to me to be one of the most important and one of the most difficult of problems of philosophy. I am tempted to say that one has not begun to philosophize until one has faced this problem and has recognized how unappealing, in the end, each of the possible solutions is.”

        Stephen: My overlapping interest in this topic is captured by the phrase,
        ‘sufficient antecedent causal condition’.

  4. First, it is worth noting that in his newest article on bootstrapping and easy knowledge “Bootstrapping, Defeasible Reasoning, and A Priori Justification” Cohen abandons the holism that he adopted originally to get out of the problem. In the article he argues that we have a priori justification for the relevant reliability claims. Given such a claim, I’m not sure if he would still need to allow epistemically circular beliefs to be justified.

    Second, Fumerton’s views on this issue seem to be more complicated as he does seems to think that one could use acquaintance to justify the existence (and possibly the reliability) of acquaintance. He just thinks that the track-record argument for acquaintance’s reliability (while it could provide justification) is sort of pointless since you have better non-inferential justification available for this claim. His “Epistemic Internalism, Philosophical Assurance, and the Skeptical Predicament” gives a more careful articulation of these issues.

    Thirdly, I think Sosa’s view on these issues have also become increasingly complex. In Reflective Knowledge volume 2 he seems to allow epistemically circular arguments to provide reflective justification (since reflective justification is given a coherentist model) but it seems as if he disallows epistemically circular arguments from providing animal justification for beliefs about reliability. His account of how “commitments” or “presuppositions”, as he calls them, of a source’s reliability are animally justified seems to make these commitments possess a quasi-a priori status.

    Fourth, it is at least interesting to note that Vogel’s gas gauge example has the exact same structure as Cohen’s bootstrapping cases. I’ve always wondered why Vogel thought that the problem was epistemic circularity generally rather than the particular form of circular reasoning found in this bootstrapping.

    And finally, I’m not sure but wouldn’t Crispin Wright’s theory of entitlement qualify as a non-skeptical epistemology that doesn’t allow epistemically circularity? (Note: I think Wright’s view shows that you need to clear up question 1 since an epistemically circular belief, as you’ve defined it, might be justified but this justification might not derive from the relevant belief source that a subject used in forming her belief about the source’s reliability. In Wright’s view the justification is a default justification and doesn’t derive from any belief source. Sorry I don’t think I explained this as clearly as I would’ve liked).

    • Sam,
      Thanks very much for these references, especially the Cohen one. (I’m happy to hear he dropped his holism.) And the Fumerton article is in that Plantinga volume, right?

      Hmm, so, like the Cohen 2002, it may be that Vogel’s argument applies to only certain sorts of epistemic circularity, not all epistemic circularity.

      On Crispin Wright, I don’t know his work too well. Any article you could point me to? I also couldn’t understand your reason for thinking that my question 1 needed to be cleared up.

  5. Wright definitely doesn’t accept epistemic circularity. A good place to start is “Warrant for nothing (and foundations for free)?” Aristotelian Society Supplementary Volume, 2004. After that you could try “The perils of dogmatism”, Themes from GE Moore, 2007.

    Roger White seems to have a view similar to Wright’s in “Problems for dogmatism”, Phil Studies, 2006.

    I’d also recommend Ram Neta, “Liberalism and conservatism in the epistemology of perceptual belief”, AJP, 2010. It has a good discussion of Wright’s views, comparing it to the views of some others.

    Any view according to which one has a priori warrant or justification for believing that your sensory faculties are reliable such as Wright’s and White’s allows for one to deny epistemic circularity without succumbing to skepticism.

    • If even if one can provide an a priori justification for thinking that one’s sensory faculties are reliable, won’t one’s justification for thinking one’s a priori “faculties” are reliable have to be epistemically circular?

      • Maybe. Suppose one holds a theory according to which one can be noninferentially justified in believing that one’s a priori faculty is reliable. If epistemic circularity is solely a property of inferences, then this belief couldn’t be epistemically circular. However, Bergmann (2006) and the McGrews (2006) both specify a sense in which noninferential beliefs can be epistemically circular. Note that the belief is both produced by the faculty it says is reliable and is justified only if its content is true. There is some sense of dependence on the source in forming the belief, but I’m not sure if that sort of case is one of problematic epistemic circularity, if circular at all.

        • Well, in my opening post, I specified the sense of ‘epistemic circularity’ I was interested in, and I was focusing on beliefs. And yes, if my faculty F directly and noninferentially produces the belief that F is reliable, then that belief is epistemically circular. So, I was wondering if there are theorists who think epistemically circular beliefs can’t be justified (where I am using ‘epistemically circular’ as defined above). This was Question #2.

          Dylan, maybe Wright would be okay with epistemically circular beliefs being justified when it is specified this way? Do you happen to know?

          • Andrew,

            I don’t know for sure, but I suspect he wouldn’t be okay with it. I think he’s motivated in large part by a desire to avoid skepticism while not endorsing epistemic circularity. In my previous post, I actually forgot some of his most obvious papers, which are the ones on transmission failure or begging the question. Check out his “(Anti-) Sceptics subtle and simple” (PPR 2002) and “Cogency and question-begging” (Philosophical Issues 2010).

            I was only thinking of perceptual beliefs. However, I think Wright wouldn’t allow even for epistemically circular justified beliefs about one’s a priori faculties (although I’ve never read or heard him address this issue in particular).

            Wright’s framework is somewhat baroque. He thinks there are “Hinge propositions”. These are propositions which, for a vast region of one’s thought, one assumes they’re true. and would be hinge propositions. You’re entitled to trust that these propositions are true. However, they’re not justified, and he thinks you don’t believe them. They’re not ever justified, and you don’t ever believe them. They are warranted, and you do trust that they’re true. Like Burge, Wright takes justified and entitled beliefs to be two species of the genus warranted beliefs. Wright denies single premise closure for justification. Aidan McGlynn has a forthcoming paper where he points out that Wright is committed to closure failing even for disjunction introduction (i.e., according to Wright’s framework, there will be cases where S has justification for believing p, but S lacks it for believing p or q). McGlynn’s paper has a nice and very understandable exposition of Wright’s framework. You might consider emailing him for the paper. But I digress.

            Anyway, as far as Wright’s view on epistemically circular beliefs about one’s a priori faculties goes, he could say is this. is a hinge proposition. Thus I’m entitled to trust that it’s true, but there’s no justification for believing it, and in fact I don’t believe it. Being entitled to trust that it’s true does serve as a basis for my having a priori justified beliefs though.

            Hopefully this sort of makes sense. There are details I’m leaving out. One detail is that I haven’t explained is why he’s forced to deny closure for justification.

  6. I think Peter Klein would fit the bill in response to your Question 2. In his “Human Knowledge and the Infinite Progress of Reasoning,” he says:

    “If there is a first premise [in reasoning] either it has appeared in the series earlier or it hasn’t. … If it has appeared earlier, then a proposition is being employed in its own evidential ancestry and circular reasoning has occurred and such reasoning cannot produce knowledge.”

Leave a Reply

Your email address will not be published. Required fields are marked *