Michael Bergmann

Purdue University


One thing all forms of foundationalism have in common is that they hold that a belief can be justified noninferentially -- i.e., that its justification needn’t depend on its being inferred from some other justified (or unjustified) belief.  In his 1999 and his 2000, Peter Klein argues that in virtue of having this feature, all forms of foundationalism are flawed.  Says Klein:

foundationalism is unacceptable because it advocates accepting an arbitrary reason at the base, that is, a reason for which there are no further reasons making it even slightly better to accept than any of its contraries. (1999, 297)

In this paper, I will explain why Klein’s objections to foundationalism fail.

     I’ll begin with a remark about a point of agreement between Klein and foundationalists. Klein is an infinitist.  An infinitist is (roughly) one who thinks that (i) a belief is justified only if it is inferred from another justified belief and that (ii) a belief B cannot be justified if B plays an essential role in the chain of reasons supporting B.  By endorsing (i), the infinitist rejects foundationalism and holistic coherentism; by endorsing (ii) the infinitist rejects linear coherentism.[1]  Klein admirably (though, in my opinion, unsuccessfully) defends this position which has been cavalierly dismissed by most philosophers.  I won’t take the time here to discuss his defense.[2]  I merely want to draw attention to a point he makes in giving it.  In his 2000 he objects to dogmatism.  Dogmatists believe they know things.  Because Klein believes that reason-giving must go on ad infinitum, he thinks that our beliefs can be provisionally justified at best.  Thus, one of the main points of his 2000 is that nothing is ever settled by reasoning.  But, contrary to what Klein suggests in certain places (e.g., 2000, 21), this last point is one with which foundationalists agree.  According to them, reasoning alone settles nothing; we need foundational starting points.  What foundationalists deny is an additional claim Klein endorses, namely, that only reasoning can settle things (‘settle’ in the sense that one can know the thing or be justified in believing it).  So to insist that nothing is ever settled by reasoning (alone) doesn’t count as an objection to foundationalism (and perhaps Klein never intended it as such).

    Let’s turn then to some lines of argument that Klein clearly does intend as objections to foundationalism.  The first employs a technical notion of ‘assent’.   According to Klein, assenting to p is being disposed to assert “I know that p because reasoning has settled that p is true” (2000, 9-10).  Klein asks us to consider a foundationalist, Fred, being interrogated by Sally about what his reasons are for holding some belief and about his reasons for those reasons.  Eventually, in response to Sally’s questioning, Fred identifies as a reason some belief b – a belief for which he has no further reason.  At this point, when Sally asks again for a reason, Fred (if he is a consistent foundationalist) will not come through with one.  Then, says Klein, “Fred ought (rationally) to give up assenting to b” (2000, 15). 

     But why think Fred ever assented to b to begin with?  Foundationalists agree that nothing is ever settled by reasoning so they will not “assent” (in Klein’s sense) to anything.  Fred may assent* to b (where assenting* to b is being disposed to assert “I know that b”).  But if he is a sensible foundationalist, he won’t assent (in Klein’s sense) to b or anything else.  So this objection creates no problem for foundationalism.

     Let’s look at another line of argument.  Klein insists that he is not claiming that he has proven that foundationalism is false.  What he claims is that “there is a deep irrationality in being a practicing foundationalist” (2000, 17).   The idea is this.  When Fred gets to a belief b for which he has no reason and Sally asks him why he thinks b is true, Klein thinks that Fred, being a foundationalist, will reply by pointing out that b is properly basic – i.e., b is justified noninferentially.  In response, Sally asks Fred why he thinks proper basicality is linked in any way to truth.  At this point, says Klein, Fred can’t (as a consistent foundationalist) give a reason for thinking proper basicality is linked in any way to truth.  Fred is thus forced to concede that he has no reason for thinking properly basic beliefs are linked in any way to truth -- he must  “embrace arbitrariness at the very foundation of all [his] beliefs” (2000, 17).  But, says Klein, “[w]hen faced with the skeptic’s question, rationality requires that the foundationalist attempt to show why the offered basic propositions are true or somewhat likely to be true” (2000, 17).  Hence the deep irrationality in being a practicing foundationalist. 

     A natural foundationalist reaction to this objection is to point out that Klein is simply begging the question.  The foundationalist thinks beliefs can be justified or rational even if not inferred from other beliefs.   Klein replies by pointing out that the foundationalist is irrational insofar as she holds a belief without being able to give a reason for it (“rationality requires that the foundationalist attempt to show why the offered basic propositions are true or somewhat likely to be true”).  But that just takes for granted that the foundationalist is mistaken in thinking a belief can be noninferentially justified or rational.  There doesn’t seem to be any objection here to foundationalism other than mere question-begging counter-assertion.

     At this point in the dialectic, we can turn from Klein’s 2000 to his 1999 and see that he will probably claim that he has been misunderstood.  In his 1999, 303 he explains what a meta-justification is.  It’s a justification designed to show that certain beliefs are justified, without reasons, in virtue of having some property P such that beliefs with P are likely to be true.  In connection with such a meta-justification he says:

I am not claiming that in order for a belief to be justified or known, either we must believe that it is justified or we must be justified in believing that it is justified.  As many have pointed out, that confuses p’s being justified with a belief about p’s justificatory status. ...  Quite the contrary, I think the foundationalist typically advocates an explicit process of reasoning that ends with beliefs which have P [the property conferring noninferential justification] rather than with epistemic beliefs about P.  The meta-justification is invoked in order to avoid the appearance of arbitrariness for it is designed to show why the “final” beliefs are likely to be true.  My point is merely that moving to the meta-level, that is, arguing that such beliefs are likely to be true because they possess a certain property, P, will not avoid the problem faced by foundationalism.  Either the meta-justification provides a reason for thinking the base proposition is true (and hence, the regress does not end) or it does not (hence, accepting the base proposition is arbitrary).  (1999, 303-4)

This is an interesting move.  The gist of it is that the foundationalist Fred cannot avoid the charge of arbitrariness by alluding to a meta-justificatory argument.  For in mentioning such an argument, he cannot be providing a reason for b since, as we’ve already said, he has no reason for b.  (And even if he were providing a reason for b, Sally would just ask Fred why he thinks the claims employed in the meta-justificatory argument are true and the regress would continue.)  But if in mentioning the meta-justificatory argument Fred is not offering a reason for b, then his mentioning it has done nothing to help him avoid the charge of arbitrariness.

     The problem with this response by Klein is that it misconstrues the role of drawing attention to the meta-justificatory argument.  If Fred is a sensible foundationalist, his discussion with Sally (in which he mentions the meta-justificatory argument) goes something like this:

Sally: Why do you think b is true?

Fred: If you are asking for a reason, the answer is that I don’t have one.

Sally:  If you have no reason for thinking b is true, then your belief that b is arbitrary.

Fred: But that’s to assume that a belief is arbitrary unless one has a reason for it.  I deny that assumption.  I say a belief for which one has no reason can avoid being arbitrary if it has some feature F such that beliefs having feature F are noninferentially justified.

Sally: Aha!  So b isn’t a basic belief for you after all.  You’ve just offered a reason for holding b. 

Fred: No I haven’t.  I’ve explained why I disagree with your claim that lacking a reason for b is a sufficient condition for b’s being arbitrary.

Sally: But do you believe b is arbitrary?

Fred: No. 

Sally: So you must believe that X where X is “b has some feature F such that beliefs having feature F are noninferentially justified”.

Fred: I wouldn’t say that I must believe X.  But I concede that, as a matter of fact, being a foundationalist who has reflected on whether my belief b is justified, I do believe X.

Sally: But that just shows that b isn’t basic for you.

Fred: Not true.  My belief that X plays no role in the justification of my belief that b. Before I ever grasped or considered or held beliefs like X, I justifiedly held beliefs like b.  Those beliefs like b were justified noninferentially in virtue of their having feature F, not inferentially on the basis of a belief like X.  The same is true of my current belief b.

Sally: But what reason do you have for thinking X is true?

Fred: We can discuss that question if you like.  But let’s be clear that you haven’t given me any reason for thinking that that question is in any way relevant to the justification of b.  It is my position that the justification of b depends on the truth of X, not on my justifiedly believing X or on my having a reason for X.

     The lesson here is this.  In drawing attention to the meta-justificatory argument, the foundationalist is not offering a reason for his basic beliefs.  Contrary to what Klein claims, the foundationalist is not invoking the meta-justification “in order to avoid the appearance of arbitrariness” by “show[ing] why the ‘final’ beliefs are likely to be true” (1999, 304).  Instead, the foundationalist is using the ideas in that meta-justificatory argument to explain why it is that lacking a reason for a belief isn’t sufficient for that belief’s being arbitrary.  The aim is to avoid the appearance of arbitrariness, not by providing a reason for b, but by casting doubt on an assumption about what is sufficient for b’s being arbitrary.

     To sum up: the objections to foundationalism given by Klein in his 1999 and his 2000 either fail to apply to foundationalism or are question-begging or are based on a misunderstanding of what the sensible foundationalist will say in response to socratic interrogation by a skeptic.[3]







BonJour, Laurence.  1985.  The Structure of Empirical Knowledge.  Cambridge, Massachusetts: Harvard University Press.


Fumerton, Richard.  1998.  “Replies to My Three Critics.” Philosophy and Phenomenological Research 58: 927-37.


Klein, Peter.  1999.  “Human Knowledge and the Infinite Regress of Reasons.”  In James Tomberlin (ed.), Philosophical Perspectives, 13, Epistemology, pp. 297-325.  Cambridge, Massachusetts: Blackwell Publishers.


________.  2000.  “The Failures of Dogmatism and A New Pyrrhonism.”  Acta Analytica 15: 7-24.


Moser, Paul.  1989.  Knowledge and Evidence.  New York: Cambridge University Press.


Plantinga, Alvin.  1993.  Warrant: The Current Debate.  New York: Oxford University Press.


Sosa, Ernest.  1980.  “The Raft and the Pyramid: Coherence Versus Foundations in the Theory of Knowledge.”  In P. French, T. Uehling and H. Wettstein (eds.) Midwest Studies in Philosophy 5, pp. 3-25.  Minneapolis: University of Minnesota Press.

[1] See BonJour 1985, 89-93 on the distinction between holistic and linear coherentism.  As Sosa (1980), Plantinga (1993, chapter 4) and Klein (1999 and 2000) rightly point out, holistic coherentism is just a version of foundationalism.

[2] But see Fumerton 1998 and Moser 1989, 56-60.

[3] Thanks to Trenton Merricks and Fritz Warfield for comments on earlier drafts.  Thanks also to the Pew Evangelical Scholars Program for funding during the time I worked on this paper.