Two new objections to explanationism

Abstract

After a period of inactivity, interest in explanationism as a thesis about the nature of epistemic justification has been renewed. Poston (Reason and explanation: a defense of explanatory coherentism, 2014) and McCain (Evidentialism and epistemic justification 2014; Erkenntnis 79:99–109, 2014) have both recently offered versions of explanationist evidentialism. In this paper, we pose two objections to explanationist evidentialism. First, explanationist evidentialism fails to state a sufficient condition for justification. Second, explanationist evidentialism implies a vicious regress.

1 Introduction

This paper is about explanationist evidentialism. Explanationist theories of justification are beginning to get traction. We think those theories are false; in this paper we offer reasons to think so. Our criticisms have two targets. Some of our arguments target the evidentialist side of explanationist evidentialism; others, the explanationist side. We do this to be thorough. We want to illuminate the general structure of explanationist evidentialist theories of justification, to highlight problems for any view in the category rather than pointing to difficulties in the details of views. Worries about technicalities can, after all, often be resolved by clever, technical argument. That is to say, our goal is not merely to show that the details of some particular explanationist evidentialist view are wanting but rather to show that explanationist evidentialism is problematic as an analysis of justification. In the next section (Sect. 2) we will briefly describe the commitments, strong and weak, of explanationist evidentialists. After describing the view’s commitments we will consider (Sect. 3) the threshold objection(s), which claims that explanationism fails to give a sufficient condition for justification. We then argue (Sect. 4) that a form of evidentialism that includes explanationist evidentialism is open to a vicious regress argument before concluding (Sect. 5).

2 The nature of evidentialist explanationism

Evidentialist explanationism is a combination of evidentialism and explanationism. Evidentialism has come to be a widely accepted view in epistemology, analyzing justification by asserting that one is justified in believing a proposition p if and only if p fits one’s evidence:

• (Evidentialism)       p is justified for S if and only if p fits S’s evidence e.

Our target in this paper is the internalist, mentalist version of evidentialism,¹ which identifies one’s evidence with one’s mental states. Explanationist evidentialism, then, is committed to the following, at least in its most popular form²:

• (Mentalism)       If e is evidence for S that p, e is a mental state of S’s.

A view that combines evidentialism with the following modest explanationist thesis will be classified as explanationist evidentialism:

• Minimal Explanationism (ME): A proposition p is justified for subject S if p is part of the best explanation available to S for S’s total evidence e.³

Explanationists may hold something stronger (usually consisting of making the thesis biconditional), but all should grant this minimal commitment. The explanationist commitment is best understood as a way of analyzing the evidential fit or epistemic support relation (McCain 2014a).⁴ Granting this it is hard to countenance an explanationist view that isn’t evidentialist and internalist. This (among other reasons to be discussed below) is because the intuitions that drive explanationism are intuitions regarding the inability to understand the internalist epistemic support relation in any other way: explanationists favor the Harman view in an old debate between Richard Fumerton and Gilbert Harman.⁵

Whereas Fumerton (1980) argues that all nondeductive inference (including inference to the best explanation) is reducible to enumerative induction, Harman (1965) argues that any good ampliative inference⁶ should be understood as an instance of inference to the best explanation because it is impossible to bridge the gap between our premises and the conclusion that those premises warrant without understanding it as a case of inference to the best explanation.

Finally, it is important to understand that explanationism is non-skeptical: an explanationist in all probability will not be a skeptic. There are several reasons that this should be clear. First of all, explanationism’s major defenders (Poston 2014; McCain 2014a, b; Conee and Feldman 2008; Feldman and Conee 2004b) all attempt to address the worries of external world skepticism, which at least suggests that they are interested in avoiding global skepticism. Further, one of the main concerns that motivates explanationism is a perceived failure of other views to account for the idea that certain ampliative inferences are justified.⁷ This motivation is impossible without a commitment against global skepticism. But this is not to say that explanationism is necessarily non-skeptical. Explanationism (logically) could be correct about the analysis of evidential fit while no one ever actually satisfies the requirements of justification.⁸

3 The threshold problem for explanationism

In this section we will propose two threshold problems for explanationism. There are two different kinds of threshold problem that could arise for explanationism. First there is the general problem of getting best explanation to yield a probability of above .5, which we address only very briefly here. This kind of threshold problem comes from the fact that something can be the best explanation of one’s evidence without being more probable than not, since it’s always possible that there are a few competing explanations. No explanationist, in our opinion, has given an analysis that can avoid this problem, but we set this problem aside. The second kind of threshold problem arises when one allows fallibly justified propositions to justify other propositions in a non-entailing way. Threshold problems of this nature arise in different ways, but the core idea is that a number of formulations of explanationism fail to state a sufficient condition for justification because they avoid analyzing justification in terms of epistemic probability.

3.1 The threshold problem for Ex-Ej 2.0

Kevin McCain, in his version of explanationism, attempts to account for cases that are intuitively inferentially justified for a subject but that, it seems, cannot be part of an explanation of a subject’s evidence. Byerly (2012) proposes a case in which one is golfing, having had a good amount of success. One putts the ball and watches, seeing it going toward the hole. One becomes justified, by this observation, in believing that the ball will roll into the hole. Intuitively, no proposition about the ball rolling into the hole in the future explains the golfer’s now having the visual experience of seeing the ball go toward the hole. The problem for explanationism is that propositions about the future often seem justified, but it is implausible that their justification is a result of being part of the best explanation of one’s present evidence.

In response to this objection from Byerly (2012) and related objections from Byerly and Martin (2015), McCain proposes Ex-Ej 2.0. Ex-Ej 2.0 states the following:

Ex-Ej 2.0- A person S, with evidence e at t is justified in believing p at t iff either

• (i) p is part of the best explanation available to S at t for why S has e, or

• (ii) p is available to S as an explanatory consequence of the best explanation available to S at t for why S has e (p would be better explained by the best explanation of S’s evidence available to S at t than not-p would). (McCain, Explanationism: Defended on All Sides, 2015)

A key weakness in Ex-Ej 2.0 is that p can be part of the best explanation of one’s evidence e but be relatively weakly supported by e.⁹ In fact, it can be just above whatever the threshold for justification may be. For explanationism, it should reasonably be thought to be just over .5 as an explanationist will allow that p better explaining one’s evidence than not-p will yield justification. All this will require is for one’s justification to yield a probability of p above .5.¹⁰ That is, the explanationist will want to gloss her theory as letting the mere fact of providing the best explanation for some evidence yield a probability of above .5. The exact probability value any particular explanation gets will depend on the strength of the explanation.¹¹

But if we set the probability yielded by the first clause to just above .5,¹² then any proposition that is justified on the second clause based on the propositions this first clause justifies to just above .5 will fall below the threshold for justification¹³ provided that the first-clause propositions don’t entail the second-clause propositions.¹⁴ But if Ex-Ej 2.0 calls ‘justified’ propositions that fall below the threshold for justification, then Ex-Ej 2.0 doesn’t state a sufficient condition for justification.

Consider the following example:

Bill the Bowler: Bill has been bowling with spotty success and is neither confident nor doubtful of his ability to get a strike on any given throw. He releases the ball as it goes toward the bowling pins. His evidence includes many times observing his ball roll toward the pins. He seems to remember (with his fuzzy and just barely reliable memory) that he gets a strike slightly more than half the time that the ball rolls toward the pins in this particular way, but he’s not sure. The conjunction of $<$This ball is going toward the pins in circumstances C$>$ and $<$Most balls going toward the pins in circumstances C will knock over all of the pins$>$ is just barely the best explanation of his experiential evidence, such that it ekes by in fulfilling Ex-Ej 2.0 clause (i). Now, $<$The ball will knock over all of the pins$>$ is better explained by the above conjunction than is $<$It’s not the case that the ball will knock over all of the pins$>,$ but barely. (“Most F’s are G’s” means “More than half of F’s are G’s.”) Therefore, $<$The ball will knock over all of the pins$>$ is explained (marginally) better than its negation.

By Ex-Ej 2.0, Bill the Bowler is justified in believing that the ball will knock over all the pins. But that is at least doubtful. This belief should be far below the threshold for justification. But Ex-Ej 2.0 implies it is justified. So, Ex-Ej 2.0 does not state a sufficient condition for justification.

Since this probability objection merely established that the conditions stated in Ex-Ej 2.0 are insufficient for justification, McCain could strengthen the connection between clause (i) and clause (ii) propositions so that what is justified by clause (ii) does not fall below the degree to which a clause (i) justifying proposition is justified. He could add a condition or strengthen the connection between clause (i) propositions and clause (ii) propositions to entailment, as he originally suggested in his (2013). But these strategies, we worry, are ill-fated. The added conditions will have to explicitly or implicitly require the overall epistemic probability of the consequence proposition not to fall below the threshold. If the view explicitly adds a probability condition, the view threatens to concede that explanatory relations are not epistemically fundamental.¹⁵ As an explanationist, McCain should propose an analysis that uses explanatory facts in order to remain in the spirit of explanationism.

In a recent paper, McCain attempts to solve the problem by adding further conditions on how good an explanation must be in order to confer justification. He says:

In order for S to be justified in believing that p it must not only be the best available explanation of S’s evidence, it must also be a sufficiently good explanation of S’s evidence. Similarly, in order for S to be justified in believing an explanatory consequence, p, of the best available explanation of her evidence it has to be that the best available explanation of her evidence would explain p significantly better than it would $\sim$p. (McCain 2015)

The problem is that, whatever “sufficiently good” amounts to in the first clause, either the problem will reappear or the account will be overly skeptical. If the first clause’s sufficiency condition requires anything higher than the probabilistic threshold of justification,¹⁶ then Ex-Ej 2.0 will leave out cases it should call justified and thus thereby promote a more skeptical view of the extent of justification than Ex-Ej 2.0 seems initially to suggest. If the standard merely requires that the proposition be justified to the threshold when justified by the first clause, then the very same problem arises as long as “significantly better” is weaker than entailment: simply set the probability conferred by the (first clause) best explanation justification to just near the threshold. The “significantly better” condition on the second clause cannot be high enough to escape the threshold problem.

Consider an example. Suppose that “sufficiently good” and “significantly better” each amount to requiring that the proposition is probable to degree .71. That is, first clause propositions will be at least justified to .71, and the second clause propositions will at least get a .71 probability conditional on the first clause propositions. This is a convenient number for our sufficiency conditions as these thresholds will be just enough to guarantee a probability of just above .5 for propositions justified on the second clause. But there is now a dilemma. Either the thresholds are .71 because this is really the minimal probability for a proposition to be worthy of belief or the thresholds are this high for some other reason. On the first horn, the original threshold problem simply recurs at a higher threshold (and, as stated above, any threshold for “significantly better” below entailment will get one below the threshold). On the second horn, the first clause of McCain’s proposal will fail to regard as justified a number of beliefs that are perfectly respectable. This applies to any belief justified on the first clause that’s at least as probable on one’s evidence as those second clause propositions which are above .5 probable but below .71. To put it another way, in this attempted solution we have either ratcheted up our standards for first clause beliefs too far or we have lowered our standards for second clause beliefs too far.¹⁷

But suppose that instead of adding a sufficiency condition to each clause, one adds an implicit probability condition that bars clause (ii) propositions from being considered justified while falling below a probability of .5. The condition would have to involve best explanations in order to remain explanationist. What would such a limit look like? It seems that an implicit characterization will either have to require that, on the whole, the justified proposition is part of the best explanation of one’s evidence (this proposal we take Byerly and Martin to have challenged with the golfer case) OR one will have to require that one’s evidence better explains p than it explains not-p. But surely McCain has avoided this with Ex-Ej 2.0 because there is no plausible sense in which my mental states, caused by the world out there, explain any proposition about those things which cause my experience. My evidence doesn’t explain facts about bowling balls hitting pins or golf-balls going into holes any more than a videotape of a tornado explains the tornado’s happening.

This problem affects explanationism generally. Sometimes we are justified in believing propositions (for example, about the future) that don’t explain our evidence/experience, but are in some way consequences of the truth of propositions that explain our evidence/experience. These latter propositions are what McCain calls ‘explanatory consequences’ of propositions that are justified by explaining evidence. Further, some of these consequence propositions that we are justified in believing are not entailed by the propositions that explain our experience (for example, the belief that the golf ball will go into the hole based on its trajectory and speed). Explanationists need to provide an account that allows for the justification of consequence propositions. The challenge we have presented is that any two-factor account like Ex-Ej 2.0 will run into this threshold problem. The problem is that, since the only thing the primary explanatory factor in these two-factor accounts guarantees is that the propositions which best explain one’s experience are more probable than not (i.e., that they are above .5 probable) there is no way (short of entailment) to arbitrarily make the second factor require a high enough probability that fulfilling this second condition will be sufficient for justification. The explanationist is caught between a rock and a hard place. The explanationist must either deny a number of plausible cases of justified belief or give up too much ground to those that think epistemic probability is more fundamental than explanation.

3.2 The threshold problem for ME $+$ fallible evidence

There is another threshold problem for explanationists who allow fallible evidence (FE) to confer justification. Suppose we take ME and we add the further thesis that we call FE.

Fallible Evidence (FE): Some fallibly justified propositions can be members of one’s body of evidence.

Explanationists able to escape the previous problem but who accept (ME) and (FE) face a new threshold problem. Recall that according to ME any proposition that is above .5 probable on one’s evidence is justified and that propositions that best explain one’s evidence are not much more than .5 probable. But assuming that FE is true, so that the probability of one’s evidence is not always 1.0, (ME) + (FE) will call ‘justified’ propositions that are less probable than their negation. Given that one’s evidence is barely well-enough explained by the proposition in question that the probability P(p/e) is just above .5, the final probability of p will fall below .5 while satisfying ME.¹⁸

This threshold problem is less general than the previous one because the explanationist can avoid it by accepting infallibilism about evidence or by denying that propositions are evidence. A non-propositionalist about evidence like McCain can avoid this problem because the only things that count as evidence for him are facts of which one is aware and seeming states, neither of which are candidates for being justified, whether fallibly or infallibly.

We conclude that explanationist evidentialism fails as an analysis of the epistemic support relation because it fails to provide a sufficient condition for justification.¹⁹

4 Evidentialism, explanationism, and regress

In this section we argue that explanationist evidentialism is committed to a vicious regress that leads to skepticism. This is a problem for the view because none of the commitments that lead to the regress can be easily discarded by explanationists. Specifically, the conjunction of evidentialism, mentalism, and the kind of awareness requirement necessary for a plausible version of explanationist evidentialism makes it impossible to become aware of the explanatory relevance of a proposition to one’s body of evidence.

Two of the three theses needed to get the regress going have already been introduced. The third is new.

(Evidentialism):

p is justified for S if and only if p fits all of S’s evidence e.

(Mentalism):

If e is evidence for S that p, e is a mental state of S’s.

(Awareness of Explanation):

If S has justification for believing that p as a result of p’s being the best explanation of e available to S, then S is aware (or disposed to become aware)²⁰ that p is the best explanation of e.

A few comments on these theses are in order. First, (Evidentialism) makes a very strong claim because it says one’s evidence and only one’s evidence contributes to justification.

Second, (Mentalism) is a one-way conditional. There is, however, a relevant biconditional that follows from (Mentalism) plus (Evidentialism): e is a justificationally-relevant mental state if and only if e is evidence.²¹ The conjunction of (Evidentialism) and (Mentalism) does not entail that all mental states are evidence, but does entail that all mental states that are relevant to justification are evidence.

Third, the relevant kind of awareness required in (Awareness of Explanation) need not be understood as propositional in nature, whether as a kind of judgment or belief that some e is best explained by p: the kind of awareness here need not be ‘noticing’ awareness (Sosa 2003, p. 120). Requiring justified propositional awareness of the explanatory connection between e and p leads to vicious regress. To believe that p on the basis of p’s best explaining e, one would have to justifiedly judge that p best explains e. To justifiedly judge that p best explains e in a way consistent with explanationist evidentialism, one would have to justifiedly judge that ‘p best explains e’ best explains $e^{\prime }.$ To justifiedly judge that, one would have to justifiedly judge that “p best explains e’ best explains $e^{\prime }$’ best explains $e^{{\prime \prime }}.$ Clearly, this will go on ad infinitum. The problem remains whether or not e is identical to the further bodies of evidence e$^{\prime },\,\mathrm{e}^{{\prime \prime }},$ and so on: per this requirement, one would still have to make justified judgments about the relevant explanatory connections. There is no reason to saddle explanationist evidentialists with an awareness requirement that by itself makes justification unobtainable. So, we’ll understand the relevant sort of awareness in (Awareness of Explanation) in a neutral way, and assume for the sake of argument that the awareness requirement is insufficient to generate problems.

(Evidentialism) and (Mentalism) are familiar. But why think (Awareness of Explanation) is an important component of explanationist evidentialist theories of justification? As we have seen, explanationists want to analyze justification in explanatory terms. Certain well-known thought experiments suggest that having justification for believing a proposition requires not only having some basis for belief, but also some awareness of the connection between the basis and the proposition supported.

Recall Laurence BonJour’s example of Norman (1985). Norman is a reliable clairvoyant. When Norman has a hunch that p as a result of his clairvoyant ability, it is usually the case that p. Norman has no reasons for or against believing that he is clairvoyant, or that anyone else is. As a result of his ability, Norman forms the belief that the President of the USA is now in New York. The belief is true. Intuitively, Norman’s belief is unjustified.

The Norman case obviously challenges a simple version of reliabilism according to which (put crudely) a belief is justified when it is the result of a cognitive process that usually produces true beliefs. But BonJour actually wanted to say more than just that. As he makes clear (1985, pp. 31, 42), he was mainly trying to show that one cannot be justified in believing that p on the basis of e unless one is also justified in believing that e makes p probable.

Ultimately, accepting that principle as a general requirement is disastrous for foundationalism.²² Showing that was early BonJour’s aim. The problem that arises is this. Take any belief that is allegedly foundationally justified. The belief will have some property in virtue of which it is foundationally justified. Call the property $\upvarphi .$ Descartes would have held that $\upvarphi$ is ‘being clearly and distinctly perceived’; Goldman (1976), that $\upvarphi$ is ‘being produced by a process that takes as inputs states other than beliefs and reliably outputs true beliefs.’ Now, according to BonJour’s requirement that one justifiably believes that p on the basis of e only if one justifiedly believes that e makes p probable, the fact that a belief has $\upvarphi$ is not sufficient for that belief to be justified. One must have another belief: that beliefs having $\upvarphi$ are probably true. As a result, having any justified belief requires having at least one other justified belief. As a result, BonJour’s principle entails that one cannot have a belief that is foundationally justified—regardless of what property one identifies with $\upvarphi$ and regardless of whether we analyze justification in an externalist way or an internalist way.

But suppose we weaken BonJour’s requirement. Instead of holding that one must justifiedly believe that one’s evidence make probable the proposition, we’ll require instead that one be aware (in some nondoxastic and nonpropositional sense) of the support relation holding between e and p. A requirement like that addresses the concern suggested by examples like Norman. By requiring awareness—even nondoxastic, nonpropositional awareness—of the relation holding between one’s evidence and the proposition it supports, we can avoid calling ‘justified’ beliefs that are formed out of the sort of ignorance that makes Norman’s clairvoyant beliefs seem defective.

Explanationists, of course, will require that the support relation be understood as an explanatory relation. To deny this would be to deny that the evidential support relation is fundamentally explanatory. So, it is plausible to think explanationists would hold both that some awareness of the connection between e and p is necessary for justification and that the connection is explanatory in nature. That is the requirement we have attempted to capture in (Awareness of Explanation).

Now that we have identified the three theses that lead to vicious regress, we ask: Does anyone accept the theses? Many are familiar with (Evidentialism) from the important work of Richard Feldman and Earl Conee in defense of the thesis. They initially defined Evidentialism this way:

• EJ Doxastic attitude D toward proposition p is epistemically justified for S at t if and only if having D toward p fits the evidence S has at t (Feldman and Conee 1985).

More recently, Feldman and Conee have endorsed (Mentalism): which, combined with (Evidentialism), yields this supervenience thesis:

• S The justificatory status of a person’s doxastic attitudes strongly supervenes on the person’s occurrent and dispositional mental states, events, and conditions (Feldman and Conee 2001).

Kevin McCain’s explanationist evidentialism is a version of (Evidentialism) (2013, 2014a). Ted Poston also defends a version of (Evidentialism) (2014, p. 92).

There are equally clear commitments to (Mentalism). McCain says “experiential states are evidence” (2013, p. 304). In his statement of necessary and sufficient conditions for well-founded justification, he writes, “[the body of evidence] E is a subset of S’s occurrent non-factive mental states and the non-factive mental states that she is disposed to bring to mind when reflecting on the question of p’s truth” (McCain 2014a, p. 118). It appears McCain is committed to (Mentalism). And as we said above, Feldman and Conee endorse S, which says justification supervenes on the mental. Poston’s explanationism is explicitly a kind of mentalism, so we can attribute (Mentalism) to him, too (2014, p. 92).

Finally, is anyone committed to (Awareness of Explanation)? While providing an example of how his explanationist evidentialism accounts for the justification of simple perceptual beliefs, McCain says the following:

The proposition $<$there is a red block$>$ is part of the best explanation of why S has an experience as of a red block that is available to S. Recall that in order for this proposition to be available to S, all that is required is for S to have the concepts required to understand $<$there is a red block$>$ and for S to be disposed to have a seeming that $<$there is a red block$>$ is part of the best answer to the explanatory why-question ‘why does S have e?’ on the basis of reflection alone. In this case, the relevant why-question can be expressed as: ‘why am I having this visual experience?’ or demonstratively as ‘why am I having this?’ [emphasis added] (McCain 2013, 306)

We italicized McCain’s awareness requirement. McCain requires S to have (or be disposed to have) a state of it seeming that p is the best answer to ‘Why e?’ in order for S to have justification for p on the basis of e.²³ The seeming-state requirement in McCain’s account is effectively a condition on justification that requires awareness of the explanatory connection between e and p: an instance of (Awareness of Explanation). But—and here is where the regress arises—seemings are justificationally-relevant mental states. Therefore, by (Evidentialism), they are evidence. According to (Evidentialism) no mental state can be relevant to justification without being evidence, and no piece of evidence can make one aware of the explanatory connection between a proposition and one’s evidence, as required by (Awareness of Explanation).

If S’s awareness that ‘p is the best explanation of e’ is a mental state relevant to S’s justification for p, then from (Evidentialism) that state of awareness is part of S’s total evidence. But now S’s evidence for p has grown: S’s evidence now includes S’s awareness of the explanatory connection between e and p. Call this new body of evidence e$^\prime .$ Well, according to (Awareness of Explanation), in order to get justification for p on the basis of the new body of evidence e$^\prime ,$ S needs to be aware of p’s being the best explanation of e$^\prime .$ This new state of awareness is, by (Evidentialism) now added to S’s yet larger body of evidence, e$^{{\prime \prime }}.$ But that won’t be sufficient either, ad infinitum. Every mental act of being aware of p being the best explanation of one’s current body of evidence gets added to one’s total evidence, which then expands to include the new evidence. Consequently, it is impossible to bridge the gap between one’s evidence for a proposition and the proposition. Instead, one just gets more and more evidence that can never be used to justify any proposition because one can never become aware of the connection between that body of evidence and a proposition.²⁴

The regress extends even to non-explanationist evidentialist views. Whatever the analysis of evidential fit—explanatory or other—if all justificationally-relevant mental states are evidence and the mental state of being aware of the fit relation is a justificationally-relevant mental state, then the regress follows. So, the solution to the regress will involve not only rejecting a distinctly explanationist view that accepts the three problematic theses, but rejecting any view that accepts the three theses.²⁵

Denying just one of the three theses is sufficient to avoid regress. Since explanationist evidentialism is built upon (Evidentialism) and (Mentalism), it would seem that explanationists might give up (Awareness of Explanation). We think, though, that (Awareness of Explanation) should be accepted by explanationists. (Awareness of Explanation) says justification requires awareness of the explanatory connection between a body of evidence and the proposition that explains the evidence. Cases like Norman the clairvoyant put pressure on those who would reject a requirement that subjects possessing justification for a proposition must be aware (in some way) of the connection between their evidence and the proposition supported by that evidence. Surely explanationists will understand the connection between evidence and proposition as an explanatory connection. Without a requirement like (Awareness of Explanation) there seems to be little reason to prefer explanationist evidentialism to a non-explanationist version of evidentialism.²⁶

A committed explanationist evidentialist may try to find a way to accept all three theses while avoiding regress. Consider this proposal.²⁷ In their analysis of properly based doxastic justification (they call it “well-founded belief”), Feldman and Conee (1985, p. 24) claim the following:

WF:    S’s doxastic attitude D at t toward proposition p is well-founded if and only if (i) having D toward p is justified for S at t and (ii) S has D toward p on the basis of some body of evidence e, such that (a) S has e as evidence at t, (b) having D toward p fits e, and (c) there is no more inclusive body of evidence e$^\prime$ had by S at t such that having D toward p does not fit e$^\prime .$

Because the regress argument targets the possibility of having adequate propositional justification to believe any proposition, we are of course disinclined to grant that condition (i) in WF is ever met if its being met depends on satisfying all three regress-generating theses that we have argued explanationist evidentialists must accept. Still, a clever explanationist evidentialist might try to modify WF for a satisfactory account of propositional justification that meets all three conditions while avoiding pernicious regress. That proposal might go like this. S has propositional justification for believing that p if and only if: p is the best explanation of e available to S, S has a larger body of evidence e$^\prime$ which includes S’s seeming that p is the best explanation of e, and there is no larger body of evidence e$^{{\prime \prime }}$ (which may be identical with e$^{\prime })$ had by S such that e$^{{\prime \prime }}$ fails to support p.

The basis of the objection seems to lie in the recognition that if the state of it seeming to S that p is the best explanation of e becomes part of e, then (in conjunction with the three theses) the regress indeed follows. So, the objector holds that the seeming-state that serves as the explanationist evidentialist (Awareness of Explanation) condition is treated as evidence [otherwise that state cannot be relevant to justification per (Evidentialism)], but not as evidence for p on the first level. Rather, the seeming is a part of a larger body of evidence that supports S’s justification for p by only by playing the negative role of not undermining the support p receives from e. While we think this is an interesting proposal, it does not avoid our regress in a way consistent with the three theses presented above. For even on this proposal it is impossible to become aware of the connection between a proposition one believes and all of one’s evidence that supports the proposition. That is because the seeming-state is part of the evidence for p. If that is so, then by (Awareness of Explanation) one must at least be capable of becoming aware of all of one’s evidence for p, including the seeming-state. However, the state of awareness in which that consists must itself be a seeming that is relevant to justification, and thus it itself is a part of one’s evidence that is relevant to the justification of p. This is the same regress as before.²⁸

We think the way to avoid the regress is to reject (Evidentialism) and recognize a distinction between evidential and non-evidential justifiers.²⁹ Poston, an evidentialist explanationist, rightly notes that (Evidentialism) rules out the possibility of non-evidential mental states that partially determine justificatory status (Poston 2014, p. 92). Our regress arises because the state of being aware of the explanatory connection between one’s evidence and the proposition it supports must be, according to (Evidentialism), a piece of evidence. Otherwise, by (Evidentialism), the state of awareness is not relevant to justification. Intuitively, there is a difference between the justificatory status of a subject’s belief that p when the subject (1) has evidence e that supports believing that p and is aware that e supports p versus when (2) the subject has e that supports p but the subject is unaware that e supports p. Intuitively, (1) is justified and (2) is not. But (Evidentialism) doesn’t allow us to make this distinction because (Evidentialism) requires us either to treat the subject’s awareness that e supports p as evidence, and so as part of e, or to reject that the subject’s awareness that e supports p is relevant to the subject’s justification. Neither of these consequences is plausible. The solution is to recognize the subject’s awareness of the connection between e and p as a non-evidential justifier. That awareness partly constitutes justification for the subject, but not because it is evidence supporting the proposition. Possession of evidence puts one in a position to obtain propositional justification; possession of evidence alone is not sufficient for propositional justification.

One alternative is accepting all three regress-generating theses and the skepticism they jointly entail. If the three theses are true it is impossible for anyone to possess epistemic justification for any proposition. This sort of very strong skepticism generates a self-defeat problem: if the regress argument shows that one cannot possess any justification for any proposition, then among the propositions one cannot possess justification for are the premises of the argument that support the skeptical conclusion, including those that express the requirements for justification. So, if you believe the skeptical conclusion, you do so unjustifiably. Perhaps this sort of skeptical argument reveals that our ordinary standards of justification simply can’t be met and that a skeptical theory of justification is correct.³⁰ We assume explanationist evidentialists will not be eager to accept that implication. In any case, these regresses defeat any non-skeptical view that accepts the conjunction of the regress-generating theses.

5 Conclusion

Explanationist evidentialism fails to give a sufficient condition for justification and further fails to provide a satisfactory nonskeptical theory of justification. The explanationist has threshold problems that not only threaten an explanationist analysis of the epistemic fit relation, but also threaten anyone who holds to the minimal explanationist thesis ME. Further, if the explanationist is interested in holding a satisfactory non-skeptical theory of justification, then the explanationist should admit that there can be non-evidential contributors to epistemic justification, and thereby reject evidentialism.