Evidentialism, Judgment, and Suspension: Meeting Sosa's Challenges

Abstract

Ernest Sosa has recently presented three challenges for evidentialism. The challenges concern what is required for epistemically justified judging and suspending of judgment. The aim of this article is to respond to these challenges on behalf of the evidentialist. Importantly, responding to Sosa's challenges requires giving substance to the idea of appreciating what one's evidence supports. This idea has been mentioned by prominent evidentialists but not given adequate development. Hence, this article marks a significant move forward in the understanding of evidentialism as well as a defense of the theory from Sosa's prima facie serious objections.

Ernest Sosa offers three important challenges for evidentialism in his book Epistemic Explanations (2021). These challenges revolve around concerns having to do with epistemically justified suspension of judgment¹ and epistemically justified judgment.² Essentially, Sosa focuses on what is required for one's conscious affirmation of p, or conscious refraining from affirming or denying p, to be justified. Sosa's first challenge is aimed at a particular form of evidentialism which says suspending judgment is justified whenever one's evidence fails to support believing p or disbelieving p.^{3, 4} He claims that this version of evidentialism cannot properly account for cases of justified suspension while one is figuring out what one's evidence supports. The second and third challenges are raised for evidentialism regardless of whether one accepts this particular take on suspending judgment. One challenge poses the problem of how one's suspension of judgment can be based on one's total evidence. The other challenge casts doubt on whether basing a doxastic attitude on the evidence it fits is sufficient for doxastic justification.⁵

Sosa's challenges are important for the evidentialist to consider not only because they seemingly threaten the viability of evidentialism but also because exploring them reveals aspects of evidentialism that provide deeper insight into the strength of the theory. That said, the plan for what follows is simple: provide an evidentialist response to Sosa's challenges. This process begins by first clarifying evidentialism (section 1). After this, each of Sosa's challenges is examined in detail and responses are developed (sections 2–4). Finally, the discussion wraps up with some concluding remarks (section 5).

1. Evidentialism

In broadest terms, evidentialism is the view that justification supervenes upon the evidence one has. Sosa captures the general idea of evidentialism pretty well when he describes it this way:

1. (a) At t, S has propositional justification for judging that p iff, at t, S's total body of <p>-relevant evidence (sufficiently) evidences <p>.

2. (b) At t, S is doxastically justified in judging that p iff, at t, S's total body of <p>-relevant evidence (sufficiently) evidences <p> and is a basis on which S judges (or judgmentally believes) that p (59).

Following Sosa, we can refer to this as “general evidentialism.” Of course, there are various more specific evidentialist theories because there are a number of ways of spelling out the components of general evidentialism. For instance, there are significant disagreements among evidentialists about what it takes for a body of evidence to “sufficiently evidence p,” or, in terms more common to evidentialists, for p to fit a given body of evidence. Likewise, there are different ways of understanding what it takes for a body of evidence to be a basis on which S judges that p. Fortunately, we will not have to settle these issues here. For the most part, we can simply work with general evidentialism.

Before moving on to Sosa's challenges, there is one additional aspect of evidentialism that needs to be briefly discussed. General evidentialism doesn't say anything about suspending judgment. However, as noted above, one plausible way of understanding this within an evidentialist framework, which Sosa challenges, is that suspension of judgment is justified when S's total body of <p>-relevant evidence fails to sufficiently evidence <p> and fails to sufficiently evidence <not-p>. Now that we have a handle on general evidentialism and a general evidentialist take on suspension of judgment, let's turn to Sosa's challenges.

2. Challenge 1: justified suspension while assessing evidence

2.1. Challenge

Sosa's first challenge is aimed at the evidentialist picture of suspension described at the end of the previous section. Here is what he says:

Problematic for that account is its treatment of suspension. A subject's proper withholding at t on whether p is defined as simply the failure of both <p> and <not-p> to fit that subject's total relevant evidence. However, suppose that <p> does fit your total relevant evidence although you have not yet had time to figure that out, and in any case you do not yet believe that p based on that possessed evidence. According to the definition, you are then at that time withholding improperly, but that seems questionable. Must you not withhold until you do properly believe? The fact that <p> fits your total relevant evidence seems insufficient to rule out your properly withholding, not while you are still deliberating. (59, footnote 10)

The evidentialist who says that justified suspension is a matter of one's evidence neither supporting believing p nor supporting disbelieving p appears to have a problem here. Suspension seems perfectly fitting in this case even though the person's evidence supports believing p.

2.2. Response

The evidentialist has an easy reply to this challenge from Sosa: the challenge rests on a mistaken assumption. The assumption is that because S does not believe that p and does not believe that not-p S must be suspending judgment about p. This isn't true though.⁶ There are various ways of being neutral with respect to a proposition without adopting any doxastic attitude toward that proposition. For instance, McGrath (2021: 469 emphasis in original) suggests that:

To suspend judgment on a question is to put off belief-forming judgment, that is, to omit it because one aims to judge it later (and not before) or when and only when certain conditions obtain (which one does not yet believe obtain).

Suspending judgment in McGrath's sense isn't a genuine doxastic attitude at all. In order to avoid confusion with the use of “suspend judgment” in the rest of the paper, let's refer to suspending judgment in McGrath's sense as “putting off judging.” In Sosa's case, it is plausible that S should put off judging whether p, and so adopt no doxastic attitude at all toward p. The evidentialist can readily admit that S isn't doing anything epistemically improper by putting off making a judgment even though believing that p is propositionally justified for S.⁷ As Stapleford and I (2020: 114) explain, evidentialism includes no general commitment to adopt doxastic attitudes. Rather, evidentialism commits one to a conditional claim: “If one adopts any doxastic attitude towards p, one (epistemically) ought to adopt the attitude that fits one's evidence.” S's evidence makes believing p the only justified doxastic attitude for her to adopt if she adopts an attitude toward p at all. In other words, believing p is propositionally justified for S, but as Conee and Feldman (2018: 79) explain “an attitude's being epistemically justified does not imply having any sort of duty or obligation to take the attitude.” Since S is still considering her evidence, it is plausible that S is putting off making a judgment and so hasn't adopted a doxastic attitude at all. Consequently, evidentialists can agree with Sosa that believing that p is propositionally justified for S in this case and that there is nothing improper about her not believing p while she continues to deliberate. Hence, there's no problem for the evidentialist here.⁸

Perhaps one would want to insist that in this case, S should do something more than put off making a judgment though. In other words, one may object to evidentialism on the grounds that S should have “an intermediate state of confidence concerning whether p, i.e., having states of confidence for p and for not-p such that one's state of confidence for p is neither strong enough for belief that p nor low enough for disbelief” (McGrath 2021: 471). Following McGrath's terminology, the idea here is that S's proper withholding in this situation amounts to being “agnostic” with respect to p. If this is correct, then Sosa's case poses a genuine problem for evidentialism because agnosticism toward p amounts to adopting a doxastic attitude toward p. As a result, the response described in the previous paragraph won't work.

Here again, the evidentialist has an easy response though. There's no reason to accept that in Sosa's case, S should be agnostic rather than put off making a judgment concerning p. In fact, the case seems to be exactly the sort of situation where one should put off forming a judgment until later when one has had a chance to consider the evidence. Consequently, the intuitively correct response to this case is one that seems perfectly compatible with evidentialism.

Unfortunately, Sosa's two remaining challenges aren't as easily set aside as this one. As we will see, they can be met but doing so requires developing aspects of the evidentialist picture that have been left largely ignored until now.

3. Challenge 2: properly based suspension

3.1. Challenge

The second challenge that Sosa presents for evidentialists concerns accounting for how the doxastic attitude of suspending judgment can be based on one's evidence:

It is hard to see how suspending judgment on a given question <p?> could ever be based on one's total evidence. Judging can plausibly be based on evidence if the evidence speaks sufficiently in favor of the content of your affirmation. But it is unclear how suspending could be based on evidence in virtue of some relation between the evidence and the content of your suspending. (59–60)

To better appreciate Sosa's challenge, let's consider two situations in which evidentialists will say that suspending judgment about p is justified for S. First, S might be justified in suspending judgment concerning p because she simply has no evidence relevant to whether p is true. Second, S may be justified in suspending judgment on p because, while she has evidence that is relevant to p's truth, that evidence is counterbalanced – it equally supports believing p and disbelieving p. Sosa discusses the second possibility and rightly notes that such evidence may be fairly complex, and in many cases, it is difficult to see how one's suspending judgment could be based on that evidence. Although Sosa's challenge can be made by considering this sort of case, it seems most forceful when considering the first possibility – S has no <p>-relevant evidence at all. One might worry that a doxastic attitude cannot be based on a lack of evidence – especially if we, like Sosa, take the basing relation to be some sort of causal relation.⁹

3.2. Response

One initially appealing response to Sosa on this point would be to insist that suspending judgment about p simply amounts to having a belief that “one cannot yet tell whether or not p” (Raleigh 2021: 2449). If suspension is itself a belief, then so long as the evidentialist can give an account of how beliefs are based on one's evidence, there's no problem here.

While this response may be on the right track, there are two difficulties facing it. The first is simply that evidentialists may not want to accept that all cases of suspension of judgment involve a higher-order belief of this sort. The second problem is that we are still faced with the challenge of explaining how this higher-order belief is justified – a challenge that may be particularly difficult in cases where S has no <p>-relevant evidence. In other words, this response may itself face a version of the very challenge it is thought to put to rest.

Fortunately, evidentialists have a way of responding that can be combined with the suspension as belief view or taken on its own without committing to that particular picture of suspension. The key to this response lies in an aspect of evidentialism that has been mentioned but inadequately explored. This underexplored element of evidentialism is what we might call “appreciation.” It is the relation that one needs to bear to one's evidence in order for the evidence to, as Sosa would say, do its evidencing. For instance, Feldman (2000: 681–82) says, “only those propositions whose connection to the evidence the person apprehends are actually supported by his evidence.” Similarly, Conee (2020: 72) claims, “[evidential support] requires the person to have some mental access to the existence of the indication relation between the evidence and the proposition, perhaps by having some potential or actual awareness of that relation.”

If evidentialists follow Conee and Feldman in requiring appreciation for justification, a response to Sosa's challenge emerges. The evidentialist can insist that justified suspension of judgment concerning whether p is not directly based on S's <p>-relevant evidence. Rather, it is based on S's appreciation of what the <p>-relevant evidence supports or fails to support. So, in the case that Sosa has in mind, whether it is a case where S has no <p>-relevant evidence or S has counterbalanced <p>-relevant evidence, it isn't that evidence itself that needs to be the basis for S's suspension. Instead, S's suspension is justified so long as it is based upon her appreciation of either her lack of <p>-relevant evidence or her <p>-relevant evidence being counterbalanced. In both cases, this would amount to S's appreciating that her total evidence fails to support p and fails to support ~p. Plausibly, suspension, like other doxastic attitudes, can be based on states of awareness like appreciation whether or not the evidentialist accepts that suspension is itself belief or instead a distinct doxastic attitude. Hence, the evidentialist can meet Sosa's second challenge by appealing to an underdiscussed aspect of evidentialism that has been a part of the picture for quite some time.

3.3. Three questions

Appreciation offers evidentialists the promise of a response to Sosa's second challenge. However, it prompts three important questions. Answering these questions are vital to making the case that this response is successful and helps make clear how evidentialism can be most plausibly construed.

Question 1: What sort of mental access is required for appreciation?

Recall, Conee's claim mentioned above that appreciation involves “some mental access to the existence of the indication relation between the evidence and the proposition, perhaps by having some potential or actual awareness of that relation.” One plausible way of fleshing out Conee's suggestion is in terms of appearances.¹⁰ Appearances are “experiences that have propositional content, mind-to-world direction of fit, and forcefulness” (McCain and Moretti 2021: 62). More precisely, appearances have content that can be expressed with that-clauses, represent the world in ways that can be accurate or inaccurate, and “have the feel of truth, the feel of a state whose content reveals how things really are” (Tolhurst 1998: 298–99). Evidentialists can reasonably maintain that appreciating what one's evidence supports with respect to p amounts to having an appearance that belief/disbelief/suspension concerning p is supported by one's total evidence.

Question 2: Can misleading appreciation supply justification?

Assuming that appreciation plays a necessary role in the justification of one's doxastic attitudes, it is natural to wonder whether the appreciated connection between one's evidence and a given proposition must be genuine. In other words, can misleading appreciation yield justification? In short, the answer is “no.” If S's evidence doesn't actually support believing that p, then it appearing to S that that evidence supports p doesn't make it so that the evidence supports believing p.¹¹

Of course, one might object that there seem to be cases where misleading appreciation does yield evidential support. For instance, Huemer (2013: 339) argues for exactly this point with his case of the unfortunate mathematician:

S has just gone through what seems to be a genuine mathematical proof of p, starting from e as the sole premise. S is adequately justified in accepting e, he has checked the argument over carefully, and it is, as far as he can ascertain, a sound and rigorous proof of p. Nevertheless, it happens that there is a subtle error in the argument, as a result of which the argument is neither valid nor cogent. What ought S to think? In the internalist sense – that is, in terms of what makes sense from the subject's perspective – it seems to me that the unfortunate mathematician should believe p. Indeed, he would be irrational not to believe p at this point… But his evidence for p does not logically support p. So it is possible to be justified in believing p on the basis of evidence that does not objectively support it.

If Huemer is correct, then it seems that misleading appreciation can make it so that a body of evidence supports p even when objectively the evidence fails to support p. Is Huemer correct though? It seems not. In Huemer's case insofar as it is plausible that the mathematician's belief is justified, it is because of his background evidence concerning his mathematical skills plus his evidence that he checked the argument for p over carefully – the premises of the purported “proof” don't provide justification for believing p. Hence, the mathematician's misleading appreciation isn't sufficient for making the evidence supplied by the premises of the purported “proof” support believing that p.¹²

Question 3: Where does appreciation fit into the evidentialist picture?

Conee and Feldman insist that appreciation is itself part of one's evidence as it is necessary for propositional justification. In other words, they maintain that even if one's evidence logically entails p in cases where one has that evidence but doesn't appreciate this connection between the evidence and p believing p is not propositionally justified. This idea appears plausible when one considers cases where S simply cannot see that p is logically entailed by her evidence. For instance, it could be that p is too complex for S to even grasp, or it could be that S can grasp p but the derivation of p from her evidence is beyond her (or perhaps any human's) abilities. In such a case, Conee and Feldman contend that S does not have propositional justification for believing that p. Perhaps an even stronger case for this position is one in which p is logically entailed by S's evidence, but S cannot appreciate this fact, and further, S has been given strong (but not entailing) evidence that p is false. In this sort of case, many would hold that S is not justified in believing p (or even assigning it a credence above 0.5) even though her evidence logically entails that it is true. In light of this, evidentialists may follow Conee and Feldman in holding that appreciation is required for propositional justification.¹³

While including appreciation as required for propositional justification is a reasonable option, evidentialists do not have to go this route. Some might think that there are problems with making appreciation necessary for propositional justification.¹⁴ Consequently, another option for evidentialists would be to require appreciation for the transition from propositional justification to doxastic justification. In other words, the evidentialist could claim that appreciation is needed for basing one's doxastic attitude upon the evidence. Locating appreciation at the level of the basing relation allows the evidentialist to make the same response to Sosa's second challenge as requiring appreciation for propositional justification.

There is a cost for the evidentialist who insists that appreciation is a feature of proper basing, however. It requires accepting that basing is more complex than we described above – simply bearing the right causal connection to the first-order evidence doesn't suffice when it comes to doxastically justified suspension of judgment. Considerations of simplicity suggest that if the right causal connection doesn't suffice for suspension, it doesn't suffice for belief/disbelief either. Hence, it seems that evidentialists who go this route should say that one's doxastic attitude toward p is doxastically justified just in case that attitude is propositionally justified and one properly bases the attitude on one's appreciation of the evidence supporting that attitude toward p.¹⁵ Importantly, this doesn't amount to S's doxastic attitude not being based on evidence (even if we don't follow Conee and Feldman in thinking that appreciation is necessary for propositional justification). The appearance that constitutes S's appreciation of what the <p>-relevant evidence supports or that S has no <p>-relevant evidence is itself higher-order evidence. Thus, in a case where S has no <p>-relevant evidence, justifiedly suspending judgment would require basing that attitude on S's higher-order evidence, specifically, her appreciation that her evidence neither supports believing p nor believing ~p. The added complexity here may be a reason for evidentialists to opt for Conee and Feldman's approach to appreciation. Be that as it may, the issue does not need to be settled here. What matters for present purposes is that evidentialists have viable options when it comes to incorporating appreciation into the evidentialist picture, and both options offer a plausible response to Sosa's challenge.

4. Challenge 3: evidential fit and proper basing isn't enough

4.1. Challenge

The third challenge from Sosa arises from his contention that whether one justifiedly suspends judgment or believes isn't simply a matter of whether that doxastic attitude is based on adequately supporting evidence. Instead, Sosa argues that one must also base the attitude on the fact that one's evidence satisfies two further conditions:

• first, that E be your total body of relevant evidence, that there be no other relevant evidence in your possession;

• second, that E be epistemically sufficient for judgment rather than suspension, that the balance of the contained evidence sufficiently favor judging as you do (60).

If one fails to base one's doxastic attitude on these additional facts then, successfully adopting the right attitude is “by luck in some measure, so that your success is not fully creditable to you. Too easily, perhaps, might you have affirmed despite failing to meet the required conditions” (64).

4.2. Response

Let's grant for the sake of argument that Sosa is correct that doxastic justification requires being creditworthy. Do the two conditions that Sosa presents pose a problem for evidentialism? There seem to be only two ways in which the conditions could pose any sort of challenge for the evidentialist. The first is if the two conditions are incompatible with evidentialism. The second is simply by challenging the evidentialist to actually incorporate the conditions into evidentialism.

The first way of challenging evidentialism can be quickly set aside. There is no conflict between Sosa's proposed conditions and evidentialism. As Conee and Feldman (2004: 101) explain, the “bedrock” of evidentialism is a supervenience claim about propositional justification:

(ES) The epistemic justification of anyone's doxastic attitude toward any proposition at any time strongly supervenes on the evidence that the person has at that time.

Additional requirements for doxastic justification don't pose a problem for this fundamental principle of propositional justification. Conee and Feldman themselves allow that more than simply fit with one's evidence is required for doxastic justification, which they term “well-foundedness.” Specifically, they contend that doxastic justification requires in addition to fit that one form the propositionally justified doxastic attitude on the basis of the evidence (Conee and Feldman 2004: 93). It is consistent with ES, and the broader evidentialist picture of doxastic justification, that satisfying the additional conditions that Sosa claims must be satisfied for doxastic justification are part of what it means to hold a doxastic attitude on the basis of one's evidence. Thus, there is no challenge for evidentialism when it comes to being consistent with Sosa's conditions.

When it comes to the second way that Sosa's conditions might be thought to be challenging, the groundwork for meeting this challenge has already been laid in the previous section. Recall the sort of appreciation we mentioned above as required for justifiedly adopting a doxastic attitude: appreciation of what one's evidence supports with respect to p (i.e., what the <p>-relevant evidence supports or that S has no <p>-relevant evidence).¹⁶ Requiring appreciation as we have described it (an appearance that one's evidence supports a particular doxastic attitude) appears to satisfy Sosa's first condition as long as we understand the evidence in question to be one's total evidence. In other words, doxastic justification requires basing the doxastic attitude in question on appreciation of what one's total evidence supports with respect to p. This doesn't necessarily satisfy Sosa's second condition though. After all, it could be that believing, say, appears to be supported one's total evidence without it thereby appearing that the total evidence sufficiently favors judging rather than suspending.¹⁷ Nevertheless, there is an easy fix to this:

S appreciates doxastic attitude D's fitting S's evidence if and only if it appears to S that D is sufficiently supported by S's total evidence.

The picture of doxastic justification that emerges satisfies both of Sosa's conditions without conflicting with the letter or the spirit of evidentialism:

S's doxastic attitude, D, toward p is doxastically justified just in case:

1. 1) D is propositionally justified, and

2. 2) S properly bases D on the appearance that D is sufficiently supported by S's total evidence.

At this point, one may be concerned that answering Sosa's third challenge as we have here will amount to overintellectualizing justification. The thought is that it is one thing for an unreflective agent to have evidence that renders p likely or unlikely to be true, but it is quite another for the agent to appreciate what her evidence's connection to p is. This is true enough. However, the evidentialist has at least two things to say in reply. The first is simply that in the context of making his challenges to evidentialism, Sosa is concerned with reflective judgment. Presumably, any agents capable of the sort of judgments that Sosa is appealing to in order to challenge evidentialism would be able to have the sort of appreciation described above. The second thing to say is that it isn't hard to have such appreciation. Although for simplicity, the account of appreciation provided here is couched in terms of what is supported by one's evidence, the appearance in question does not need to include concepts such as EVIDENCE, SUFFICIENCY, etc. The required appreciation consists of the sort of appearance that plausibly even young children can have. For example, young children can have appearances that they know that p is true, or that they don't have a clue about p. These are sufficient for satisfying the required appreciation. So, there doesn't appear to be an overintellectualization problem here. Thus, evidentialists can meet Sosa's third challenge too.

5. Concluding remarks

We have explored three challenges Sosa presents for evidentialism. The first challenge is easily overcome by distinguishing between suspending judgment and other states where one doesn't believe or disbelieve. The second and third challenges can be met but doing so requires the evidentialist to add certain appreciation requirements for justification. When all is said and done, Sosa's challenges do not undermine evidentialism. Instead, they push evidentialists to clarify and refine important ideas. Arguably, the result of addressing Sosa's challenges is a more developed and more plausible evidentialism.¹⁸