*Based on the Conee & Feldman reading.

Evidentialism refers to a particular concept of epistemic justification according to which a belief's justification is a matter of the nature and degree of the evidence one has for the belief. More specifically, Conee and Feldman spell out evidentialism under the following principle:

[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.

Conee and Feldman understand evidence to include both propositional evidence and experiential evidence. Hence, the evidentialism they advocate is a broad sort of evidentialism.

Justification is an evaluative term. It indicates the right sort of cognitive response to evidence. There are three sorts of cognitive responses to evidence: believing p, believing not-p, and withholding belief (i.e., believing neither p nor not-p). Each stands in a particular relation to evidence.

If p fits our evidence, then we ought to believe p.

If p does not fit our evidence, then we ought to believe not-p.

If our evidence is equally balanced, then we ought neither to believe p nor believe not-p.

One objection to [EJ] is that it implies that humans have direct voluntary control over their beliefs. This is the so-called doxastic voluntarism thesis. According to this view, a cognitive attitude (belief, disbelief, or withholding of belief) is justified only if the cognitive attitude is within our direct voluntary control. However, there is good reason to suppose that this thesis is false, and hence if [EJ] entails it, [EJ] must also be false. Conee and Feldman argue that evidentialism need not have this implication at all. They suggest, like many, that the voluntary control thesis is false. But unlike some philosophers, they do not think evidentialism is committed to doxastic voluntarism. They argue that some involuntarily adopted cognitive attitudes are justified and that some unjustified cognitive attitudes are involuntary. Someone who spontaneously believes that the lights are on in his apartment, as the result of the evidence of his senses, is justified in this belief, even though it is not a belief chosen at will. A paranoid person is unjustified in his belief that someone is following him if he has no evidence for this, even though the belief is formed uncontrollably as the result of his paranoia.

Conee and Feldman think that this objection to evidentialism emerges as the result of a confusion between evaluation of belief and evaluation of the believer. "Blame" and "praise" are terms that indicate an assessment of a person. And it is widely held that we can blame or praise people only if we praise or blame them for something under their voluntary control. You praise a person's action of helping the poor only if the person freely chose to do this. And we certainly do not think that someone should be blamed for acting rudely or offensively if the action was the result of some cognitive disorder over which the person had no control. (Of course, we may certainly evaluate people non-voluntary actions in other ways. We may, for instance, assess whether the behavior is defective or not normal in some way). Conee and Feldman think that justification has little to do with assessments of people, but rather it involves an assessment of a belief held by a person. Asserting that John is justified in believing that the lights are on does not entail that we should praise John for anything. Similarly, saying that the paranoid man is unjustified in his belief that someone is following him does not entail that the man should be blamed for holding this belief.

According to Conee and Feldman, a theory of epistemic justification sets a standard for the evaluation of beliefs. Epistemic standards establish requirements or obligations. To say that a belief has satisfied such standards is to say that it is justified. If it does not meet such requirements, then it is unjustified. For the Conee-Feldman type evidentialist the standard is "fit with evidence." "One ought to have the doxastic attitude that fits one's evidence. We think that being epistemically obligatory is equivalent to being epistemically justified" (p. 172). But there are different ways of spelling out this normative or deontological aspect to justification. Some of them differ from Conee and Feldman.

A. Chisholm's view:

According to Roderick Chisholm, our intellectual duty is to try to bring it about that for every proposition we consider one believes it only if it true. The technical formulation is as follows:

[CJ] Doxastic attitude D toward proposition p is justified for a person S at time t if and only if S considers p at t and S's having D toward p at t would result from S's trying his best to bring it about that S believes p at t iff p is true.

"Trying one's best" may be taken two ways. First, it may be taken as "trying in such a way that one has the best result," which in this case would be actually acquiring a true beliefs and avoiding false ones. But one can easily believe a truth one considers and not believe a false proposition one considers without any relation to evidence. For Conee and Feldman we are not justified because we believe what happens to be true. This sort of success is not justification. Secondly, though, we might read Chisholm as implying that "trying one's best" is simply a matter of believing in accordance with one's evidence. But even this formulation differs from the Conee and Feldman thesis. Chisholm only says that one does one's best to do this. Clearly, though, one might do one's best to see to it that one believes only those propositions that fit one's evidence without actually succeeding in this. For Conee and Feldman what matters is not trying your best but your belief actually fitting your evidence.

B. Kornblith's View

According to Hilary Kornblith, the relevant obligation is "to seek truth and gather evidence in a responsible way" (p. 173). But Conee and Feldman think that this view is neither necessary nor sufficient for justification, given the evidentialist view. First, one might have been very irresponsible in gathering evidence, but the evidence might nonetheless actually fit one's belief. According to Kornblith, one could not be justified in this instance, whereas for Conee and Feldman the person would be justified. Secondly, a person might have been very responsible in seeking truth and gather evidence. But suppose the person believes p at the end of the investigation, even though the belief does not in fact fit the evidence, even if the person supposes that it does. Conee and Feldman argue that in such a situation the person is unjustified, whereas Kornblith would say the person is justified.

What we have here, then, are different ways of spelling out what our epistemic obligations are. Chisholm says that it is a matter of doing one's best to see that one believes what is true and does not believe what is false. Kornblith says that the obligation is to engage in truth seeking and evidence gathering in a responsible way. But Conee and Feldman argue that the obligation is to have the appropriate doxastic attitude given your evidence. We ought only to believe what fits our evidence.

Critical Note:

There is a reason, of course, why Chisholm and Kornblith hold the views they do. Each recognizes that people have obligations and they do only if they can voluntarily engage in that to which they have an obligation. This creates a bit of a problem for Conee and Feldman who reject the notion that justification involves an assessment of a person but only a belief, and thus is not committed to a doxastic voluntarism thesis. But it does not seem reasonable to suppose that we could have obligations toward doing x if our doing x was not voluntary. Of course, a person who met his epistemic obligations should not be praised, at least not on the assumption that people should not be praised for simply having done their duty. Nevertheless, we are inclined to attribute blame to people who fail to meet their duty. The case for doxastic voluntarism follows just from the notion of obligation and blame, not praise. It looks like Conee and Feldman's attempt to make evidential justification normative cannot avoid the implication that people are to blame when their beliefs fail to be justified. But this does imply the voluntary control of belief thesis, at least if it is true that we can only be blamed for voluntary actions.

Conee and Feldman do make it clear, however, that though justified cognitive attitudes are obligatory in some sense, they are not the same moral and prudential obligations. Consider an example of the latter. A woman who suspects that her husband has been unfaithful may have good evidence for this belief and thus be epistemically justified in holding it, but she might be better off in other respects for not believing that her husband is unfaithful (e.g., if this were to lead to the end of a marriage that she desires to sustain). Practical ends are sometimes best served by not believing what fits our evidence. Why? Because beliefs that do not fit our evidence can be useful for many things and beliefs that fit our evidence can undermine the obtaining of various goals. And the same difference emerges with respect to moral obligations and epistemic justification.

Conee and Feldman refine their evidentialist thesis in the light of an additional criticism. The criticism aims to show that having evidence for a belief or a belief's fitting with one's evidence is not sufficient for justification. Take the case of paranoid Jack. Suppose that he believes that his neighbor is spying on him. Suppose further that this belief fits with Jack's evidence for this belief (e.g., having seen him look over his fence during the day and evening, running into him at the store). But it is possible that Jack's belief that his neighbor is spying on him might not be based on the evidence that supports it. In other words, a person might have a belief that fits his evidence but not hold the belief because of the evidence. Many argue that in such a situation it is incorrect to say that the person is justified in holding the belief.

This objection stems from the source relevant view of justification advocated by many philosophers. The source relevant view of justification maintains that a necessary factor determining justification is how the belief was formed or how it is sustained, i.e., the psychological source or origin of the belief. A belief's fitting one's evidence, even if necessary to justification, would not be sufficient for justification since this is logically compatible with a person holding the belief on the basis of something other than the evidence that supports it.

To address this issue, Conee and Feldman have recourse to a distinction between justification and well-foundedness, two different epistemic assessments of belief. Justification, they contend, is solely an issue of whether a belief fits our evidence. It has nothing to say about whether the evidence is the reason why we hold the belief. This is not to say that the idea of the cause or psychological source of belief is irrelevant. Conee and Feldman argue in addition to justification, a belief that is held on the basis of the evidence with which it fits is a fell-founded belief. Hence, they view the source relevant constraint as something that is indicative of an episemic desideratum distinct from justification. So some justified beliefs may fail to be well-founded, but no well-founded belief will fail to be justified.

Question for further thought: Does knowledge require well-foundedness?

For further discussion on evidentialism, see Sudduth, Evidentialism and Deontology.