Imagine visiting a museum and coming upon a great hall that contains a full-size ship on display. On an informational placard, you read that the ship used to sail the seas in the nineteenth century, but that it is still sea-worthy today. You might then continue your tour with a belief that corresponds to your evidence: the ship is sea-worthy. Most would intuitively consider this belief justified, as you got it from a source that is probably trustworthy, and you do not have a reason to further investigate the ship to gather more evidence. However, if we consider a shipowner forming a belief about the ship, the situation is different. Perhaps it is time to send the ship out to sea once again, for a long voyage with many passengers. If we imagine the shipowner to have roughly the same level of evidence as the museum visitor for the belief that the ship is sea-worthy, it seems that our intuitions about the shipowner’s epistemic status differ from those about the museum visitor. One might be inclined to expect the shipowner to gather more evidence, to strengthen their epistemic status concerning the belief in question. This can be explained by the fact that the stakes for the shipowner are much higher: what they believe will guide how they act, and depending on how they act, many lives could be endangered if the ship is not seaworthy.

This situational sketch illustrates intuitions about knowledge and justification which have posed a challenge for epistemological purism, the standard position in academic epistemology. Purism is characterized by exclusively considering truth-relevant factors of beliefs. 1 This view is challenged by pragmatic encroachment on knowledge: whether one knows something can depend (in part) on what is at stake in whether the proposition is true.2  Epistemologists who adhere to this, who allow pragmatic factors to be relevant to knowledge, are called pragmatists. At the end of the day, the standard purist and pragmatist approaches to knowledge differ fundamentally on the matter of values. Purists value epistemic excellence. Pragmatists hold that beliefs are valuable for being a reason for action or for furthering one’s interests. Though purists have trouble explaining certain cases like the ship-example, pragmatists run into a plethora of different problems when extending their pragmatic encroachment to cover greater areas of epistemology. 3 I am interested in a stable, yet pragmatically sensitive epistemic theory which does not face these problems by taking a middle ground. As such, in this paper, I will develop a soft pragmatist theory which does justice to both kinds of value by relying heavily on an evidentialist framework while allowing a spot for pragmatic value. A soft pragmatist position usually combines purist norms with a pragmatist justification, as they argue that basing one’s beliefs on evidence is the best way to further one’s interests.4 Along these lines, I will formulate a proposal to positively answer the following question:

Can the epistemic value of evidentialist norms be warranted by a pragmatic justification of these norms?

To achieve this, I will defend a soft pragmatic position to include a form of justification encroachment of epistemic standards, which accepts epistemic value in reasonable beliefs. This means that I will argue that practical factors are relevant in determining whether a subject is generally rational in forming or suspending a belief. As my theoretical starting point, I shall explain Richard Feldman’s evidentialist norms and his
theory of epistemic value as reasonable beliefs.

Subsequently, I will argue how a pragmatic justification of these norms can fit this framework, considering the level of epistemic risk and the value of suspension of belief. Then, I will further develop this proposal by advancing a couple of detailed ways this soft pragmatic approach can be understood, including some foreseen problems and solutions. Finally, I will evaluate the final status of my version of soft pragmatism and its alternatives and consider their strengths and weaknesses.

What is epistemic value?

In part II of ‘The Ethics of Belief’, Richard Feldman defends an evidentialist variation on W.K. Clifford’s claim: “It is wrong, always, everywhere, and for anyone to believe anything upon insufficient evidence.5 According to Feldman’s evidentialism a ‘believer’ achieves epistemic excellence if they form many reasonable beliefs. Furthermore, it is rational to follow one’s evidence to form a belief. To elucidate Feldman’s position, I will use a model that Veli Mitova presented in ‘Why W.K. Clifford Was A Closet Pragmatist’ as she distinguished two normative questions about belief:

1) The norms-question: How ought I to form my beliefs?

2) The justification-question: Why should I accept the epistemic procedures recommended by (1)? 6

Following these questions, Feldman provides an evidentialist answer to the norms-question: I ought to form beliefs in a reasonable way, which is to say based on my evidence. Consequently, his answer to the justification-question, as to why one ought to form beliefs in that way, is: because following one’s evidence leads to something of epistemic value.

In principle, epistemic value is the goal we hope to achieve through our epistemic norms. A person can reach epistemic excellence by being a ‘good’ believer. According to Feldman, assigning epistemic value to true belief or knowledge encounters substantial problems for the fallibilist, evidentialist thesis.7 The goal of believing all and only truths is not an attainable or desirable goal. Likewise, following one’s evidence is not always the best way to get at the truth. Hence, this is not a good candidate for being our epistemic goal. Additionally, identifying knowledge with epistemic value runs into the same trouble when considering someone who has strong evidence for an untrue proposition, as following one’s evidence will not lead to knowledge in that case. As such, the sought-after evidentialist epistemic value is best designated by the idea that one must strive for reasonable beliefs. Someone who gains many irrational true beliefs is not doing well, epistemically. A person who, on the other hand, forms many rational but false beliefs is doing well epistemically. This also means that if the evidence is strong enough for a justified belief which turns out to be true, one will gain knowledge where one can.8 In other cases, one is left with valuable reasonable beliefs. As such, epistemic value is not the trueness of a belief or even the level of justification, but a different property of the belief, being reasonableness. This is a perfect way of achieving epistemic value, according to Feldman.9

So, what are reasonable beliefs? And why are they valuable? The epistemic norm can be understood as the rule for believing as one ought to. One maximizes epistemic value, or believes as they ought to, by adopting a rational attitude toward a proposition. Adopting a rational attitude is simply understood as following one’s evidence in a rational way. This means that, depending on the strength or ‘direction’ of the evidence, one will believe the proposition, disbelieve the proposition, or suspend judgment about the proposition. 10 The result is guided by purist properties of the evidence, meaning that the strength of the evidence does not rely on pragmatic factors. Following one’s evidence is the best way to maximize epistemic value, because someone who, against all evidence in, believes a true proposition, is not forming reasonable beliefs, although they might have true belief. Reasonable beliefs are valuable because they are widely attainable, and they will guide one to knowledge whenever it is available. Apart from this, one will be doing the best they possibly can, epistemically, in the light of misleading evidence and irrational convictions.

 Can pragmatic justification be reasonable?

As one might see, evidentialism of this kind will have trouble explaining the ship-example, because the strength of the evidence is similar in both situations, which would mean that their epistemic status should be the same, too. I would like to diverge from Feldman’s framework with a scheme in which pragmatic factors are relevant to determining if a person is rational in adopting a certain cognitive attitude to a proposition. This means that pragmatic factors are relevant to which beliefs are reasonable by being relevant to the strength of the evidence relative to situational factors. For epistemic norms to be warranted, they need to provide epistemic value. The question that I believe needs to be answered now is: can a soft pragmatist theory of knowledge provide a scheme where, despite the pragmatist justification, epistemic value is maximized? I will argue that pragmatically justified epistemic norms can still provide epistemic value, as pragmatic encroachment can exist in the evidentialist framework without affecting its fundamental value.

The ‘reasonable’ quality of a belief depends on how one gained that belief, or more specifically if one adopted a rational attitude towards the supporting evidence. However, one problem of evidentialism is that it is not clear what ‘evidential support’ exactly entails: multiple propositions may follow from evidence, the connection between the evidence and the proposition may be distant, or sometimes one has some modest amount of evidence supporting a proposition according to which one may believe or suspend judgment. 11 The question is how one should follow the evidence to be rational. This means we need to determine the strength of the evidence that is needed to gain reasonable beliefs. Feldman tends more towards forming a belief rather than suspending judgment when confronted with epistemic risk. Nonetheless, he remains ambiguous about the determination of this epistemic standard: “Rather, people simply have varying attitudes toward this.”12 ‘This’ refers to the level of epistemic risk that is allowed for the belief to reasonably follow from the evidence.

To provide a standard to which to measure the allowed risk, I propose that in the context of inquiry pragmatic factors can help ascertain the level of epistemic risk one should take. When epistemic risk meets pragmatic risk, one takes a rational attitude to the evidence if one tries to avoid risky beliefs. Pragmatic risk should be understood as how much there is at stake based on whether the belief is true. I would argue that it is not less than reasonable to be more careful with beliefs in risky situations. The epistemic norm in these scenarios can be purist in nature: one only ought to form beliefs when one has more than a modest amount of evidence. This is standard evidentialism, but with higher epistemic standards. The justification of these norms is pragmatic: this is good because the stakes are high, and it furthers my interests.

What will result from this scheme, is that there will be cases in which the epistemic standards will be raised, and as a result, after one rationally considers one’s evidence, one should suspend their judgment more often. If you suspend judgment more often, you gain fewer (justified and true) beliefs because you form fewer beliefs in general, meaning you will not get knowledge whenever it is available. Still, it seems like a more reasonable way of forming beliefs. Feldman wants to maximize value by removing high standard, as I do, but I raise the standards again. So, why do I end up with a similar amount of epistemic value if I hold on to largely the same framework? One may argue that I end up with less reasonable beliefs and therefore lose epistemic value, which means that I did not warrant the value of the epistemic norms. I object by proposing that the suspense of judgment is also an epistemological position which may have epistemic value. It may be valuable because one may arrive at this cognitive attitude by taking a rational attitude to the evidence. On the other hand, if one would reject the inherent value of reasonable suspense of judgment because it is not a belief, a response may be to assign it instrumental value: the higher standards will incite one to search for more evidence. This results in a stronger reasonable belief when it may be formed, which has a higher chance of leading to knowledge. The epistemic value is postponed, but the final position carries more weight.

Substantializing the proposal

One interpretation of these considerations can be developed as a contextualist theory of epistemic standards. The concurrence of pragmatism and contextualism is commonplace, though not in every case about pragmatic encroachment on epistemic standards. 13 Here, I will initiate such an endeavor. It may be imagined as such: in low stakes contexts, one adheres to the standard evidentialist norms as proposed by Feldman. In a situational context of high pragmatic risk, on the other hand, one should adhere to the following epistemic standards:

1. The norms-question: Only form beliefs that have low epistemic risk.

2. The justification-question: This is good because it furthers my interests.

This may be best illustrated with reference to the ship-example. The museum visitor is in low stakes context, whereas the shipowner in a high stakes context. Following the contextualist model, the museum owner is reasonable in believing an epistemically risky belief according to normal evidentialist norms by following their evidence, as there is no pragmatic risk involved. Contrarily, the shipowner is not reasonable in believing the risky proposition and must strengthen their evidence to form a belief with low to no epistemic risk, according to the pragmatically warranted norms. This interpretation does not touch on third-person knowledge attributions, which may be seen by some as the strength of pragmatic encroachment by explaining the intuitions in the ship-example. 14  My interpretation considers the context of inquiry, which amounts to an epistemic standard about when someone is considered to have reasonable beliefs. So, instead of an explanation of intuitions about why we attribute knowledge to one or the other, this approach presents an explanation concerning the epistemic value the believers acquire.

However, this contextualist interpretation suffers from a critique coined by Feldman concerning borderline cases. He states that any theory proposing that one should only believe a proposition when the evidence is sufficiently strong will encounter cases in which the strength of the evidence lies on the border between believing and suspending judgment. 15 A variation of this critique also goes for the contextualist interpretation: how does one determine when the stakes are sufficiently high enough to switch to the other context? There are countless cases in which there is something pragmatically at stake, but the stakes are not high enough to demand such high epistemic standards. However, it appears that this is not different from the problem a regular contextualist theory of knowledge faces. Thus, one may consider this to be an intuitive switch and let people decide this according to, as Feldman says, ‘their own attitude’. But, perhaps, this is not convincing just yet.

A second way one may develop for my proposal is a gradual interpretation, similar to Reed’s in the conclusion of his article, but in this case including pragmatic encroachment. 16 This would entail a negative relationship between the level of pragmatic risk and epistemic risk that may be taken to consider the risky belief reasonable. If the pragmatic risk increases, the level of epistemic risk that is allowed decreases. This means that, if indeed the belief is epistemically risky, one needs stronger evidence to decrease the risk of the belief before being reasonable in believing it. How much stronger the evidence must be, gradually increases or decreases with the stakes. However, this approach also runs into the problem of borderline cases. This time it has the character of an ‘infinite-regress’ argument. Imagine finding in front of you two bombs, one of which will go off after the timer runs out. There is no evidence you can gather about which bomb is the one that will go off and you only have time to dismantle one. As such, epistemically, you ought to suspend judgment about which bomb needs to be dismantled: the belief carries great epistemic and pragmatic risk and, as such, the standard of evidence needed to form a belief about the bombs to dismantle one of them keeps getting higher as the bomb nears its explosion. This infinite increase would allegedly leave one in a state of inertia as one could form no reasonable belief on which to act. This dilemma seems to pose a problem to the gradual interpretation of soft pragmatism but is easily mitigated. As one cannot know which bomb is the one about to go off, the only thing one can do in this situation is do something. As such, all that is needed for the ‘believer’ to respond correctly is that they believe that they should dismantle one of the bombs. This is enough to spur them into action, even if they lack a belief about which bomb they should choose. As such, objections of this type need not be cause for concern.

To conclude these contemplations, I would like to compare the strengths and weaknesses of the different general positions one may take in this debate with those of the soft pragmatic theory I have attempted to develop. First, it seems to fare better than the evidentialist approach in giving a reasonable account of high stakes situations, like the ship-example. This is a strength that my theory shares with pragmatic encroachment in general. Second, it appears to me that it fares better than some pragmatic approaches since it warrants epistemic value. As Brian Kim claims, pragmatists need to explain away the intuitive appeal of purism and importance of epistemic value.17This is something my soft pragmatism circumvents by accepting and warranting evidentialist norms. However, a strong point that pragmatic encroachment enjoys is the explanation of the intuitions about third-person knowledge attribution in ship-examples. As stated before, I have disregarded those knowledge attributions in this paper. However, I believe my position does account for a more fundamental idea. Essentially, the intuition in the ship-example demonstrates the fact that the pragmatic action-guiding property of knowledge is important to us, which is at odds with our other purist intuitions. Thus, by weaving the pragmatic into one’s epistemic theory, I have nevertheless been able to warrant the value upon which the intuition is based.

However, it may be considered a weakness of my theory that it does not account for knowledge. This means that the theory carries less weight, as most people would consider knowledge as epistemically more important than beliefs. This flaw also brings along an advantage, as avoiding saying anything about knowledge means the theory does not face the high standards knowledge must own up to. Pragmatists often encounter the criticism that pragmatic encroachment destabilizes or devalues knowledge. As such, it boils down to the question of epistemic value: what is it that our epistemic theories should be talking about? I have presented Feldman’s answer, but others, such as Engel, have argued that epistemic value must be assigned to knowledge.18 Nonetheless, I would argue that epistemic value should be assigned to reasonable beliefs, for it makes epistemic value realistically available. The fallibilist principle of reasonable beliefs gives a better account of epistemic value if we want to insist that it exists. Because if epistemic value would exist but be nearly impossible to attain, I would wonder why the pragmatic value of knowledge is not the principal standard.

In summary, I have attempted to assign pragmatic justification a place in an evidentialist framework. Although pragmatic encroachment contradicts the purist norms of evidentialism, I believe the evidentialist fundamentals have remained warranted on more fundamental than merely conflicting norms. I have argued that pragmatic factors affect a person’s epistemic excellence since it is tied to the definition of ‘a rational attitude’, which is determined by pragmatic factors. This was substantiated by arguing for higher epistemic standards in high stakes situations. Essentially, I have interfered with the debate about the difference between the pragmatic and the epistemic. From the foregoing considerations, I believe one has reason to think that, in the realm of the reasonable, the dichotomy between the pragmatic and the epistemic blurs.


LinkedIn profile Emma Young.