Conceptions of Epistemic Value

Abstract

This paper defends a conception of epistemic value that I call the “Simpliciter Conception.” On it, epistemic value is a kind of value simpliciter and being of epistemic value implies being of value simpliciter. I defend this conception by criticizing two others, what I call the Formal Conception and the Hybrid Conception. While those conceptions may be popular among epistemologists, I argue that they fail to explain why anyone should care that things are of epistemic value and naturally undercuts disputes about what is of epistemic value. I end by sketching and locating my conception within some increasingly popular views in meta-normativity.

Introduction

But I doubt that anyone endorses epistemic standards on the grounds that they are one's own; presumably they are invoked because they mark out something valuable, either intrinsically or extrinsically. Goldman (1991: 189)

The overall aim of this paper is to defend a conception of epistemic value that I call the Simpliciter Conception. On it, epistemic value is a kind of value simpliciter and being of epistemic value implies being of value simpliciter. I defend this conception by criticizing two alternatives. According to the Formal Conception, epistemic value is not a kind of value simpliciter; and being of epistemic value does not imply being of value simpliciter. According to the Hybrid Conception, epistemic value is not a kind of value simpliciter but being of epistemic value does imply being of value simpliciter. I argue that these conceptions fail to explain why anyone should care that things are of epistemic value. Additionally, they undercut philosophical disputes about what is of epistemic value.

In section 1, I characterize the target notion of final epistemic value a little more fully. In section 2, I exposit the Formal Conception. In section 3, I object that the Formal Conception fails to explain why it is appropriate for people to care that certain things are of epistemic value; I also object that the Formal Conception naturally undercuts the philosophical significance of disputes about what is of epistemic value. In section 4, I extend my objections to a Hybrid Conception of epistemic value, which maintains that the standards of the epistemic point of view are constituted by our states (beliefs) or activities (reasoning). In section 5, I exposit the Simpliciter Conception and show how it avoids the problems of the Formal Conception and Hybrid Conception. Finally, in section 6, I explore what kinds of positions the Simpliciter Conception sits most naturally with, suggesting, without defending, a position I call Unified Simpliciter Conception about Epistemology.

1. Final epistemic value

Let's begin with an initial characterization of epistemic value and a few distinctions. First, we can initially characterize epistemic value as follows:

Something is of epistemic value if and only if it is valuable from the epistemic point of view.

This naturally raises the question of what the epistemic point of view is. I take the epistemic point of view to contain the standards for epistemic evaluations. These standards take the form of statements articulating the conditions – necessary, sufficient, or paradigmatic – of when things have epistemic properties. The epistemic point of view contrasts with other points of view, such as the ethical point of view or the aesthetic point of view, as well as perhaps the religious, legal, prudential, etc. points of view.

Second, we can mark a distinction between two types of epistemic value. Some things are of instrumental epistemic value. Something is of instrumental epistemic value when the overall epistemic value of its consequences is positive. But of course the value of consequences do not just go on forever; they ultimately bottom out in things that are not of instrumental epistemic value. While one can find a variety of terms in the literature, I'll use the modifier ‘final’ and provide the following characterization:

Something is of final epistemic value if and only if it is valuable in and of itself, or for its own sake, from the epistemic point of view.

And, of course, to understand the overall epistemic value of consequences we need more than just the concept of final epistemic value. We need its contrary:

Something is of final epistemic disvalue if and only if it is disvaluable in and of itself, or for its own sake, from the epistemic point of view.

With the concepts of both final epistemic value and disvalue, we could provide an adequate characterization of the overall epistemic value of consequences in terms of them.¹

None of these distinctions, so far, tell us what is of final epistemic value. Of course, any time something is of final epistemic value there is some fact or obtaining state of affairs specifying that the thing is of final epistemic value. So, alternatively put, none of these distinctions so far tell us what the facts of final epistemic value are. There is a general, though not exceptionless, consensus that truth is important for final epistemic value and falsehood is important for final epistemic disvalue. For purposes here, we can assume just a simple formulation of this idea to grease the wheels of discussion:

(T) If a belief is true, then it is of final epistemic value.

As this claim is merely illustrative, it doesn't matter if it is true for our purposes here. So I won't defend it.

2. Formal conception of final epistemic value

I've characterized final epistemic value in terms of the epistemic point of view. But none of these terms are ordinary ones, even if what they refer to might be commonplace. Consequentially, it is important to further develop a conception of final epistemic value. In this section, I will exposit what I call the Formal Conception of final epistemic value.

The Formal Conception of final epistemic value embraces several theses. First:

Epistemic Point of View: The epistemic point of view contains true statements expressing the standards for epistemic evaluation.

Since I am focusing on epistemic value, I am focusing on a subset of the epistemic point of view that states the standards for epistemic value. Since I am assuming (T), I am also assuming (T) is part of the epistemic point of view.

Second, a claim about the structure of justification internal to the Epistemic Point of View:

Axiomatic Justification: Some of the statements in the epistemic value point of view are justified in virtue of their relation to other statements in the epistemic value point of view; but some are not.

So the epistemic value point of view has something like an axiomatic structure. There are some statements in it that have an “axiomatic” status; they are not justified by appealing to anything else in the point of view. Other statements have a “theorem” status; they are justified by appealing to one of the axiomatic statements. The simplest way to justify a theorem is good ole fashioned deduction. But there might be looser ways of justifying it as well.²

Some authors might prefer the more modern terminology of “fundamental” and “derived.” Some of the statements in the epistemic point of view are “fundamental” in that they are not justified in virtue of their relation to other statements in the epistemic point of view. Other statements in the epistemic point of view are “derived” in that they are justified in virtue of their relation to other statements in the epistemic point of view. (How they are “derived” is an issue we don't have to settle here; many views are possible.) Using this language, we could then say that certain values or obligations are “fundamental” where this means that they are referred to by at least one fundamental statement in the epistemic point of view. I do not think anything turns on whether we used “axiomatic/theorem” language or “fundamental/derived” language. The important point is that the standards in the epistemic point of view have an internal structure.

The next two claims require a little setup. Contrary to error-theorists (e.g., Mackie 1977) and others (e.g., Geach 1956; Thomson 2008; Kraut 2011), I accept the existence of final value (sometimes also called intrinsic value or absolute goodness). I will not defend this view, simply noting that I don't find criticisms of this view persuasive.

I characterize final value as follows:

Something is of final value if and only if it is appropriate or correct to value it for its own sake or in and of itself.

By ‘to value’ here I have in mind a range of pro-attitudes that include things like: liking, caring about, desiring, admiring, advocating for, enjoying, taking pleasure in, etc. Lot of philosophers who accept the existence of final value might fuss with this approximation, even if they agree with the general idea. Specifically, they might disagree about whether the right way to “connect” final value to valuing is by appealing to what is “appropriate or correct.” For instance, they might say instead that something is of final value if and only if we ought to value it; or there is reason to value it; or it is rational to value it; etc. This inhouse fight doesn't matter here.

Third, the Formal Conception takes a stand on the relationship between final epistemic value and final value (or final value simpliciter). The Formal Conception contains the following two theses:

Kind Failure: Final epistemic value is not a kind or subspecies of final value.

Implication Failure: Being of final epistemic value does not imply being of final value.

Implication Failure is weaker than Kind Failure. Having one property might imply having another because the former is a type or kind of the latter. But having one property might imply having another without the former being a type or kind of the latter. Being crimson implies being red and being extended; but while being crimson is a type or kind of being red it is not a type or way of being extended. As I formulate it, the Formal Conception embraces both of these theses.

To be clear, it is consistent with Formal Conception that some things are both of final epistemic value and final value. But this would not necessarily be because those things have the property of being of final epistemic value. On the Formal Conception, there is no conceptual, metaphysical, or deep causal connection between the properties of final epistemic value and final value. An analogy. One could imagine the “white supremacist” point of view that articulates the standards for white supremacy. And that point of view might imply that the experiences of pleasure of some whites is of final white supremacist value. And, as a matter of fact, those experiences of pleasure might also be of final value. But this wouldn't be of final value because they are of final white supremacy value; it is merely because there is coincidental overlap with final value.

We can now see why I call this conception of epistemic value a “formal” conception. On it, certain statements – including perhaps (T) – play a purely “formal role” inside the epistemic point of view. They are used to justify statements about epistemic standard inside the epistemic point of view. But as these last two theses make clear, on their own, they play no role in connecting epistemic value to final value.

Several authors endorse the Formal Conception or something close enough.³ Consider the following discussion from Ernest Sosa. He begins by reminding us that:

We humans are zestfully judgmental across the gamut of our experience: in art, literature, science, politics, sports, food, wine, and even coffee; and so on, across many domains. (Sosa 2007: 70)

He continues:

Any such domain of human experience admits values of two sorts: the derivative, and the fundamental – that is to say, the derivative or fundamental for that domain. A value might be irreducible to other values distinctive of a given domain, without being fundamental absolutely. (Sosa 2007: 70)

Sosa goes on to consider whether only truth is of the fundamental epistemic value. In describing issues regarding this view, he writes:

Truth may or may not be intrinsically valuable absolutely, who knows? Our worry requires only that we consider truth the epistemically fundamental value, the ultimate explainer of other distinctively epistemic values. (Sosa 2007: 72)

Similar claims can be found in other authors. Here is Duncan Pritchard.

When I say that truth is the fundamental epistemic good, I mean that from a purely epistemic point of view it is ultimately only truth that we should care about … A key point to note about treating truth as a finally epistemically valuable good is that it does not follow that truth is finally valuable simplicter. For instance, it does not follow that the truth always generates a pro tanto goodness. The point is just that from a purely epistemic point of view truth is rightly to be valued non-instrumentally. But this is entirely consistent with arguing that the general view of truth is merely instrumental (or indeed, that it has no value at all). (Pritchard 2014: 113)

In a slightly earlier work, Pritchard characterized “fundamental epistemic goods” as “any epistemic good whose epistemic value is at least sometimes not simply instrumental value relative to a further epistemic good” (2010: 11–12). He claims that “the very idea of there being such a thing as epistemic value … presupposes that some epistemic goods are fundamental. Something, after all, needs to act as the terminus for the instrumental regress of epistemic value” (Pritchard 2010: 12). Nonetheless, “while a fundamental epistemic good can act as the terminus of the instrumental regress of epistemic value, this is entirely compatible with that good not being finally valuable simpliciter” (Pritchard 2010: 12).

Sosa and Pritchard endorse the Formal Conception, or something close enough. I highlight them because of their prominence. But they are not alone. A number of other authors have suggested similar views; see, inter alia, Resnik (1994: 35), Matheson (2011: 391–2), McHugh (2012: 22–3), Ahlstrom-Vij (2013a: 21, 2013b: 158–9), Ridge (2013: 194ff.), and Sylvan (2018: 385).

3. Against the Formal Conception

In this section, I criticize the Formal Conception. First, I raise a flat-footed objection that the Formal Conception fails to imply that it is appropriate to care about the fact that some things are of epistemic value. While I find this objection compelling, others might think it a boon that it lacks this implication. For, if the Formal Conception lacks this implication, people may continue to theorize about epistemic value while immunizing themselves from skepticism about final value. I then argue that this is a mistake – the Formal Conception naturally undercuts the significance of theorizing about epistemic value. It is thus not a boon to lack this implication.

3.1. A flat-footed objection

Suppose I am unconcerned about epistemic value. I concede that various things have epistemic value. But I am not interested in having them for that reason; I neither desire them nor seek after them; I am not pleased when I learn something is of epistemic value; and I have no interested in promoting those things of epistemic value. Am I making a mistake? Is there something incorrect or defective or problematic about my attitudes, even if only slightly so? Would my lack of care tell you something about my virtues or vices? If the Formal Conception is correct, the answers to these questions are negative. I am not making a mistake; there is nothing problematic, defective, or incorrect about my attitudes; and my lack of concern tells you little about my virtues or vices. However, I think such a person would be making a mistake; that there is something incorrect about such a lack of concern. For this reason, I reject the Formal Conception.

Let me clarify this objection. When I speak about being concerned about epistemic value, I have in mind not just caring about individual things that are of epistemic value but caring about the fact that such things have epistemic value.⁴ Something might be of final epistemic value and also of final value or even instrumental value. In such a situation, a person might care about that thing but only because it is of final value. But this doesn't mean that they care about the fact that the thing is of final epistemic value.

This clarification defuses a potential rejoinder to this flat-footed objection. Some philosophers claim that true beliefs are of final epistemic value and false beliefs are of final epistemic disvalue. They then provide the following story. Having true beliefs, instead of false beliefs, helps contribute to living longer. (Usually reference is made to Quine's pithy, though characteristically lax, statement that “Creatures inveterately wrong in their inductions have a pathetic but praiseworthy tendency to die before reproducing their kind” (1969: 126).) Finally, these philosophers might claim that living longer is, in general, valuable. Thus, they might claim it is appropriate and correct to value true belief; in general not doing so does reveal something of one's character, namely, a lack of concern for one's future well-being.

For sake of discussion, let's grant that living longer is, in general, valuable and true beliefs, in general, are more likely to lead to living longer than false beliefs. Nonetheless, none of this indicates that it is correct to value final epistemic value. For the fact that true beliefs are of final epistemic value, and false beliefs are of final epistemic disvalue, is entirely epiphenomenal to this story. Otherwise put, this story – if successful – might explain why it would be correct to value true beliefs (at least instrumentally), but it would not explain why it would be correct to value the fact that true beliefs are of final epistemic value.

I've characterized final value as something that is appropriate to value for its own sake, so that not valuing it is, in one sense, a mistake. Given this characterization of final value, proponents of the Formal Conception do not accept:

If X is of final epistemic value, then it is appropriate to value X for its own sake.

But they might still accept that there is some connection between epistemic value and valuing. They might propose one of the following two principles as explaining that connection:

If X is of final epistemic value, then it is appropriate to epistemically value X for its own sake.

If X is of final epistemic value, then it is epistemically appropriate to value X for its own sake.

Let's take these proposals in turn.

First, as noted above, when I talk of “valuing” – or the perfect tense ‘value’ – I have in mind adopting a valuing attitude or pro-attitude such as liking, caring about, being concerned with, desiring, advocating for, being in favor of, admiring, etc. Thus, when it is appropriate or correct to value something, that means it is appropriate to adopt some of these attitudes towards that thing. However, the first proposal refers to “epistemically value X.” That, in turn, suggests that among the pro-attitudes there are some that are the “epistemically valuing” attitudes in contrast to the “ethically valuing” attitudes or the “politically valuing” attitudes, etc. But I confess I can't make sense of that idea. None of these attitudes seem especially “epistemic” or “ethical” or “political” or whatever. They are all attitudes that can be had towards a range of things. So I can't make sense of this first proposal because I can't make sense of the idea of a distinctive range of “epistemically valuing” attitudes.

The second proposal does not suffer from that problem. When it refers to “value X” it can have in mind any or all of the pro-attitudes. It says that sometimes these pro-attitudes can be “epistemically appropriate.” The problem with this proposal is trying to understand the idea of some pro-attitudes being “epistemically appropriate” (as well as its contrary “epistemically inappropriate”).

One proposal might be this:

A pro-attitude directed towards something is epistemically appropriate if and only if it is appropriate to direct pro-attitudes towards that thing for their own sake and the attitude has property P (where ‘P’ names some property that explains what makes the attitude epistemically appropriate).

We don't really have to evaluate potential candidates for P here. For this proposal is inconsistent with the Formal Conception, specifically Implication Failure. From the claim that something is of final epistemic value, it will follow that it is epistemically appropriate to value it for its own sake; from the claim that it is epistemically appropriate to value it for its own sake, it follows that it is appropriate to value it for its own sake; and from the claim that it is appropriate to value it for its own sake, it follows that it is of final value. But this contradicts Implication Failure.

An alternative proposal might be this:

If one is using the epistemic point of view, then it is appropriate to value for its own sake what the epistemic point of view identifies as being of final epistemic value.

This may be what Duncan Pritchard has in mind when he writes, “The point is just that from a purely epistemic point of view truth is rightly to be valued non-instrumentally” (2014: 113).

I understand this proposal and do not think it is logically inconsistent with the Formal Conception. Nonetheless, it is quite implausible. The basic problem is that even if a point of view identifies something of its kind of final value and you are using that point of view, it does not follow that it is appropriate to value for its own sake whatever that point of view identifies as valuable. Here's an obvious example. Suppose Andrew is using the white supremacist point of view. And the white supremacist point of view identifies intimidating ethnic minorities and immigrants as being of final white supremacist value. In no way does it follow that it would be appropriate for Andrew to value the intimidation of ethnic minorities and immigrants, regardless of what point of view Andrew is using.

Summing up, the Formal Conception fails to imply that it would be appropriate or correct to value the fact that things are of epistemic value. Since I think we should care about such facts, insofar as we can, I think we should reject the Formal Conception.

However, some proponents of the Formal Conception might have a different perspective. They might take it as a boon that they reject Implication Failure. For they might be worried about skepticism about final value.⁵ Endorsing the Formal Conception may be a gambit to simultaneously allow for theorizing about what is of final epistemic value while immunizing such theories from worries about final value. However, I will now argue that this gambit won't pay off. Endorsing the Formal Conception naturally undercuts the philosophical significance of disputes about what is of epistemic value.

3.2. The Formal Conception and other points of view

To introduce this objection,⁶ let us start with perhaps the major problem in contemporary work on epistemic value. The problem takes for granted three assumptions:

True beliefs are of final epistemic value.

Knowledge is of final epistemic value.

Knowledge is of more final epistemic value than the corresponding true belief.

The problem is to explain how each of these assumptions are true together or, if they are not all true, which assumption should be rejected.

I'll mention three potential solutions. One potential solution is to argue that knowledge has some further condition besides true belief and that meeting this condition is a fundamental final epistemic value distinct from true belief. This is, essentially, Sosa's (2007) solution. Sosa's solution amounts to maintaining that knowledge requires apt belief and that apt belief is, like true belief, a fundamental epistemic value (2007: 87–8). A second solution rejects the third assumption. Instead, it maintains that frequently knowledge is of more instrumental epistemic value than true belief, but is not itself always or even frequently of more final epistemic value than true belief. This solution is, more or less, that of Goldman and Olson (2009). A third solution rejects the first assumption. Instead, it claims that justified belief (not true belief) is a fundamental epistemic value and claims that justified belief is necessary for knowledge. Though not presented in quite this way, Feldman (2002) contains such a solution.

Earlier I said that the epistemic point of view contains the standards for epistemic evaluation. We can understand these different solutions as being equivalent to making claims about the epistemic point of view and its axiomatic standards. The first solution says that the epistemic point of view contains as a further axiomatic standard besides (T) something like:

(A) Apt beliefs are of final epistemic value.

The second solution, among other things, denies that (A) is a further axiomatic standard in the epistemic point of view. The third solution denies that (T) is an axiomatic standard of the epistemic point of view to begin with and proposes in its place:

(J) Justified beliefs are of final epistemic value.

However, at this point, let us consider some parody proposals. According to the first parody proposal, Sosa is correct about the epistemic point of view and the axiomatic statements of it. However, on this proposal, there is also the epistemic* point of view that contains the standards for epistemic* evaluation. It also has an axiomatic justification structure. Additionally, according to it, there is only one axiomatic statement:

(T*) True beliefs are of final epistemic* value.

And, like epistemic value given the Formal Conception, when something is of final epistemic* value it does not follow that it is of final value.

According to the second parody proposal, Sosa is correct about the epistemic point of view and the axiomatic statements of it. However, on this proposal, there is also the epistemic** point of view that contains the standards for epistemic** evaluation. It also has an axiomatic justification structure. Additionally, according to it, there is only one axiomatic statement:

(J**) Justified beliefs are of final epistemic** value.

And, like epistemic value given the Formal Conception, when something is of final epistemic** value it does not follow that it is of final value.

Each of these parody proposals has the same structure. It concedes that some author – e.g., Sosa – is right about the epistemic point of view and what has final epistemic value. It then articulates an alternative point of view concerning a different kind of value, understood in terms of that point of view, where the things that have that kind of value are different from the things that have final epistemic value. Using parody proposals, we can construct a proliferation of points of views with their associated values.

However, given this proliferation of points of views, with their different value properties, a natural question is whether any of them should be privileged over others. For each of these parody proposals agrees with (e.g.) Sosa about what is of epistemic value. When it comes to their views about the distribution of epistemic value, there is no disagreement. The potential disagreement is about which point of view should be utilized and adopted. Is there anything about the epistemic point of view that privileges, vindicates, or otherwise legitimates it over these other points of view? Something that would explain why someone who opted out of the epistemic point of view for the (e.g.) epistemic** point of view was making a mistake? My contention is that if the Formal Conception is true, then nothing would privilege it.

But philosophical disputes about epistemic value are significant only if the epistemic point of view is privileged. For philosophical disputes about epistemic value are significant only if it is important to use the epistemic point of view over others. After all, as the parody proposals illustrate, it is possible for us to evaluate the world in various ways, even ways that parody or mimic participants in disputes about epistemic value. And it is important for us to use the epistemic point of view over others only if it is privileged over these other points of view. But if the Formal Conception is true, then the epistemic point of view is not privileged over others.

Thus, the Formal Conception naturally undercuts the significance of philosophical disputes about epistemic value. To be clear, one could still participate in such disputes (perhaps such participation is amusing or is useful for livelihood). But given the Formal Conception, any such bold proclamations about the profound significance of such debates, that philosophers tend to make, cannot be substantiated.

Proponents of the Formal Conception might propose several ways of trying to privilege the epistemic point of view over others. First, someone like Sosa (2007) might argue that one cannot use the epistemic* point of view to resolve the problem we started with. And this provides a reason to prefer using the epistemic point of view over the epistemic* point of view. Sosa is right that one cannot use the epistemic* point of view to resolve that problem. That's because that point of view is not concerned with epistemic value. But he's wrong that this provides a reason to prefer the epistemic point of view over the epistemic* point of view. For the epistemic point of view cannot be used to solve problems about epistemic* value.

Second, someone like Pritchard (2014) might claim that the activity of inquiry provides a reason to prefer the epistemic point of view over others like the epistemic* point of view. Inquiry is a goal-directed activity that is successful when it achieves its goal. And the goal of inquiry is co-extensive with what is of final epistemic value. So, someone like Pritchard might claim, this gives us a reason for preferring the epistemic point of view over other points of view like the epistemic* point of view. However, we can concede all these points about inquiry, but they do not provide a reason for preferring the epistemic point of view over others. For let us consider inquiry*. Inquiry* is a goal-directed activity that is successful when it achieves its goal. And the goal of inquiry* is co-extensive with what is of final epistemic* value. But so far we've been given no reason to privilege inquiry over inquiry*. So appealing to a connection between inquiry and final epistemic value will not help.

Finally, proponents of the Formal Conception might appeal to broadly pragmatic features. They might propose that we are already using the epistemic point of view, instead of others. Or perhaps it might be more difficult to use other points of view like the epistemic** point of view. Or maybe people just prefer the epistemic point of view over others for whatever reason.

It is not obvious that these claims about pragmatic features are correct. (For instance, it has proven quite elusive to adequately analyze the contemporary concept of knowledge; maybe a point of view without such a concept would be more useful.) But let's concede them for sake of discussion. One issue is that these pragmatic features are highly variable and contingent. They might fail across time or right now across different populations. So this response might work now, but not for tomorrow; it might work for some populations, but not others. A different issue is that this response is conservative. Individuals or communities can keep using their certain points of view because that's what they've always done; but bolder individuals or community might opt for different ones. Variety is the spice of life, they say.

Additionally, some philosophers – such as Pritchard (2010: 16–18) – find unsatisfying views that maintain that knowledge is not more epistemically valuable than true belief but only more practically valuable than true belief. However, it would be very hard to maintain that view while also maintaining the view that the primary reason why we should stick with the epistemic point of view, and epistemic value, are practical and pragmatic reasons!

Let me sum up the dialectic at this point. First, I argued that, given the Formal Conception, it is not appropriate or inappropriate to not care about epistemic value. Since, I claim, it is appropriate to care about epistemic value, there is reason for rejecting the Formal Conception. Second, I considered a response by proponents of the Formal Conception who prefer to distance themselves from claims about final value so as to make way for important theorizing about what is of epistemic value. I've argued that this gambit fails. For the Formal Conception fails to give priority to the epistemic point of view over others.

4. A Hybrid Conception?

I formulated the Formal Conception as embracing two theses:

Kind Failure: Final epistemic value is not a kind or subspecies of final value.

Implication Failure: Being of final epistemic value does not imply being of final value.

However, one might consider a view that accepts the first thesis while rejecting the second thesis. I will call such a position a Hybrid Conception of epistemic value, since it combines elements of both the Formal Conception and my preferred alternative, the Simpliciter Conception.

One might develop the Hybrid Conceptions in various ways. For the sake of moving discussion forward, I'll consider one specific way of developing it in along broadly “constitutive” lines. This way of developing the Hybrid Conception will have three additional theses. First, the standards of the epistemic point of view are “constituted” by our psychological states (such as beliefs) or psychological activities (such as reasoning). Second, since those states or activities are universal for human beings, who meet certain conditions, the epistemic point of view is likewise universal for such human beings. In this way, the standards of the epistemic point of view are inescapable for human beings like us (who meet the relevant conditions). Finally, the relevant properties in the standards do not bear any conceptual, metaphysical, or deep causal relations to their counterpart properties. So, on this approach, final epistemic value does not bear any conceptual, metaphysical, or deep causal relations to final value. In slogan form, on this approach to developing the Hybrid Conception, epistemic normativity is constitutive normativity and constitutive normativity ain't real normativity.⁷

I will briefly argue that the Hybrid Conception – including the constitutive way of developing it – does not avoid my objections to the Formal Conception.

First, the Hybrid Conception merely in virtue of rejecting Implication Failure does not avoid my first objection to the Formal Conception. To see this, recall the White Supremacist example. The White Supremacist point of view may identify something as being of final white supremacist value. And that thing may be of final value. But insofar as it is of final value it is not because it is also of final white supremacist value. After all, the property of being of final white supremacist value bears no conceptual, metaphysical, or deep causal relations to final value. There is merely overlap here, nothing more.

Likewise, even if the Hybrid Conception rejects Implication Failure, that rejection by itself is insufficient to avoid my first objection. For even if Implication Failure is false, there may still be no conceptual, metaphysical, or deep causal relations between final epistemic value and final value. To be sure, there would be a total modal overlap – any time something is of final epistemic value, it is also of final value. But for all that has been said, it may still be that such overlap is merely a modal overlap and nothing more. Additionally, appealing to the constitutive way of developing the Hybrid Conception does not help. For, on the formulation used here, it explicitly denies that there are any conceptual, metaphysical, or deep causal relations between final epistemic value and final value.

One might consider an important variant of the constitutive way of developing the Hybrid Conception. On this variant, meeting the constitutive standards for certain states (e.g., beliefs) or activities (e.g., reasoning) does bear conceptual, metaphysical, or deep causal relations to what is of final value. But it would be hard to see how to plausibly develop this view. In general, from the mere fact that a state or activity has a constitutive standard, it does not follow that such states or activities have any final value. After all, there are many deplorable, disgraceful, or otherwise disgusting states or activities with constitutive standards – ethnic cleansing comes easily to mind – but from the mere fact one has those states, or is engaging in those activities, clearly nothing about final value follows.⁸

These points hold even if the relevant state or activity is inescapable for the agent.⁹ After all, pointing out that something is inescapable for an agent is not necessarily a way of vindicating it; it may be a way of revealing one's plight. To use a literary example, Sisyphus is cursed by Zeus to roll a boulder up a hill for all eternity. We may even suppose that Zeus makes certain desires and intentions to roll the boulder up the hill inescapable for Sisyphus as well. So for Sisyphus, there are certain actions, desires, and intentions for him that are inescapable. And, I may concede, those actions, desires, and intentions may have their own constitutive standards. But none of that would show that Sisyphus’ activity was of any great value. And, in his more reflective moments, he might ask whether he has reason to engage in these activities – he doesn't suddenly (mysteriously?) lose the ability to ask such questions simply because these activities are inescapable for him. And he might, rightly, conclude he has no reason to perform them and rightly wish – or even desire – not to have his fate. Thus, even if certain states (beliefs) or activities (reasoning) have constitutive standards and are inescapable for human persons like you or I, it would not thereby follow that meeting those standards bears any conceptual, metaphysical, or deep causal connections to final value.

Additionally, I don't think that the constitutive approach to developing the Hybrid Conception avoids my second objection to the Formal Conception. At first blush, it may seem to avoid it. For, one might claim, what “privileges” the epistemic point of view, over others, is that the epistemic point of view is the one that is constituted by our actual states (beliefs) or activities (reasoning). In other words, the epistemic point of view is privileged over others because it is “inescapable” for human beings like us, with our states and activities.

But from the mere fact that some point of view is inescapable for us may not be sufficient to vindicate a choice of it over other points of views. Inescapability is normatively neutral. Additionally, upon considering other points of view, we might prefer them and thus try to resist utilizing the standards of the epistemic point of view. Of course, if the constitutive approach to the Hybrid Conception is correct, we will never be able to fully avoid the epistemic point of view. But we can still adopt other points of view as well.

To illustrate my response, let us suppose that given our evolutionary history – something beyond the control of any of us – human beings are innately disposed to reason in a certain way. Specifically, we are innately disposed to use “in-group” reasoning, where individuals are given preference in action and thought simply in virtue of arbitrary features that are similar to whatever agent happens to be reasoning. There are, we may suppose, constitutive standards for “in-group” reasoning; not just any form of reasoning could be “in-group” reasoning. For human beings such as ourselves, these dispositions to “in-group” reason might be inescapable. But we still might not see that as any reason to privilege them. In fact, we might take steps to avoid, mitigate, or compensate for such reasoning.

A final, more extreme illustration. Suppose I am – à la Dr. Frankenstein – in the business of creating some new creatures. I recognize that running and slithering are different ways of moving, each with their own constitutive standards. Nonetheless, noting the differences by itself wouldn't settle which way of moving would be more preferable, good, or ideal for my new creatures. Likewise, I might recognize that believing and “delieving” – a state similar to believing but distinct¹⁰ – are different psychological states, each with its own constitutive standards enshrined in distinct points of view. Nonetheless, noting the differences by itself wouldn't settle which way of designing the psychology of these creatures would be more preferable, good, or ideal.¹¹ Thus, even if there are constitutive standards for our states or activities, and even if they are inescapable for us, and even if those standards are enshrined in the epistemic point of view, this need not be enough to vindicate the epistemic point of view over others.

Proponents of the Hybrid Conception – and this constitutive approach to developing it – may have their responses. But I hope to have provided enough justification for an alternative approach, which I now turn to.

5. The Simpliciter Conception

In this section, I want to sketch an alternative account. I'll call it the Simpliciter Conception. The Simpliciter Conception embraces the first two claims of the Formal Conception:

Epistemic Point of View: The epistemic point of view contains statements expressing the standards for epistemic evaluation.

Axiomatic Justification: Some of the statements in the epistemic value point of view are justified in virtue of their relation to other statements in the epistemic value point of view; but some are not.

However, it rejects the next two claims:

Kind Failure: Final epistemic value is not a kind or subspecies of final value.

Implication Failure: Being of final epistemic value does not imply being of final value.

Instead it embraces:

Kind Inclusion: Final epistemic value is a kind or subspecies of final value.

Implication: Being of final epistemic value does imply being of final value.

The Simpliciter Conception gets its name because it maintains that final epistemic value is a kind of final value or value simpliciter. The general picture here is that there is such a thing as final value that sub-divides into distinct categories with final epistemic value being one such category.¹²

The Simpliciter Conception avoids the issue plaguing the Formal Conception. The flat-footed objection was that the Formal Conception did not imply that it is appropriate to care about epistemic value. The Simpliciter Conception does imply this. Given it, if something is of final epistemic value, then it is of final value. And if it is of final value, then it is appropriate to care about it. Additionally, I complained that it was hard to make sense of the idea of it being “epistemically appropriate” to value something. However, given the Simpliciter Conception, a natural account is available. Some things are of final value; it is appropriate to value them. It is epistemically appropriate to value something just when it is of final epistemic value, which is a subset of final value. The qualifier ‘epistemically’ doesn't mark out a special kind of “appropriateness” or a special kind of valuing attitude; it merely marks out the object of the attitude.

Turning to the sophisticated objection, the Simpliciter Conception opens up a move that the Formal Conception does not. Some points of view do not “track” anything of value simpliciter; some do. According to the Simpliciter Conception, final epistemic value is a kind of final value. Thus, the epistemic point of view, which is concerned with final epistemic value, tracks what is of final value. There's no reason to think that the other points of view do. Thus, there is reason to prioritize the epistemic point of view over others – it, at least partially, tracks what is of final value.

6. Metanormativity

In this final section, I'm going to relate the Simpliciter Conception to various issues in what can be called, for lack of a better word, “meta-normativity.” I do not have the space to defend some of the more controversial positions. I'm rather more interested in exploring what kinds of views best fit with the Simpliciter Conception. To borrow a turn of phrase from Peter van Inwagen, I've argued for a view about final epistemic value, and I'm adopting some views about meta-normativity.

6.1. The domain of the epistemic

First, according to the Simpliciter Conception, final epistemic value is a kind of final value. But presumably there are other types of final value (e.g., final ethical value). One question is what demarcates final epistemic value from other types or kinds of final value; what demarcates the domain of the epistemic?

While I have views about this, I don't intend to exposit them here. But notice that this is a question for any theory of epistemic value that accepts the existence of epistemic value alongside the value of other things. Thus, proponents of the Formal Conception will also have to answer the question: what demarcates the domain of the epistemic? To be sure, they don't have to answer the particular way that question takes shape, given the Simpliciter Conception. But they still must face the general question. So while proponents of the Simpliciter Conception will have to demarcate epistemic value from other things, so will everyone else.

6.2. Extended normative pluralism

Second, I have assumed that we can meaningfully speak of final value simpliciter or san phrase. Indeed, on the view defended here, final value is presumably more conceptually basic than final epistemic value, with the latter being understood as a subset constructed out of the former. Some authors might be skeptical of this, who are not yet error theorists. Specifically, consider a position sometimes referred to as “normative pluralism.” A cornerstone of this position is that there's no such thing that we ought to do simpliciter or sans phrase, though there may be things we ought to do morally, prudentially, etc. (See, e.g., Copp 2007: Ch. 9; Tiffany 2007; Sagdahl 2014; Baker 2018.)

Now normative pluralism is not a position about final value simpliciter; it is a position about ought simpliciter. But presumably proponents of normative pluralism would extend their position to other (broadly speaking) normative properties like value properties. (For instance, they frequently extend their position to normative reasons so that there are no such things as reasons simpliciter but rather moral or prudential reasons.) I'll call extended normative pluralism a position that extends this idea to value, maintaining that there is no such thing as final value simpliciter or sans phrase (even if there are types of value with phrase). So understood, extended normative pluralism is inconsistent with the position here.

However, extended normative pluralism is inconsistent with a basic assumption of my discussion, that there is such a thing as final value simpliciter. It is not necessarily inconsistent with my view about the relationship between final epistemic value and final value. So extended normative pluralism does not provide any reason to doubt my proposal in particular. Indeed, even proponents of the Formal Conception might reject it if they think there are reasons for accepting final value that are independent of issues about epistemic value. Additionally, there are reasons for rejecting normative pluralism – and presumably, by extension, extended normative pluralism – that are independent of our discussion here. (See, inter alia, McLeod 2001; Dorsey 2013; Case 2016.) So I'll simply note that the Simpliciter Conception I'm advocating for is inconsistent with extended normative pluralism.¹³

6.3. Epistemology and ethics

Finally, it is relatively common to see ethics as studying final value. On my view, epistemic value – specifically final epistemic value – is a kind of final value. This naturally raises the question of what view on the relationship between ethics and epistemology comports well with the Simpliciter Conception.

One can naturally generalize the distinction between Formal and Simpliciter Conceptions of epistemic value to other epistemic concepts. After all, epistemology does not just describe epistemic value. It is shot through with other concepts like epistemic virtues and vices, epistemic justification, epistemic obligations, epistemic reasons.¹⁴ For each of these concepts, one can consider Formal, Hybrid, and Simpliciter accounts. For instance, the Formal Conception of epistemic virtues might say that the epistemic point of view contains the standards for what constitutes an epistemic virtue; but it might deny that epistemic virtues are a kind or species of virtue more generally. One could also formulate a Simpliciter Conception of epistemic virtue that also includes the claim that epistemic virtues are a kind or species of virtue.

Given a Simpliciter Conception of epistemic value, the most natural position is a Simpliciter Conception of these other concepts in epistemology. We might call this the “Unified Simpliciter Conception of Epistemology.” On it, epistemic reasons are a type or species of normative reasons more generally; epistemic virtues are a type or species of virtues more generally; etc. To use a slogan, the qualifier ‘epistemic’ doesn't switch topics, it merely restricts our topic.

The Unified Simpliciter Conception of Epistemology is a much stronger position that the Simpliciter Conception that I have been defending. For instance, the Simpliciter Conception of epistemic value I have been defending on its own takes no stand on these additional (apparently) normative categories. And it is possible that when we examine (for example) epistemic obligations we will find that they are importantly different from obligations simpliciter. Nonetheless, I suggest that given the Simpliciter Conception of epistemic value, the Unified Simpliciter Conception of Epistemology is quite natural.

The Unified Simpliciter Conception is not without precedent. Both Chisholm (1991) and Zagzebski (1996) suggest that epistemic evaluation and concepts is a type or subspecies of ethical evaluation and concepts. If by “ethics” they have in mind something that includes the study of value simpliciter, virtue simpliciter, reason simpliciter, etc. then their views are very similar to the Unified Simpliciter Conception of Epistemology.

More recent attention has come from work in meta-ethics or, perhaps better yet, meta-normativity. Specifically, Cuneo (2007) defended a version of moral realism by defending a version of epistemic realism and arguing that if epistemic realism were false, then so would be moral realism. In this way, moral realism and epistemic realism run “in parallel.” Cuneo's work has generated a lot of work, with some authors defending the parallel (e.g., Bedke 2010; Case 2019; Cuneo and Kyriacou 2018; Das 2016, 2017; Rowland 2013, 2016) and others criticizing it (e.g., Cowie 2014, 2016; Olson 2018).

There are some differences between positions like Cuneo's and Zagzebski's. Zagzebski describes epistemology as a sub-branch of ethics. Cuneo describes two different positions – epistemic realism and moral realism – and does not seem to identify one of them as being a part of the other. Nonetheless, this difference between authors may be primarily terminological. After all, it is clear that Cuneo thinks that epistemic and moral reasons share certain features – e.g., both are categorical. So presumably he would not object to an overarching subject matter that describes those shared features. This would put his position quite close to what I've called the Unified Simpliciter Conception.

Again, the Simpliciter Conception of epistemic value does not imply the Unified Simpliciter Conception. So one could embrace the Simpliciter Conception while rejecting the Unified Simpliciter Conception because one thought (for example) epistemic reasons and moral reasons were fundamentally different. Nonetheless, the Simpliciter Conception does sit comfortably with this increasingly popular position in meta-normativity.