2. Getting Clear on the Explanandum

Questions about how one should or ought to manage one’s own corpus of doxastic states—e.g. “What should I/they believe, given my/their circumstances?” “Should I/they believe p?”—often arise for us as live, even pressing, in the course of everyday life. Sometimes these sorts of questions are at the forefront of our minds. But much more often these sorts of questions present themselves underneath the surface of our conscious thought and are settled tacitly, as it were, without explicitly framing deliberation or reflection. Sometimes epistemic evaluations play a prospective role in shaping the ways in which we actually do regulate our doxastic states. But other times, the role is retrospective. We seem to expect conformity with epistemic norms both from ourselves and from others, and we are quick to criticize what we diagnose as our own or what we take to be each other’s failures on this front. We chastise one another for, e.g., jumping to conclusions, making overly hasty generalizations, conducting biased reviews of the available evidence, judgments that have been improperly clouded by anger or fear, wishful thinking, motivated reasoning, and so forth. We make efforts to help one another to avoid or overcome these sorts of failures and instead cultivate what we take to be proper, correct, or appropriate habits of doxastic state regulation (e.g. habits of fair-minded, impartial or unbiased, and appropriately critical thinking, of charitable interpretation, etc.).

It is particularly telling that, in the face of epistemic criticism, we tend to expect correction (and/or, perhaps, an excusing explanation for the relevant failure). If I tell my partner that their interpretation of the text of a colleague’s email is grossly uncharitable, I expect my partner to revise their beliefs about what their colleague intended to communicate (or, perhaps, to provide further justification in support of their original interpretation). Relatedly, coming to realize that one has made some sort of mistake, error, or misstep in regulating one’s own corpus of doxastic states typically involves experiencing an accompanying pressure toward correction: “I’m looking at the weather forecast: it’s going to rain this afternoon.” “Dummy! You’re looking at the weather forecast for Burlington, NC (and we’re in Burlington, VT).” “Oh, whoops! Ok, I have no idea whether it’ll rain here this afternoon… Hang on and let me pull up the weather forecast for Burlington, VT.” In pointing out that I’ve accidentally consulted the weather forecast for a different locale, my friend aims to spur me to give up my belief that it will rain today. And accepting my friend’s criticism does, in fact, prompt me to do just that. More generally, then, in recognizing that one has fallen short in conducting the business of being a doxastic subject, one is typically thereby moved to regulate one’s doxastic states differently so as to correct, repair, or at least ameliorate the relevant defect or shortcoming.

So, our ordinary thought and talk really does suggest that we think conformity with epistemic norms matters, all things considered. But more than this, at least for the sorts of doxastic subjects that ordinary thought and talk regard as paradigmatic, it seems that we think conformity with epistemic norms always matters, all things considered. Our pretheoretical commitment here is a commitment to the normative significance of epistemic evaluation as inescapable. This is somewhat striking. After all, there is a wide variety of domain-specific forms of evaluation which seem to have normative significance in some contexts and not in others. Any young, unmarried woman violates the norms of Victorian etiquette by being alone, unchaperoned, with a man. But this fact has no bearing on whether one of my undergraduates who happens to be an unmarried woman ought to spend some time alone and unchaperoned studying with a classmate who happens to be a man. What it would take for my undergraduate to conform with norms of Victorian etiquette is simply irrelevant to how my undergraduate ought, all things considered, to comport herself in this context. A woman living in London circa 1850, however, might well find herself in a context where it does matter, all things considered, that she avoids being alone, unchaperoned, with a man. Evaluations with respect to the norms of Victorian Etiquette, then, seem to have normative significance in certain contexts, for certain subjects but not others.

In contrast, ordinary thought and talk seems to presume that, at least for the sorts of doxastic subjects that ordinary thought and talk regards as paradigmatic, the normative significance of evaluation with respect to epistemic norms, standards, or prescriptions is not context-dependent.Footnote ² Facts about context can make evaluation with respect to the norms of Victorian etiquette as wholly inconsequential or irrelevant, all things considered, for my undergraduate. But facts about context can never make epistemic evaluation wholly inconsequential or irrelevant, all things considered, for a paradigmatic doxastic subject. At least for paradigmatic doxastic subjects, then, ordinary thought and talk presumes that the normative significance of epistemic evaluation is guaranteed.Footnote ³

Relatedly, ordinary thought and talk also suggest that the normative significance of epistemic evaluation is not dependent on, grounded in, or explained by appeal to any of the particular ends that any particular doxastic subject happens to have. Indeed, ordinary thought and talk treats facts about a subject’s own ends as straightforwardly immaterial or irrelevant to the question of why it matters, all things considered, that the subject conforms with epistemic norms. By way of illustration, the following strike us as rather dramatically off the mark: “You should base your beliefs about the effects of the practice of lynching on Black communities in the United States during the Jim Crow era on reliable sources because failing to do so risks compromising your ability to achieve your own ends.” or “That astronomy professor said that nuclear fusion in the sun’s core transforms hydrogen into helium, so you should believe that—and the reason is that believing the experts about what happens in the sun’s core will help you to achieve your ends.”

Perhaps even more telling, ordinary thought and talk rejects the possibility that disavowing, renouncing, or relinquishing any particular end can neutralize the normative significance of epistemic evaluation for a particular doxastic subject. Consider the following illustrative exchange: “I know this traffic is putting us a tiny bit behind schedule getting to the station, but don’t worry: the train will be at least a few minutes late” “you’ve got no evidence that the train will actually be late” “sure, but right now I don’t have any reason to care whether the train will actually be late—we can’t do anything about the traffic, so we have no way to get to the train station faster…” “What?”

It is difficult to interpret the last response here at face value, as expressing genuine belief. Instead, we are inclined to hear this last response as an invitation to join in the pretense that the train will be late as a kind of soothing fantasy, or perhaps as expressing some kind of hope or wish that the train will be late paired with an invitation to ignore, for the moment, the possibility that the relevant hope/wish will be unfulfilled.Footnote ⁴ Ordinary thought and talk resists interpreting this subject’s suggestion that believing the train will be late in no way frustrates, and indeed, furthers our ends as a genuine attempt to explain why it makes sense, all things considered, for us to believe that the train will be late. Thus, the fact that a subject’s particular ends are not, in the circumstances at hand, served by conformity with epistemic norms, standards, or prescriptions does not appear to rob conformity with epistemic norms, standards, or prescriptions of normative significance for the subject. And this suggests that the normative significance of epistemic evaluation is not (or at least not only) hypothetical.

Finally, ordinary thought and talk presume that epistemic evaluation is not (at least not always) decisive in settling what and how a subject ought to believe, all things considered. Rather, the degree to which conformity with epistemic norms, standards, or prescriptions matters, all things considered, seems to weigh against other things that matter. And, ordinary thought and talk certainly recognizes the possibility that a subject believes as they ought to believe, all things considered, and yet that the subject is epistemically criticizable for believing as they do.Footnote ⁵ Epistemic evaluation, then, seems to be but one piece of the puzzle, so to speak, when it comes to determine whether a subject has conducted the business of being a doxastic subject as they ought, all things considered.Footnote ⁶ Nevertheless, ordinary thought and talk suggests that other (non-epistemic) considerations can never make the normative significance of epistemic evaluation disappear, dissolve, or evaporate in the way that finding out one has taken a hallucinogenic can rob one’s perceptual experience as of a pink elephant of all evidentiary value. In this sense then, epistemic evaluation exhibits pro tanto normative significance, all things considered.

So, the datum for epistemic theorizing to explain now more clearly in view is a pretheoretical commitment to the normative significance of epistemic evaluation as inescapable, nonhypothetical, and pro tanto, at least for paradigmatic doxastic subjects.

3. The Instrumentalist Strategy

How might epistemic theorizing explain this datum? One explanatory strategy that holds significant theoretical appeal and has garnered significant philosophical attention is inspired by a familiar explanatory story we tell about when, how, and why conformity with garden-variety domain-specific norms, standards, and prescriptions comes to have normative significance.Footnote ⁷ The normative significance of, e.g. Victorian etiquette is plausibly instrumental: if it matters, all things considered, that S conform with a norm of Victorian etiquette in a particular context, C, this is because conforming with the relevant norm is among the necessary means S would need to take in order to achieve certain of her ends in C. Analogously, then, an instrumentalist explanation of the normative significance of epistemic evaluation maintains that the normative significance of epistemic evaluation ultimately derives from the normative significance of certain of the ends that doxastic subjects actually have. And, accordingly, an instrumentalist’s explanation of our pretheoretical commitment here comes in two steps. The first identifies certain ends that doxastic subjects (at least paradigmatic doxastic subject) in fact have. And the second shows that conformity with epistemic norms, standards, or prescriptions constitutes a means for achieving these ends.

Why is the instrumentalist strategy appealing? Well, for the instrumentalist, epistemic evaluation inherits normative significance from a more explanatorily fundamental source. But, crucially, an instrumentalist is not in the business of explaining why or how or in virtue of what this source, itself, has normative significance. This makes an instrumentalist explanation a kind of buck-passing explanation. And any buck-passing explanation will be, at least in one sense, partial or incomplete: it will explain why and how epistemic evaluation matters, all things considered, by taking for granted that something else matters, all things considered. Nevertheless, a buck-passing explanation can be perfectly theoretically respectable and, indeed, significantly illuminating just in case the explanation (i) identifies an independently plausible proximate source of normative significance and (ii) offers an independently plausible model of inheritance. And, indeed, the instrumentalist strategy is exceptionally well-positioned on both these fronts. The instrumentalist locates the relevant proximate source of normative significance in the practical sphere: this proximate source is (certain of) the (reasonably adopted) ends doxastic subjects have. But, of course, the thesis that a subject’s (reasonably adopted) ends often have normative significance is exceedingly plausible on independent grounds.Footnote ⁸ Moreover, the instrumentalist supplies an independently plausible account of inheritance: the instrumentalist’s model of inheritance is just an application of a kind of general story that we already accept about the way the normative significance of ends transmits to means that would secure those ends. In essence, the instrumentalist strategy promises to explain the normative significance of epistemic evaluation by subsuming the way in which conformity with epistemic norms, standards, and prescriptions comes to matter, all things considered, under an exceedingly familiar explanatory model that we already rely on elsewhere. And this is a significant theoretical virtue.

Painting with a broad brush, at least two, distinct approaches to implementing the instrumentalist strategy have received attention in contemporary discussions. But, neither one is capable of delivering a thoroughgoing vindication of the kind of pretheoretical commitment to the normative significance of epistemic evaluation made manifest in ordinary thought and talk.

According to the first approach, epistemic evaluation inherits normative significance from the idiosyncratic ends each doxastic subject has, whatever these ends happen to be.Footnote ⁹ In order to vindicate our pretheoretical commitment to inescapability, however, this approach must show that conforming with epistemic norms, standards, or prescriptions turns out to be the most effective (or, perhaps, necessary) means to realizing any potential set of ends that a paradigmatic doxastic subject might happen to have. And this looks to be a very tall order, indeed.

According to the second approach to implementing the instrumentalist strategy, the normative significance of epistemic evaluation derives from distinctively epistemic end(s) (e.g. believing truth and avoiding error) that doxastic subjects (or, at least paradigmatic doxastic subjects) in fact, have. To vindicate our pretheoretical commitment to inescapability, the instrumentalist who favors this approach will need to identify an end for which conformity with epistemic norms, standards, or prescriptions plausible constitutes a means that, plausibly, all (paradigmatic) doxastic subjects always have. And, again, this looks to be a very tall order.Footnote ¹⁰

It seems, then, that both these approaches will struggle, albeit for different reasons, to vindicate the apparent inescapability of the normative significance of epistemic evaluation.Footnote ¹¹ But, more than this, it should be obvious that neither approach can vindicate the apparently non-hypothetical character of the normative significance of epistemic evaluation. The instrumentalist strategy is, after all, committed to identifying explanatory grounds for the normative significance of epistemic evaluation among the ends that doxastic subjects actually have.

Of course, it is not as if epistemic theorizing is required to provide a thoroughgoing vindication of the pretheoretical commitments made manifest in ordinary thought and talk. I’ve assumed that epistemic theorizing ought to explain our pretheoretical commitment to the normative significance of epistemic evaluation. And a vindicating explanation of our pretheoretical commitment here certainly achieves this much. But alternatively, epistemic theorizing might explain our pretheoretical commitment to the normative significance of epistemic evaluation by telling an error-theoretic story that diagnoses (aspects of) our pretheoretical commitment as manifesting a kind of unmysterious, understandable, and even predictable mistake or confusion. Indeed, the epistemologist might develop the instrumentalist strategy with this alternative sort of partially error-theoretic explanation as the explicit goal or accepted result. At present, however, I wish simply to explore the prospects for a full theoretical vindication of our pretheoretical commitment to the normative significance of epistemic evaluation: what, if any, sort of potentially viable vindicating explanation of this datum can epistemic theorizing generate? Accordingly, I leave two important tasks—the evaluation of the theoretical virtues of candidate error-theoretic explanations and the comparative assessment of candidate error-theoretic explanations versus candidate vindicating explanations—for a subsequent occasion.

4. The Constitutivist Strategy

Instead, I’ll turn my attention to an alternate explanatory strategy that may seem, at least at first pass, better positioned to fully vindicate our pretheoretical commitment to the normative significance of epistemic evaluation.

Like the instrumentalist, the constitutivist looks to subsume the way in which conformity with epistemic norms, standards, and prescriptions comes to matter, all things considered, under an already-familiar explanatory model. But for the constitutivist, the relevant explanatory model is not the instrumentalist’s favored means-ends model. Rather, it is the explanatory model to which we appeal in explaining the normative significance of domain-specific norms, standards, or prescriptions that help to constitute certain roles, activities, and performances as the sorts of roles, activities, and performances that they are. Evaluation with respect to these sorts of constitutive norms, standards, or prescriptions derives whatever normative significance they have from the normative significance of occupying the role, engaging in the activity, or executing the performance that they help constitute. Put differently, these sorts of constitutive norms, standards, and prescriptions bear on how a subject ought to be and on what a subject ought to do if, when, and because occupying the relevant role, engaging in the relevant activity, or executing the relevant performance itself has normative significance. Analogously, then, the constitutivist proposes that epistemic norms, standards, or prescriptions help to constitute the business of being a doxastic subject as the sort of business that it is. And, accordingly, epistemic evaluation derives normative significance from the normative significance of this business going well, on its own terms. Conformity with epistemic norms, standards, or prescriptions matters, all things considered, if, when, and because being a doxastic subject (or, alternatively, engaging in the activities and executing the performances that constitute the business of being a doxastic subject) itself has normative significance.Footnote ¹²

We deploy the constitutivist’s favored explanatory model to make sense of when, why, and how evaluation with respect to an impressively wide variety of mundane domain-specific norms, standards, or prescriptions come to have normative significance. And, indeed, some of the appeal of the constitutivist strategy is rooted in precisely this fact. But, of course, evaluations with respect to most mundane domain-specific constitutive norms, standards, or prescriptions do not seem to exhibit inescapable normative significance. On the contrary, the normative significance of occupying a specific role, engaging in a particular activity, or executing a specific performance seems to be standardly context-dependent. Whether and how, e.g., the rules of chess bear on what a subject ought, all things considered, to do seems to depend on the contingent or idiosyncratic features of the subject’s circumstances. In a context where it is, all things considered, of no consequence whatsoever that a subject (continue to) play chess—or, indeed, if the subject has overwhelming reason to withdraw or abstain from playing chess—then conformity with the rules of chess quite plausibly lacks normative significance entirely. Of course, even in such a context, we can still coherently evaluate certain ways of moving the pieces around the board as chess legal and others as violations of the rules of the game. The point is these evaluations do not always matter, all things considered.Footnote ¹³ Engaging in the activity of playing chess has normative significance in some contexts and not at all in other. And it is the normative significance of playing chess for a particular subject, given their circumstances, that determines both whether and how the chess legal status of any way of moving a chess piece around a board bears on the question of how the subject ought, all things considered, to behave.

By way of illustration: imagine that, in the middle of our chess match, my opponent is stung by a bee and has the sort of life-threatening anaphylactic reaction that requires immediate medical intervention. In this scenario, what it would take for me to execute a chess legal move given the arrangement of pieces on the board in this moment simply does not matter at all, all things considered. Alternatively, imagine I find myself engaged in a chess match with a companion whom I know to be an extremely sore loser. If I know that I am going to have to spend the rest of what will be an extremely unpleasant afternoon with this companion should they lose the match, and I realize that the match has progressed to this point in a way that now makes my companion’s loss inevitable should we continue playing, then perhaps the all-things-considered best path forward is for me to feign chess-incompetence by executing an illegal move as a means of gracefully exiting the game without victory. Or, imagine that I am demonstrating illegal chess moves in the context of explaining to a friend how the game of chess is played. The fact that advancing my rook one single square along a diagonal violates a constitutive norm of chess does not guarantee that I have any reason at all to avoid executing precisely this move. Indeed, quite the opposite seems to be true: the fact that moving the rook piece in this way violates a rule of chess might give me some reason to execute precisely this move. In these scenarios, it is not that the normative significance of an evaluation of the legal status of a potential move gets outweighed or overridden by other considerations. Rather, in these scenarios, an evaluation of the legal status of a potential move is just entirely irrelevant to answering the question of how I ought, all things considered, to proceed. There is no sense in which, from the all things considered perspective, my failure to execute a chess-legal move in any of these scenarios is a regrettable, but justified cost of doing what I can to help save my opponent’s life or avoid an unpleasant afternoon or teach a friend how the game of chess is played. Rather, my failure to execute a chess-legal move is no cost at all. The rules of chess here have no normative significance whatsoever.

This much does not yet show that the constitutivist strategy is unable to vindicate our pretheoretical commitment to the inescapable normative significance of epistemic evaluation, at least for paradigmatic doxastic subjects.Footnote ¹⁴ But it does show that if the constitutivist is to vindicate our pretheoretical commitment here, then the constitutivist will need to provide grounds for thinking that the business of being a doxastic subject and, e.g., playing chess are disanalogous in at least the following crucial respect. An account of what it takes to play chess will always require substantive supplementation in order to make any kind of explanatory progress illuminating the normative significance of conforming with the rules of chess for a particular subject in a particular context. But, in contrast, a plausible account of the business of being a doxastic subject can, without supplementation, make significant explanatory progress illuminating the normative significance of epistemic evaluation for paradigmatic doxastic subjects.Footnote ¹⁵ Put differently, the constitutivist about the normative significance of epistemic evaluation must show that the nature of the business of being a (paradigmatic) doxastic subject can, on its own, supply a genuinely illuminating (even if not complete) explanation of the normative significance of epistemic evaluation. The constitutivist is committed to deriving the normative significance of epistemic evaluation from the normative significance of engaging in the business of being a doxastic subject. Thus, vindicating the inescapable character of the normative significance of epistemic evaluation requires that the constitutivist tell a story about the nature of the business of being a doxastic subject, which guarantees that engaging in this business has normative significance for all paradigmatic doxastic subjects. And if the constitutivist’s story here is to be genuinely explanatory (rather than merely stipulative), then the constitutivist must show that understanding the nature of the role of doxastic subject and/or of the activities or performances that characterize the business of being a doxastic subject helps us better understand, not merely that, but how and why occupying this role and/or engaging in these activities or performances comes to have normative significance for all paradigmatic doxastic subjects.

But, alas, this is a burden that familiar implementations of the constitutivist strategy in epistemology have struggled to meet. On perhaps the most familiar way of implementing the constitutivist strategy, to be a doxastic subject (or, alternatively, to engage in the activities and execute the performances that characterize the business of being a doxastic subject) is, constitutively, not merely to be assessable with respect to epistemic norms, standards, or prescriptions, but further to be inescapably committed to and/or bound the all things considered significance of these norms, standards, or prescriptions. On this sort of constitutivist proposal, it is an explanatorily fundamental fact about the nature of doxastic subjects (or, alternatively, of the activities or performances that characterize the business of being a doxastic subject) that, for any doxastic subject, conformity with epistemic norms, standards, or prescriptions really does matter, all things considered. The normative significance of epistemic evaluation here is built into the nature of being a doxastic subject (or, alternatively, to engage in the activities and execute the performances that characterize the business of being a doxastic subject) at the ground level, so to speak, as an intrinsic feature thereof. It neither requires nor admits of further explanation. But this sort of move is, in effect, a bald rejection of the demand to explain how and why constitutive facts about the business of being doxastic subject are sufficient to guarantee the normative significance of epistemic evaluation. And it is reasonable, I think, to worry that a bald rejection of this explanatory burden is dialectically unjustifiable.Footnote ^16, Footnote ¹⁷ Moreover, I think the constitutivist strategy can be implemented in a way that positions the constitutivist to simply accept and then discharge this explanatory burden.

5. How to Implement the Constitutivist Strategy: Developing an Explanatory Story

The key, I think, to implementing the constitutivist strategy in a way that accepts and discharges this explanatory burden, is to embrace a certain sort of metaphysical picture according to which the business of being a doxastic subject is constitutively connected to the business of agency. And my aim in this section is to sketch a constitutivist explanation of the normative significance of epistemic evaluation by developing and exploiting a metaphysical picture of precisely this sort.Footnote ¹⁸ The sort of constitutivist explanation of the normative significance of epistemic evaluation I aim to bring into view here involves two moves. The first move involves embracing a thesis about the normative significance of acting, as I’ll say, well, over time and across context. The second move involves embracing an account of the business of being a doxastic subject according to which this business exhibits a particular sort of constitutive connection to the capacity to act well over time and across contexts. Let me spell out each of these moves in turn.

The constitutivist’s explanatory story takes as a point of departure a familiar feature of our agency. The feature is this: from the perspective of what really and truly matters, all things considered, agents like us genuinely can, as I’ll say, act more or less well. Footnote ¹⁹ Put differently, how we carry out the business of agency does, we think, have normative significance; it really does seem to matter, all things considered, that we exercise our agency in certain ways and not others. Both in ordinary thought and talk and in philosophical theorizing, this much is an exceedingly familiar and widely accepted feature of the sort of agency that we familiarly exercise.Footnote ²⁰

What, exactly, does it take to act well—that is to act in ways that one ought, all things considered, and to refrain from acting in ways that one ought, all things considered, to refrain from acting? And why does it ever matter, all things considered, how we act? Why is it that our actions are subject to any normatively significant evaluation at all? These are important questions that deserve the philosophical attention they have received and continue to receive in ethics and metaethics. But it is perfectly reasonable for an epistemologist who aims to vindicate our pretheoretical commitment to the normative significance of epistemic evaluation to aim to remain as ecumenical as possible here. And, in any case, these are not questions to which an explanation of the normative significance of epistemic evaluation owes an immediate answer. An albeit partial, but nevertheless genuinely illuminating explanatory story that spells out why and how the normative significance of how we exercise our agency can serve as a proximate explanatory ground for the normative significance of epistemic evaluation would, on its own, achieve significant explanatory progress.

It is independently plausible that the business of agency is (for whatever reason) subject to genuinely normative evaluations (whatever these turn out to be). But, more than this, the business of agency is, at least paradigmatically, a diachronic business. And it is independently plausible that we appropriately evaluate it as such. Thus, the extent to which paradigmatic agents manage to act well quite plausibly matters, not (only) in a one-off kind of way, but cumulatively, over time and across contexts. For diachronic agents, acting now in a way that risks damaging or undermining one’s capacity to act well later is, it seems, always regrettable or at least non-ideal, all things considered (and even if the action the agent performs now is the action that the agent ought, all things considered, to perform given their particular circumstances). When an agent’s capacity to act well over time and across contexts is somehow compromised, this much is never thoroughly inconsequential, all things considered. Evaluations of this sort of capacity to act well over time and across contexts are never appropriately dismissed with complete and utter indifference. So, at least for diachronic agents, the capacity to act well over time and across contexts inescapably matters, all things considered. Evaluations assessing the strength, health, or robustness of this capacity have inescapable normative significance.Footnote ²¹ The first move of the explanatory story coming into view here is to identify this feature of the sort of diachronic agency that we familiarly exercise as the proximate source of the normative significance of epistemic evaluation.

But, of course, the constitutivist still owes an account of inheritance or transmission that explains how the normative significance of epistemic evaluation for paradigmatic doxastic subjects-sum-agents derives from the normative significance of cultivating and sustaining a capacity to act well over time and across contexts. And, thus, the second major move of the constitutivist’s explanatory story here aims to supply such an account by embracing a metaphysical picture according to which the business of being a paradigmatic doxastic subject and the business of diachronic agency are constitutively connected in a particular way. According to this picture, the business of being a doxastic subject and the business of diachronic agency paradigmatically come as a particular sort of metaphysical package. At least for the sorts of doxastic subjects-cum-agents that our ordinary thought and talk regards as paradigmatic, the business of being a doxastic subject is properly understood as a constitutive component part of the business of diachronic agency.

The particular species of constitutive connection between the business of being a paradigmatic doxastic subject and the business of being a diachronic agent to which the constitutivist appeals here is not unfamiliar. It is the species of constitutive connection that characterizes the relationship between a complex skill such as, e.g., playing the violin and its various constituent subskills such as, e.g. fingering or bowing. It is also the species of constitutive connection that characterizes the relationship between a complex mechanical system such as a standard car engine and its various constituent subsystems such as, e.g., the fuel injectors or spark plugs. Consequently, perhaps the easiest way to bring this species of constitutive connection into sharp focus is via appeal to the more mundane. To that end, let me take a moment to work through yet a third example in greater detail: the way in which medicine conceives the business of insulin production within the human pancreas in response to the evolving conditions within the human body that a functional analysis of pancreas’s role in regulating blood sugar levels yields. The business of insulin production within the human pancreas in response to the evolving conditions within the human body subserves the business of regulating blood sugar levels. The system responsible for insulin production in response to evolving conditions in the body instantiated in the human pancreas operates as one functionally-specified component of a more complex functionally organized system that is for maintaining healthy blood sugar levels. And, indeed, the sort of insulin-production system we find in the human pancreas is constituted as the particular sort of insulin production system that it is by its distinctive functional profile within this more complex functional system. Accordingly, this business of insulin production goes well qua the kind of business that it is precisely when the pancreas produces insulin at rates that are well-suited (on the assumption that other components of this functionally organized system are operating as the paradigmatically do) to play a specific role in helping maintain appropriate blood sugar levels as conditions in the body change over time. Put differently, what, exactly, it takes to conform with the particular functional norms governing the system responsible for insulin production instantiated within the human pancreas is determined by reference to the demands of the specific role played by the insulin-production system operating as one component of a more complex functionally organized system that is responsible for maintaining appropriate blood sugar levels in the human body. And, in turn, this more complex functionally organized system’s capacity to maintain appropriate blood sugar levels reliably over time as conditions change is partially constituted by well-functioning insulin production in the pancreas. The insulin-production system instantiated in the human pancreas and the more complex functional system for regulating blood sugar levels in the human body of which the former is a functional component are a particular kind of metaphysical package: the particular sort of capacity for insulin-production in response to evolving conditions instantiated in the well-functioning human pancreas both helps to constitute and is constituted by the particular sort of capacity for blood sugar level regulation instantiated in a healthy human body.

Analogously, then, the proposal here is to embrace an account of the business of being a doxastic subject that conceives of this business as embedded as one constitutive component, operating alongside other constitutive components, each of which will have a distinct profile, in a more complex system that is for the sort of diachronic agency in which we paradigmatically engage. Accordingly, constitutive norms that spell out what it takes for the business of being a doxastic subject to go well qua the sort of business that it is are just norms that spell out what it takes for the business of being a doxastic subject to equip paradigmatic diachronic agents with the sort of first person, subjective perspective on the facts that is well-suited to play one particular sort of distinctive role or make one particular sort of contribution in the service of acting well over time and across contexts. The capacity to act well over time and across contexts is partially constituted by the business of holding and regulating a corpus of doxastic states constituting a first person, subjective perspective on the way things are.Footnote ²² And it is the distinctive way in which the business of being a doxastic subject is embedded as a constitutive component of the capacity to act well over time and across contexts that determines what it takes for this business to go well on its own terms as the distinctive sort of business that it is. What conformity with the constitutive norms, standards, and prescriptions that govern the business of being a doxastic subject requires is fixed by the particular demands that the sort of diachronic agency we familiarly exercise makes on this business.

With both major moves on the table, the constitutivist’s explanatory story comes into view: For paradigmatic doxastic subjects-cum-agents, maintaining the capacity to act well over time and across contexts is a matter of normative significance. But what it takes to conduct the business of being a doxastic subject well is (partially) constitutive of and constituted by the distinctive contribution to the capacity to act well over time and across contexts that the business of being a doxastic subject paradigmatically makes in the context of this capacity. Thus, at least for paradigmatic doxastic subjects-cum-agents, conducting the business of being a doxastic subject well is a matter of normative significance.

This sort of explanatory story is well-positioned to vindicate inescapability. After all, for paradigmatic doxastic subjects-cum-agents, the normative significance of the capacity to act well reliably over time and across contexts is quite plausibly inescapable. And as such, assessments of any kind of excellence or shortcoming with respect to this capacity always matter, all things considered. But if the business of being a doxastic subject is constitutively connected to the capacity to act well over time and across contexts in that way that the constitutivist has proposed here, then epistemic evaluations are just one class of assessments of excellence or shortcoming with respect to this capacity. Thus, at least for paradigmatic doxastic subjects, the normative significance of epistemic evaluation is inescapable.Footnote ²³

Crucially, assessments of excellence or shortcoming with respect to this capacity to act well reliably over time and across contexts assessments are normatively significant simply in virtue of the fact that the sort of diachronic agency that we familiarly exercise can go more or less well, all things considered. But it seems implausible that acting well—that is as one ought, all things considered—is merely a matter of achieving the ends that one happens to have. Thus, if, as the explanatory story on offer here proposes, the normative significance of epistemic evaluation derives from the normative significance of the capacity to act well reliably over time and across contexts, then the normative significance of epistemic evaluation is not straightforwardly hypothetical.

Finally, the explanatory story on offer here can vindicate the idea that the normative significance of epistemic evaluation is always pro tanto. For paradigmatic doxastic subjects-cum-agents, cultivating and sustaining the capacity to act well over time and across contexts matters. But, of course, this capacity is not the only thing that matters. And if/when the normative significance of this capacity is outweighed or overridden by competing normatively significant considerations, these competing considerations will thereby outweigh or override the normative significance of epistemic evaluation. Still, if the capacity to act well over time and across context is inescapably a matter of some (potentially outweighed or overridden) normative significance for S, whether S falls short epistemically is never a matter of complete inconsequence, all things considered.

6. Clarifications

On the vindication of our pretheoretical commitment to the normative significance of epistemic evaluation that I’ve proposed, the proximate explanatory grounds for the significance of epistemic evaluation are among the constitutive features of the business of being a doxastic subject. And thus, at least at a certain level of abstraction, what I’ve proposed shares a great deal with more familiar implementations of the constitutivist strategy. First, they share a commitment regarding where the explanatory story that epistemic theorizing provides about when, why, and how conforming with epistemic norms matters, all things considered, will bottom out. The shared commitment here is a commitment to locating explanatory grounds for the normative significance of epistemic evaluation among the constitutive features of the business of being a doxastic subject. Relatedly, the proposal on offer here and familiar implementations of the constitutivist strategy share a commitment to subsuming the normative significance of epistemic evaluation under a familiar explanatory model that we readily deploy to account for the normative significance of a wide variety of mundane species of evaluation: evaluations with respect to constitutive norms, standards, and/or prescriptions derive normative significance from the normative significance of the role, the activity, or performance that these norms help to constitute. Thus, the explanatory strategy I’ve proposed here is aptly characterized as a version of constitutivism.

But familiar implementations of the constitutivist strategy are often more ambitious than the sort of explanation I’ve developed here along at least two distinct dimensions. First, at least some implementations of the constitutivist strategy aim to supply a complete explanation of the normative significance of epistemic evaluation. Unlike what I’ve proposed here, these implementations of the constitutivist strategy purport to identify ultimate (rather than merely proximate) explanatory grounds for the normative significance of epistemic evaluation. Second, certain familiar implementations of the constitutivist strategy explicitly seek to preserve the independence of the normative significance of epistemic evaluation from the normative significance of any sort of (broadly speaking) practical evaluation. Indeed, the goal here seems to be to account for the normative significance of epistemic evaluation as wholly sui generis . In contrast, the sort of explanatory story that I’ve proposed here is straightforwardly buck-passing. Moreover, this explanatory story grounds the normative significance of epistemic evaluation in the practical sphere: epistemic evaluations are normatively significant for doxastic subjects like us because how subjects like us act over time and across contexts is normatively significant.

Instrumentalists, of course, also embrace a buck-passing explanation of the normative significance of epistemic evaluation that “passes the buck” to the practical sphere. And so, in these two crucial respects, the sort of explanatory strategy I’ve pursued here aligns with certain of the core commitments of the instrumentalist strategy. But the sort of explanation I’ve developed here is not aptly characterized as instrumentalist for two reasons: it does not appeal to the ends doxastic subjects-cum-agents actually have, and it does not appeal to the transmission of normative significance from ends to means.

The sort of explanatory story I’ve developed here is aptly characterized as a form of pragmatism about the epistemic. After all, this sort of story identifies the proximate source of the normative significance of epistemic evaluation as the normative significance of evaluation with respect to whatever norms, standards, or prescriptions characterize what it takes for an agent to have the capacity to act well over time and across contexts. But, crucially, what I’ve proposed here is not the sort of pragmatism that simply collapses or reduces normative significance of epistemic evaluation to the normative significance of any other familiar kind of (broadly speaking) practical evaluation.Footnote ²⁴ The form of pragmatism on offer here is not one according to which it matters that a subject conform with epistemic norms, standards, or prescriptions because it matters that the subject regulates their doxastic states in ways that, in fact, given the subject’s present circumstances, conduce to actually acting well. Rather, the sort of explanatory story I’ve developed here proposes a significantly more complex derivation from the normative significance of acting well to the normative significance of conducting the business of being a doxastic subject well.

It is worth highlighting that the sort of explanatory story I’ve developed here has implications for first-order epistemic theorizing that aims to spell out the contents of epistemic norms, standards, and/or prescriptions. In particular, this sort of story commits us to a certain way of spelling out what conformity with first-order epistemic norms, standards, or prescriptions actually involves. In particular, this sort of explanatory story positions the business of being a doxastic subject within a more complex structure or system that subserves the constitutive goal or aim of acting well over time and across contexts. Accordingly, this story maintains that what it takes to act well plays a crucial role in helping to fix what it takes to conduct the business of being a doxastic subject well. Compare: What it takes for the human pancreas to release insulin well is just a matter of what it takes for the pancreas to release insulin in ways that equip the human endocrine system (which the pancreas is one component) to regulate blood sugar levels in the way that it paradigmatically does on the assumption that other parts of the endocrine system function properly. Analogously, then, what it takes to conduct the business of being a doxastic subject well is a matter of what it takes to conduct the business of being a doxastic subject in ways that sustain the doxastic subject-cum-agent’s capacity to act well reliably, over time and across contexts, on the assumption that other parts of the “agential system” operate properly. It is an important question—but one the exploration of which I must leave for another occasion—how revisionary the resulting functional account of the contents of epistemic norms, standards, and/or prescription turns out to be.

Finally, embracing the sort of explanatory story I propose here does not preclude also embracing the possibility that epistemic evaluation sometimes derives additional normative significance from other sources. For example, in certain contexts and for certain subjects, a fairly standard instrumentalist story might apply: imagine one of my ends is to identify and purchase the best-value toaster on the market. It might be that conforming with epistemic norms, standards, or prescriptions as I conduct my market research is, for me, the best means to this end. If so, then there is no reason for the proponent of the sort of explanatory story I propose here to deny that (at least when it comes to the topic of toasters) conformity with epistemic norms, standards, or prescriptions also gains instrumental normative significance for me in this context. Indeed, the proponent of the sort of explanatory story I propose here can even allow that holding and regulating doxastic states about a particular topic might, at least in certain contexts and for certain subjects, matter non-derivatively and for its own sake.