Guidance and mainstream epistemology

Abstract

According to one prominent critique of mainstream epistemology, discoveries about what it takes to know or justifiedly believe that p can’t provide the right kind of intellectual guidance. As Mark Webb puts it, “the kinds of principles that are developed in this tradition are of no use in helping people in their ordinary epistemic practices.” In this paper I defend a certain form of traditional epistemology against this “regulative” critique. Traditional epistemology can provide—and, indeed, can be essential for—intellectual guidance. The reason is that, in many cases, how you should proceed intellectually depends on what you already know or justifiedly believe: how you should treat counterevidence to your beliefs, for example, can depend on whether those beliefs count as knowledge. Therefore, to get guidance on how to proceed intellectually, it will often be essential to be able to figure out what you know or justifiedly believe. And to do that it will often be helpful to try to figure out what it takes to count as knowledge or justified belief in the first place. To do this is precisely to engage in mainstream epistemology.

If you ask a professional epistemologist for help figuring out what to believe about some contentious issue—say, whether mask mandates help reduce the transmission of COVID-19—one book you probably won’t want them to hand you is Roderick Chisholm’s Theory of Knowledge. This is so, despite the fact that Chisholm himself took the purpose of traditional epistemology to be

to correct and improve our own epistemic situation... to do our best to improve our set of beliefs—to replace those that are unjustified by others that are justified and to replace those that have a lesser degree of justification with others that have a greater degree of justification. (1989, 1)

These look like the words of a philosopher concerned with intellectual guidance. But even passing familiarity with Chisholm’s work makes clear that Theory of Knowledge isn’t a particularly helpful instruction manual:

We have taken “at least as justified as” as an undefined locution. Obviously, we have to take some locution as undefined. But the fact that we have not defined it does not mean that we cannot say anything about what is intended by it. If we set forth certain axioms for the locution, we can illuminate just what it is that we intend to express by it… (1989, 12) We note, finally, that our undefined epistemic concept and the axioms that may be provided for it enable us to set forth a hierarchy of epistemic concepts. This hierarchy involves 13 epistemic categories—13 steps or stages each capable of being occupied by countless propositions. (1989, 15)

The reader will be perhaps be grateful that I am not rehearsing Chisholm’s axioms or 13 stages at length here.¹ It’s hard to see how reflection on those axioms or stages could provide much help in day-to-day intellectual activity, or even when untangling the knottier of life’s problems. Such, at least, is the contention of a recent critique of traditional epistemology: traditional epistemology, to its detriment, can’t guide intellectual conduct.

I defend a form of traditional epistemology against this critique; traditional epistemology can provide intellectual guidance. In Sect. 2, I spell out the traditional view I defend. I elaborate the critique of traditional epistemology in Sect. 3. The positive defense of the guidance-potential of traditional epistemology comes in Sects. 4, 5, and 6. I close in Sect. 7 by responding to an objection.

But I start with some general reflection on the nature of guidance.

1 Helpful and unhelpful advice

You’re filling up your bicycle tire. How do you figure out when it’s full? Fortunately, bicycle tires stamp the answer on the tire itself. That alone won’t tell you when to stop inflating, because it won’t tell you when your tire has reached that pressure. So you’ll want an air pressure gauge. But the pressure gauge won’t help unless you know what pressure counts for your tire as full.

You don’t only want guidance when filling up your tire. You want guidance on lots of things, from the mundane to the momentous: when to take out the garbage, when to remove the hard boiled egg from the water, when to have a child, what to wear to an interview, what verdict to issue in a trial, whether to check that the stove is off, whether to respond to someone’s social media post, how much credence to give that social media post, and what to believe about the efficacy of mask mandates.

We epistemologists are interested in how we might provide guidance on those latter questions. For example, one hope is that we can provide what Jane Friedman (2020) has called ‘zetetic norms,’ which are norms governing the conduct of inquiry. One kind of zetetic norm might govern when it is proper to stop inquiring into some matter. If we’re wondering when we can stop inquiring, it would be helpful to know what the zetetic norm is. Of course, in one way the answer is easy: you should stop inquiring when inquiry is finished. But this isn’t any more useful than the analogous answer to the question when to stop inflating your tire: when it’s full. You want to be told when the tire is full. As noted, tires come with a helpful stamp. Inquiry does not.

Some epistemologists have advice about this. Friedman (2017), for example, argues that,

If one knows the answer to some question at some time then one ought not to be investigating that question, or inquiring into it further or wondering about it, or curious about it, and so on, at that time… (310)

Christoph Kelp (2021, 10), Peter van Elswyk and Yasha Sapir (2021, 5838), Dennis Whitcomb (2017, 152), and Alan Millar (2011, 63), among others, make similar recommendations. These epistemologists argue that inquiry into some question is finished when you know the answer.

If true, this advice is helpful only if you can recognize when you know the answer. This again makes the advice about as helpful as the easy answer to the question when to stop inflating the tire: when it’s full. That advice is also helpful only if you can recognize when the tire is full. Fortunately, you often can, even if you don’t have a pressure gauge. As long as you have a general sense of what a full tire feels like—not too squishy—then you’ll be able to figure out whether the tire is full and so will get guidance on when to stop filling it up. Likewise with the advice about inquiry. You might not know the details of what knowledge requires, but you might have a general sense of what knowledge feels like: not too squishy. Sometimes that might be good enough.²

But as noted in the magazine Mountain Bike Action, “pressing your thumbs against the tire’s sidewall may result in a misleading psi measurement. It’s better to trust your pump’s gauge or invest in a tire pressure tool.” (2011, 24) The same goes for judging when to stop inquiry: sometimes you’ll want more than the squishy conception. If the stakes are very high, or you’re in doubt or there is some disagreement about whether you know, you’ll want more principled grounds for concluding, before you end inquiry, that you know the answer. But how can you have that assurance without a more detailed conception of what knowledge requires—the kind of conception of what a full tire consists in that’s provided by the stamp on the tire? In the normal course of inquiry, there is no such stamp.³ If only there were a discipline whose job it is to investigate what knowledge requires!

Before getting to that, let’s look at what kinds of answers to the question of what knowledge requires might be helpful, assuming that knowledge properly ends inquiry. We can make headway by looking at other instances in which you might seek guidance. I’ll start with mushrooms. Which mushroom should you pick? Here’s an unhelpful answer: the non-poisonous one. You want to know what is required for a mushroom to be poisonous. But there are lots of different kinds of answers we can give to the question. Here’s Merriam-Webster’s definitional answer:

X is poisonous iff X “causes injury or death when absorbed or ingested.”

Definitions can helpfully distinguish different aspects of an issue and so can help you localize your investigations. In legal theory it is common to define a criminal action as at least a conjunction of actus reus and mens rea. This division into components is helpful because it allows investigators to focus their burden of proof into manageable bits; “O.k., first we need to prove that a criminal act occurred. Then we need to prove there was a guilty mind.”

However, at least when confronted with the question of what mushrooms to pick, the definition of ‘poisonous’ won’t be what you’re looking for.⁴ You want to know how to distinguish mushrooms that satisfy that definition from mushrooms that don’t. For example, helpful guidance might include surface features of the mushroom, as indicated here:

“Never eat any mushroom unless you know what it is.” (Hallen, 2015)

Whoops! Not that one! That’s back to epistemology. I meant this one:

“You have to know the exact species/genus of each mushroom you want to eat.” (Still, 2022)

Wait! Ignore that one too. Here it is:

• 1. Avoid mushrooms with white gills, a skirt or ring on the stem and a bulbous or sack like base called a volva…

• 2. Avoid mushrooms with red on the cap or stem…

• 3. Finally don’t consume any mushrooms unless you are 100% sure of what they are. I know I have already mentioned this but it is by far the most important rule. (Biggane, 2014)

That’s the one. Don’t look at rule #3. Just the others. One way to get guidance on whether to pick a mushroom is by being told what poisonous mushrooms look like.

In some cases, though, you’ll want to find out more than what mere signs characterize some relevant feature or its absence.⁵ The chemical composition of the poison is not a mere sign of poisonousness; it constitutes the poison. If you can directly measure chemical composition, you’ll be in pretty good shape with respect to whether to pick the mushroom, as long as you know what chemical compositions are poisonous. Likewise, pounds-per-square-inch (psi) is not a mere sign of a tire’s fullness. The psi measured in a tire constitutes the fullness of the tire. That’s why it’s helpful to measure the psi of the tire and have a stamp that tells you what psi counts as full.

Sometimes the tire pressure should deviate from that specified by the stamp; the stamp is just a rough guide. Mountain Bike Action notes that “Numerous factors have to be taken into account when selecting tire pressure, including rider weight, trail conditions and tire size” (Unknown, 2011, 24) and provides a chart and notes detailing appropriate psi by rider weight, tire size and position, terrain, and experience level (Fig. 1).

Fig. 1

Recommended PSI by rider weight and tire size

This chart and notes can be applied by many bikers in many conditions. If you’re wondering when to stop inflating your tires, it will be helpful to have a chart like this no matter who you are or what kind of bike you are riding or where or when you’re riding it or what level of cyclist you are. You might, for example, be unsure whether the one-size-fits-all recommendation stamped on the tire is good for you. This chart tells you the answer (unless you weigh more than 230 pounds or less than 100).

The values on the chart aren’t arbitrary; they are grounded by a unifying explanatory principle: “appropriate tubeless tire pressure is essential for getting the most out of your ride” (24, emphasis added). This evaluative feature is a function of other features that “increase performance”: when your tires “grip well,” don’t “ping off rocks” or “provide a harsh ride,” “burp,” or “wreak havoc on your rims,” etc. (24).

So, to give guidance on a question about what to do, it’s good to provide principles that ground functions from discoverable states (like surface mushroom features or measurable psi) to actions (like picking a mushroom or stopping inflation). And it’s good to generate universal results that anyone can apply regardless of who they are and what conditions they’re in.

2 Good Ol’ fashioned S-p epistemology

These considerations help defend a kind of traditional epistemology against the critique that traditional epistemology can’t guide intellectual conduct. Call the kind of traditional epistemology I defend “good ol’ fashioned S-p epistemology” (SPE, for short).⁶ Much SPE concerns what it is for some subject, S, to know that p. But SPE can be about other attitudes and states as well: justified belief, understanding, or wisdom.

SPE is a practice distinguished by features of the theory that the practice aims at. The theory the practice aims at has three primary features:

• 1. It’s principled in that it is composed of general truths about what it takes to have some epistemic attitude or be in some epistemic state (like knowledge, justified belief, or understanding).

• 2. It’s theoretical in that the principles’ truth is independent of how we actually conduct ourselves intellectually.

• 3. It’s context- and identity-neutral in that some of the principles apply to everyone regardless of their identity or context and that these principles ground any other difference in what it takes for two subjects to have the relevant epistemic attitude or be in the relevant state.

The principles are supposed to be general, in that they are independent of the object of the attitude or state in question. They concern what it takes to be in some state or have some attitude with respect to p, for any p—what would suffice for, or at least contribute to, absent defeaters, knowledge/justified belief that p, for any p.⁷ Some principles are definitional. For example, it may be true by definition that S knows that p only if p is true. But Descartes (1641/1985, 24) imposes a substantive and non-definitional condition on rightful certainty when he argues that certainty requires clear and distinct perception. Therefore, Descartes can count as an SP epistemologist as can more recent, ‘knowledge-first’ epistemologists like Timothy Williamson (2000) who argue for non-definitional conditions on knowledge.

Second, theoreticality: SPE aims to tell us what it takes to know or justifiedly believe, not (by itself) whether you have knowledge or justified belief in some specific case. Whether you know in some specific case usually depends on contingent facts about how you’ve conducted yourself intellectually in that case. But the truth of the principles aimed at by SPE is independent of contingent facts about how you have conducted yourself in that case, and independent of contingent facts about how humans generally conduct themselves.

It might seem, then, that SPE is committed to so called ‘ideal theory.’ Ideal theories in epistemology aim to produce principles about what it would be for ideal cognitive agents to have knowledge or justified belief in ideal contexts. According to one worry about ideal theory (see, e.g., (Skipper & Bjerring, 2022)) because none of us are ideal in this way, investigation into what it is for us to know or justifiedly believe must take into account the ways in which we depart from ideality, including in the ways we in fact conduct ourselves intellectually. But SP epistemologists can include principles that make predictions about non-ideal agents and contexts. As always, what you know depends on facts about how you have conducted yourself and other facts about your specific context. That doesn’t mean that there aren’t theoretical principles that ground truths about those non-ideal contexts whose truth does not itself depend on how you contingently conduct yourself.

Finally, neutrality: again, there is a difference between whether you know something, and what it takes for you to know that thing. As standpoint epistemology teaches, what you know might well depend on contingent facts about your social identity. Perhaps “the sex of the knower is epistemologically significant” (Code, 1991, 7) or individuals with certain social identities are better positioned to have certain kinds of evidence. These views are consistent with there being neutral principles about what it takes for all individuals to know⁸ things and for these principles to ground more context- and identity-variable truths. What counts as knowledge can remain constant even though some have a harder time getting it.

Again, this is not to say that, say, how much evidence you need in order to know something can’t vary according to your social situation or other contextual factors, much as what pressure your tire should be inflated to might vary depending on road conditions and rider skill. But, just as there is an explanation for the appropriate psi that is common across contexts (that is, “getting the most out of your ride”), SPE aims to provide principled explanations for the different requirements that might apply to you depending on your social situation and other contextual factors.

Here’s what SPE need not be:

• 1. Committed to the value-free ideal

• 2. Invariantist rather than contextualist

• 3. Methodist rather than particularist (in Chisholm’s (1982, 66) sense)

• 4. Rationalist rather than empiricist

• 5. Internalist rather than externalist

Many philosophers of science (e.g., (Douglas, 2009)) are skeptical about the value-free ideal—the ideal that non-epistemic values (what we prefer or fear, what’s morally good, what goals are worth adopting) should play no role in what theories to adopt. In epistemology, there is a similar ideal (defended by, e.g., Reed, 2010): that whether you know that p depends only on purely epistemic factors. SPE is committed to neither. SPE aims only at discovering principles about what it takes for you to know something—principles whose truth doesn’t vary with context. The principles can allow that whether you know that p depends on whether you can rightly act on p (as argue, for example, (Fantl & McGrath, 2009), (Hawthorne & Stanley, 2008), and (Stanley, 2005)). Since whether you can rightly act on p can depend on what’s valuable for you, SPE can be value-laden.

Second, SPE is not essentially invariantist. Contextualists (e.g., DeRose, 2009) say that the truth of knowledge-attributions can vary from one speaker-context to another. Therefore, whether it is true to say that S “knows” that p can vary from context to context. But at most SPE requires aiming at discovering truths about what it takes to truly say S “knows” that p that do not vary with context. This is consistent with contextualism and presupposed by most contextualists.

Third, SPE is compatible with both methodism and particularism. Methodism holds that general epistemological principles are applied to individual cases to determine what is known in those cases. Particularism says that general principles are derived from prior discoveries about what is known in individual cases. SPE is neutral between these two ways of deriving general principles (as long as they don’t depend on contingent features of how we inquire). It says nothing about how principles are to be discovered, just that we aim at discovering them. Likewise, fourth, SPE is not essentially a priori: whether a principle or its discovery requires some kind of rational insight or contribution from experience, the principle can be discovered by SPE. Finally, fifth, there have been internalists and externalists who can count as SP epistemologists.

You might be an SP epistemologist too.

3 The regulative critique

If so many philosophers can count as SP epistemologists, why don’t some epistemologists like it? I am primarily concerned with one specific critique: attacks on the emphasis on attitudes and states (what Christopher Hookway calls “static evaluations” (2003, 193)), rather than intellectual conduct. The central worry animating this kind of critique is the one noted at the outset of this paper: SPE—with its focus on attitudes and states—can’t guide intellectual conduct.

Those who recommend a turn to the study of the evaluation and regulation of intellectual conduct sometimes call themselves “regulative epistemologists.” As the regulative epistemologist Nathan Ballantyne puts it,

Regulative theories come in many forms: principles, rules, guidelines, instructions, checklists, informative examples, detailed descriptions of intellectual virtues and vices, and more. (2019, 63)

Many regulative epistemologists provide heuristics and rules that have proven reliable given that refined statistical reasoning is time-consuming and beyond the reach of many inquirers. Some are statistical shortcuts (Bishop and Trout (2005) call them “Statistical Prediction Rules”). Others are rules about how to behave during reasoning, evidence gathering, or engagement with others. For example, though not calling themselves regulative epistemologists, Alex Worsnip advises us to diversify our sources (2019) and Casey Johnson (2018) and Jennifer Lackey (2018) recommend publicly voicing dissent in many cases.⁹

While none of this seems to essentially require SPE, it’s not essentially at odds with SPE, either. Indeed, Ballantyne argues that.

My call for expansion of the field should not offend practitioners. For one, expanding the field’s boundaries in one direction does not necessarily mean shrinking them in another. Indeed, regulative and non-regulative projects can be mutually encouraging and collaborative. (19)

SPE and regulative epistemology, then, do not preclude each other, just as normative and applied ethics do not preclude each other. But even Ballantyne claims that.

Learning about the nature of justified belief won’t usually tell us how to gain or preserve justified belief when we’re faced with complex bodies of evidence and ignorance about the specifics of our situation. (5)

And then there are philosophers who Ballantyne labels “radical” regulative epistemologists (18) who recommend to various degrees replacement of SPE with regulative epistemology. Again, the motivating reason is that SPE can’t guide intellectual conduct.

The most SPE-sympathetic of the radical regulative epistemologists—or, at least, the most sympathetic to SPE’s concern with states and attitudes like knowledge—are Robert Roberts and W. Jay Wood (2007). Still, they argue that.

Regulative epistemology… does not aim to produce a theory of knowledge… Instead, it tries to generate guidance for epistemic practice… rather than just an interesting theoretical challenge for philosophy professors and smart students. This kind of epistemology aims to change the (social) world. (21)

More radical are Michael Bishop and J. D. Trout, who “disagree with most traditional epistemologists in terms of what epistemology is about. This difference couldn’t be more fundamental.” (8) In particular, they criticize what they call “Standard Analytic Epistemology” (SAE):

The fundamental aim of SAE is to deliver an account of epistemic justification or knowledge… The fundamental aim of our approach to epistemology is to provide an account of reasoning excellence. Is this really a deep difference? Yes it is. (16)

SPE’s problem, they contend, is that it cannot guide inquiry and reasoning:

SAE will never provide effective normative guidance, and so it will never achieve the positive, practical potential of epistemology. In fact, we sometimes despair about whether most contemporary epistemologists have lost sight of this potential—and, indeed, of our obligation to seek it. (22)

Mark Webb makes a similar point:

Can epistemology be of any use to non-philosophers? … Not by doing what they have been doing for most of the twentieth century. That is, they can’t help by developing general epistemological theories, or analysing notions like knowledge or justification. The kinds of principles that are developed in this tradition are of no use in helping people in their ordinary epistemic practices. (51)

Why should we think that SPE can’t provide guidance? One reason I’ve already alluded to at the outset of the paper: even a cursory glance at the Gettier literature or the exemplars of SPE from the past century seems to show the futility of relying on SPE for substantive advice about how to conduct yourself intellectually here and now. Here is Robin McKenna (2023) describing a “bemused reaction” (2) to traditional epistemology—a reaction McKenna will go on to defend (2):

If you have taken a epistemology class, you were likely told that epistemology is ‘the theory of knowledge’… So understood, epistemology is concerned with the nature of knowledge, which for many epistemologists really means the conditions under which someone knows something (‘S knows that p if and only if…’).

If your epistemology class (I am assuming you had an epistemology class) was about this, you might have had the nagging feeling there was something odd about the whole thing… [I]t may be that you decided the whole thing was not for you. Either way, it was probably clear to you that, whatever epistemology is, it is an extremely abstract and theoretical enterprise. You might have asked what relevance it has to your life. You might have wondered whether it has any applications to ‘real world’ issues and problems. (1)

The first reason, then, for concern about the guidance potential of traditional epistemology is experiential: we’ve read the stuff; it doesn’t help.

A second reason is more principled and harkens back to Ballantyne’s concern, quoted earlier, that it’s hard to see how doing SPE could tell you how to gain or preserve knowledge. Suppose, in line with Chisholm, that what we want out of SPE’s intellectual guidance is guidance about how to get knowledge—guidance about how to get the various attitudes and states that are the subject matter of SPE. And suppose that the best SPE can do is tell you what those various attitudes and states are. Generally, finding out what something is won’t thereby tell you how to get it. Finding out what money is, how to identify money, and what distinguishes money from things that aren’t money won’t tell you the best ways to get money. At best it will allow you to identify money when you have it (as if you couldn’t do that already).

Likewise, even if SPE is successful in what it aims at—producing true principles about what counts as knowledge—this by itself won’t tell you how to get it. Even if SPE can help you identify when you have achieved knowledge the principles that helped you do this won’t thereby have provided you with the methods by which knowledge is best produced. Therefore, even if your intellectual goals are those very attitudes and states that SPE aims at revealing, SPE won’t thereby provide guidance about how to get them. And if your intellectual goals aren’t the acquisition of knowledge, justified belief, and understanding, then SPE looks to be in even worse position.

4 The defense of SPE, part I: epistemic links

I defend SPE by arguing that, contra the regulative critique, SPE is valuable for intellectual guidance.¹⁰ The reason is not that intellectual conduct aims at acquiring knowledge and that you need to know what it is in order to figure out how to get it.¹¹ The principled regulative critique just rehearsed is therefore beside the point.

The reason is that how you should conduct yourself intellectually—and generally—depends on what you already know.¹² Proper intellectual conduct can vary with variations in whether you know some relevant proposition. Therefore, to get guidance on how you should conduct yourself intellectually you’ll want to figure out what you already know. And it will help you figure that out to know what it takes to know something. Regulative epistemologists aim to provide advice about how to conduct yourself intellectually. As I will argue, that advice must be sensitive to what you already know. Therefore, for you to be guided by that advice, you must be able to recognize whether you know various propositions. To do that it can be helpful to be told what it takes to know something, just like it can be helpful to be told what psi counts as full in order to recognize when your bicycle tire is full. That’s what SPE aims to do.

The argument for the importance of SPE, then, depends on whether there are true principles—epistemic links—tying knowledge to proper intellectual conduct. Here are four possible ways knowledge might be so tied:

• (A) If you know that p, then you may be properly closed-minded toward arguments against p.

• (B) You properly assert p only if you know that p.

• (C) In cases subject to the ‘reasonable doubt’ standard, knowledge (on the admissible evidence) that the defendant is guilty permits a verdict of guilty.

• (D) If you know that p, then you may end inquiry into whether p.

Not to mention ways knowledge is practically important, like:

• (E) Never eat any mushroom unless you know what kind it is.

These principles link knowledge to some activity. Versions of these principles have been defended, e.g., respectively, from A–D, in (Fantl, 2018), (Williamson, 2000), (Moss, 2021), and, as noted above, (Friedman, 2017), among others.¹³

If true, the epistemic links constitute an impressive cumulative case that what you know affects proper intellectual conduct. Even if these principles admit of exceptions, as long as there are some situations in which proper conduct depends importantly on what you know, it will be helpful in those situations to figure out what you know or justifiedly believe. The very fact that we sometimes ask, “How do you know that?” when challenging someone’s conduct tells us that how you should conduct yourself depends at least sometimes on what you know. If so, figuring out what it takes to know can provide guidance.

I think that at least some of these links are true. Just to focus on C (the knowledge norm on inquiry), Friedman points out that.

Assertions like, ‘I know whether Bob went to the party, but I wonder whether he went’ or ‘I know who won the election, but I’m curious about who won’ sound awful. (2017, 309)

Having these questioning attitudes, according to Friedman, is part of inquiry. Therefore, knowing the answer to a question—since it is incompatible with properly holding these questioning attitudes—is incompatible with properly inquiring. I noted above that she is joined in this conclusion by Kelp (2021), van Elswyk and Sapir (2021), Whitcomb (2017), and Millar (2011). None of these philosophers are on record as sympathetic to the regulative critique. What’s important is that even those regulative epistemologists who have concerns about the guidance potential of SPE also end up producing norms that implicate epistemic attitudes and states; even the regulative epistemologists’ own guidance is committed to some such epistemic links.

5 The defense of SPE, part II: the evaluation of counterevidence

Intellectual conduct involves evaluating counterevidence and alternative positions: not just discovering truths via positive evidence, but testing those discoveries against arguments, evidence, and positions to the contrary. Regulative epistemology, then, must and does include recommendations for how and when to seek out counterevidence and alternative points of view, and how to respond when discovered.¹⁴

For example, Ballantyne (2019) proposes that you become less sure of yourself when confronted with significant amounts of evidence that you haven’t yet explored (he calls this “the problem of unpossessed evidence” (173)). If you haven’t done much research on mask mandates, then even if you find yourself with a strong opinion one way or the other, knowing that there are arguments on both sides the details of which you are unaware recommends reduction in confidence. Sometimes, of course, you might have independent evidence that the evidence you haven’t explored will be unreliable. But in many cases you lack this kind of specific defeater defeater.

Bishop and Trout, too, emphasize the importance of reducing overconfidence.¹⁵ The primary technique they recommend is what they call the “consider-the-opposite” strategy, one example of which involves explicitly pausing during reflection “to consider why your judgment might be wrong.” (147) They point out that one difficulty with this strategy is its range: “it would surely be a mistake for us to adopt the strategy for every belief we entertain.” (148) So they ultimately recommend a narrowed version, in which you adopt the strategy only when considering beliefs relevant to “highly significant problems in which short-term reliability is very important.” (148).

So, regulative epistemologists must and do offer advice about how and when to seek out, consider, and respond to counterevidence and alternative points of view. For all SPE tells us, the advice offered by regulative epistemologists might be good. But SPE is vindicated if this advice is essentially state- or attitude-involving.¹⁶

Consider Ballantyne’s injunction to reduce confidence if you find a mass of evidence that bears on p. He allows that this injunction can be defeated if you have “compelling evidence” in favor of your position. For example, Ballantyne endorses Peter Van Inwagen’s assertion that you can defeat the disagreement of members of the Flat Earth Society by invoking the fact that the total evidence rationally compels belief that the earth is spherical (126). Similarly, he singles out “human-caused climate change” as a case in which “the available evidence seems to demand a particular doxastic response”: namely, belief. (127, emphasis added).

Therefore, according to Ballantyne, you need not reduce your confidence in the face of evidence you haven’t yet explored as long as your positive evidence “demands a particular doxastic response.” In order to determine whether your evidence demands a particular doxastic response, it can be helpful to first determine what it takes for that particular doxastic response to be demanded. This is just what SPE aims to do.

Likewise with any plausible version of Bishop and Trout’s narrow “consider the opposite” strategy. Recall that Bishop and Trout recommend that when you are reasoning about a highly significant problem in which short term reliability is very important, you explicitly consider why your belief might be false. But must you consider this about any relevant belief? That’s implausible. Suppose you were a Republican representative considering whether to vote to impeach President Trump for inciting an insurrection. You might have believed that the members of the riot were Antifa infiltrators, that voting for impeachment would damage you politically, that there was massive fraud on election day, etc. All of these beliefs seem relevant to your decision, and because the decision was momentous and short term reliability was important, Bishop and Trout rightly recommend considering the opposite.

But it was also relevant to your decision whether the U.S. Capitol building exists, whether voting for impeachment would inspire a Martian invasion of the earth, whether you had a material body (and thus were capable of raising your hand to vote), and a number of other beliefs about yourself, the mechanics of U.S. government, and the state of the universe. These, I presume, you need not have subjected to Bishop and Trout’s strategy. Why not? The natural answer is that your evidence for them demanded a particular doxastic response—namely, belief.

Nor is it enough to say that the possibility, say, that voting for impeachment would have inspired a Martian invasion, wasn’t a relevant alternative and so didn’t need to be considered at all. The question is what made the possibility irrelevant. Suppose that your total evidence lent significant support to the possibility that voting for impeachment would have inspired a Martian invasion: the possibility was significantly probable on your evidence. Then, surely, that’s a possibility that you should have considered and a possibility that would have been relevant to your deliberation. It’s only because the possibility was so improbable on your evidence that it was irrelevant: it was only because your evidence demanded a particularly doxastic response that the possibility was irrelevant and, thus, did not need consideration.

Therefore, you only need to invoke these regulative strategies if your current beliefs are not what your evidence demands. To use more familiar epistemological language, you only need to reduce your confidence if you don’t know that your beliefs are true.¹⁷ More generally, how you should respond to counterevidence can vary depending on what you already know. Therefore, in order to know whether to take the advice offered by the regulative strategies—in order to be guided by them—you’ll want to first figure out if the relevant beliefs count as knowledge.

In many of the key cases, there will be doubt about whether you know the relevant beliefs to be true. It might seem to you that you know your belief to be true, even though you really don’t. You might defend your resistance to, say, reading various articles arguing that masks are ineffective, and your continued belief that those articles are misleading, by citing your knowledge that masks are effective. Your interlocutor might disagree: “you can’t know that; you haven’t even read the article I sent you!” You might wonder whether you or your interlocutor is right. Are you correct to dismiss your interlocutor’s counterevidence on the basis of your knowledge that masks are effective? Or is your interlocutor correct that you can’t know that masks are effective without engaging with their reasoning?

If you’re looking for guidance about how to proceed with your inquiry, you’ll want answers to these questions. To get them you’ll need to figure out whether you have enough evidence to know that masks are effective. Even if you already know what evidence you have, to figure out whether you know that masks are effective, you’ll need to figure out how much and what kind of evidence it takes to know that masks are effective. These are questions SPE aims to answer.

6 Vindicating two of SPE’s three features

The fact that what it takes to know something is prior in many cases to the question of how you should respond to counterevidence vindicates at least two of SPE’s aims.¹⁸ To try to get guidance from principles about what it takes to know¹⁹ that p requires that we try for principles that are content-, context- and identity-neutral: principles that answer the question what it takes to know that p for any p, for any context, and for any potential knower. We want an answer to the question what it takes to know that p that is going to be helpful no matter who you are, what situation you’re in, and what you’re not sure you know.

We can treat all of these possibilities together. That’s because you might be in doubt or there might be disagreement about whether the content of your belief allows you to know it while you fail to know other things, or whether your identity or context allows you to know something while others can’t know that thing. Suppose you wonder whether some features of your identity or context make the difference to whether you know that masks are effective. Your interlocutor might say they don’t, insisting that the relevant features do not grant knowledge to you while denying knowledge to those who lack those features. Again, your interlocutor might be wrong; perhaps the relevant features do make it easier to know certain things. But if you’re looking for guidance—as the regulative epistemologist maintains—then you’ll want to try to get an answer to the question of why some features of content, identity, or context make the difference between knowing or not. That, again, is an answer that SPE—not regulative epistemology—aims to provide.

You usually seek out guidance when you’re in doubt about how to conduct yourself. It’s when you’re in doubt about when to stop inflating your bicycle tire that you might turn to the chart to figure out when the tire is full. It’s when you’re in doubt about whether to end inquiry that you might look for the features that determine when inquiry is finished. If the epistemic links are true, one such feature is that you know the answer to the question you’re inquiring about. If you’re in doubt about whether you know the answer, then you’ll want guidance on that issue as well. Do you know? Perhaps, if standpoint epistemology is true, members of certain groups have an easier time knowing certain things. But you might be in doubt whether you’re a member of such a group or, rather, whether the group you’re a member of is one of the groups that has an easier time knowing the answer to the question you’re inquiring about.

Guidance would be helpful: guidance that provides functions from your context, your identity, your target question, and your evidence to whether you know the answer to that question. You’d like a chart analogous to the bicycle tire psi chart: general and neutral, usable by anyone independent of their contingent features, needs, and goals even though (and in fact because) the chart makes reference to contingent features, needs, and goals. It is the task of SPE to provide such a chart. It will be a simpler chart if standpoint epistemology is false, and a more complicated chart if it’s true. Either way, you’d like a chart like that if you’re in doubt about whether and what you know. SPE aims to provide it (Fig. 2).

Fig. 2

Degree of evidential support required for knowledge

More to the point, we want a way to generate such a chart. If we fill in a cell arbitrarily, this won’t provide satisfying guidance when you are already in doubt about the cell’s value. Suppose you have doubts about what it takes to know that p, rather than q, or what it takes for someone with your social identity or in your context to know that p. If our advice is arbitrary, you may still be guided, if you just take our word for it. But you might hope for more; you might hope that we’ve tried for more. Ideally, we’ll have tried to discover principled, general, and neutral reasons (akin to “getting the most out of your ride”) that ground the values in the cells.

7 A particularist objection

If the epistemic links are true, the only way that SPE won’t be able to provide guidance is if SPE can’t help you figure out whether you know that p. There are a few ways that could happen: (1) there are no true principles that have the features SPE aims at or (2) there are such principles, but they are undiscoverable or 3) there are such principles and they are discoverable, but they don’t help decide whether you know in the cases in which guidance is necessary.

Of course, even if any of 1–3 are true, there’s no implication that it’s therefore impossible to figure out in some case whether you know in that case. Just as it might be difficult to articulate principles of grammar even though you can recognize in some case what proper sentence formation is, you might be able to recognize when you know that p even if you can’t articulate SPE principles.²⁰ This is perhaps one way to view Roberts and Wood’s simultaneous respect for knowledge, concern about necessary and sufficient conditions, and emphasis on the inculcation of intellectual virtues. If there are no principles that help you determine, in some given instance, whether you know that p, then you are left to your best judgment—phronesis—to determine whether you know that p. Perhaps the ability to judge correctly whether you know that p, in the absence of helpful principles, can be enhanced through training in various virtues.

Still, the most virtuous knowers can disagree about whether the available evidence is enough for knowledge, and can be in doubt about whether they themselves know. If you’re such a knower, even you might face such disagreement and doubt, and so even you will be in a situation in which guidance is needed. Without helpful principles about what it takes to know, you’re stuck.

If so, you’d like to know whether you’re genuinely stuck, or whether help is in the offing. So, you’d like to know whether any of 1–3 are true. But it’s going to be hard to figure even that out without doing some SPE. How can you know whether there are no SPE principles or whether the ones there are can’t help decide the matter unless you do (a lot) of SPE? To decide the matter in advance of doing some SPE is like deciding that there’s no cure for cancer before actually trying to research cures for cancer.

It might be objected—as noted earlier—that we have done a lot of SPE and none of the advice is helpful. But is that really so sure? SPE, for example, tells us that knowledge requires only fallible justification and that testimony is sufficient for knowledge. You might think these conclusions are obvious and that we didn’t need SPE to tell us. But if you think that, you’ve never tried to teach epistemology to first-year students, many of whom immediately respond to discussions of Plato’s Theaetetus by insisting that knowledge requires seeing for yourself. Nor have you listened to much talk radio or visited many internet comments sections. And, of course, some epistemologists (e.g., Climenhaga, 2021; Keren, 2023), respectively) still think knowledge requires infallible support and that it is at least plausible that testimony doesn’t deliver knowledge. That’s to say that there are current debates in SPE whose resolution is necessary for complete guidance.

Still, maybe none of this can be resolved, because there are no deeper principles about what knowledge requires. Perhaps the best we can get is an arbitrary chart and no deeper explanation of why you know and I don’t, why you know that p but not that q, and why you know in one situation but not in another. This point might be pressed by advocates of the epistemic analog of Jonathan Dancy’s (2004) moral particularism. Moral particularism is the view that there are no general moral principles—that moral truths about what you should do in specific situations don’t derive their truth from more general or universal moral rules. Epistemic particularism (not to be confused with ‘particularism,’ in Roderick Chisholm’s sense, discussed above) says something similar: there are no general principles about what is known (see, for example, Bradley, 2019). Whether you know in some situation doesn’t derive from any more general or universal epistemic principle.

Note that this criticism of SPE is different from the regulative critique addressed in this paper. The regulative critique (and most of the regulative epistemologists cited earlier) either grants or is neutral on whether there are true general principles about what it takes to know. Their concern is that even if there are such truths, they can’t provide guidance. That is the concern I hope to have addressed in this paper.

But even if epistemic particularism is true and there are no general and neutral principles that entail that knowledge is present or absent, this is not the death knell either for traditional epistemology or the claim that it’s important to do traditional epistemology in order to get intellectual guidance. For, if particularism is true, then given the epistemic links outlined above, it sure would be useful to know whether particularism is true. I fail to see how we can know that without doing more traditional epistemology… and failing. Therefore, whether there are general principles or not, traditional epistemology can be helpful, if not essential to those in doubt about what they know and, thus, what to do.