1. Introduction
Epistemic evaluation plays a central role in our lives as believers and inquirers. We are beings who seek truth and knowledge, in some cases for practical and moral reasons, and other times just for the sake of it. Even when not actively reflecting or inquiring, we are inundated with testimonial and perceptual data. In all cases, depending on our interest in a matter, we sort and evaluate – contemplate, ignore, keep, or discard – in accordance with (or in violation of) various epistemic norms.
The epistemic norms that I have in mind here are the standard norms of belief which privilege evidence, and certain methods of gathering evidence, as comparatively and reliably likely to land us with what we might call “epistemic goods,” such as truth, understanding, or knowledge, but I also here include norms of inquiry.Footnote 1 Do some research instead of flipping a coin; ask a trusted source instead of consulting a tabloid; minimize distractions when inquiring over some matter; try counting heads instead of guessing; seek out multiple sources of evidence instead of relying on hasty generalization; and so on.
Epistemic norms are norms of rationality, and while they can be idle in telling us what to care about or whether to inquire over some matter, in the first place, they provide a rational basis for belief and for halting lines of inquiry which have reached their limits. We rely on these norms in our own epistemic pursuits, and we deploy them to evaluate the epistemic lives of others, positively and negatively. It is routine to judge others for their keen insight and quick thinking, for their ability to see gaps in arguments and to close them with expert reasoning, for shoring up strong evidential backing for their beliefs, and for knowing when an inquiry has hit a wall. In these cases, from the standpoint of rationality, we think that these people have done something right.
Equally, we judge people for not believing in accordance with the evidence, for being dogmatic, and for relying on wishful thinking. And in cases where people form beliefs using unreliable methods or hold beliefs in the teeth of the evidence, we evaluate those beliefs as being without merit, or, as we say traditionally, unjustified. We also judge people for needlessly inquiring into matters that are settled, and for not pursuing matters that, given their epistemic goals (or ours), we think they ought to. In all these cases, from the standpoint of rationality, we judge that they have done something wrong.
Sometimes, when we judge that someone should know better, our judgment comes bundled with feeling. When we think someone is believing badly, inexcusably so, we might feel corresponding anger or resentment, or other similar negative “reactive attitudes” (Strawson, Reference Strawson1962). This affective reaction is a familiar blame response, with clear analogues in the moral realm, and it is at the heart of our ordinary understanding of what it is to blame someone. This conception of blame is commonsense and unifies blame responses across different domains. As such, I take it to be the essence of what it means to blame someone (contra Boult, Reference Boult2024b and elsewhere), and it will be central to the discussion that follows. As I see it, blaming just is having a negative emotional reaction.
In the epistemic realm, we might blame others when there are harmful consequences to someone’s bad believing – for instance, your epistemic carelessness in not confirming the start time of the play means that I missed the opening act, leaving me indignant. And when the consequences are more serious, indignance can flame to anger. But sometimes blaming a would-be knower arises absent any obvious moral or practical consequences – for instance, I know that you are in possession of conclusive evidence that Donald Trump legitimately lost the 2020 American Presidential election, which is why your refusal to believe as the evidence dictates (you adamantly suspend judgment) confounds and irritates me.
Cases like this one give rise to the question of whether there is a distinctively epistemic form of blame, one which goes beyond negative evaluation (i.e. the evaluation that your belief is unjustified) and which is not reducible to moral or prudential blame. Recently, Brown (Reference Brown2017, Reference Brown2020a, Reference Brown2020b, Reference Brown, Lackey and McGlynn2025), Boult (Reference Boult2021, Reference Boult2023, Reference Boult2024a, Reference Boult2024b), and others have argued that there is.Footnote 2 I think this is right, but I question what we are to make of this. In Section 2, I argue that blaming is a common ingredient in our epistemic practices and arises when we believe that someone has inexcusably veered away from truth, understanding, or knowledge, coupled with the desire that they had not done so. As I discuss below, Brown’s belief-desire account of epistemic blame captures this well. However, as I argue in Section 3, from the standpoint of rationality what counts is blameworthiness, and epistemic blame is neither a necessary nor sufficient condition for blameworthiness. That is, instances of epistemic blame fail to reliably pick out instances of epistemic blameworthiness – they do not reliably tell us if the blame is fitting or appropriate. Epistemic blame is an imperfect metric of epistemic goods, revealing as much about the person doing the blaming as about truth, understanding, or knowledge. As such, even if we grant that epistemic blame is a distinct kind of phenomenon, its value or significance, from an epistemic standpoint, is unclear. In Section 4, I consider the possibility, turning to Boult, that its epistemic significance lies in helping us regulate epistemic relationships, by signaling when we ought to modify trust (or intentions to trust, or expectations that someone is trustworthy). But here, again, I find that epistemic blame is neither a necessary nor sufficient condition for identifying inexcusable violations of norms of rationality, i.e. blameworthiness. I conclude, in Section 5, by considering the ways in which epistemic blaming practices have social utility within epistemic communities, and I offer some reflections on their role in cultivating epistemic agency.
2. Epistemic blame
The two most prominent accounts of epistemic blame in the current literature are both modeled after accounts of moral blame. Jessica Brown’s belief-desire account is modeled after Sher’s (Reference Sher2006) account of moral blame. It is a two-tiered account consisting of a frustrated desire-belief pair that is the consequence of an inexcusable epistemic norm violation. The basic idea is that epistemic blame consists in characteristic affective dispositions, such as feelings of anger, frustration, and resentment, which arise when we believe both that someone has not conformed their beliefs to the evidence, without excuse, and desire that they had not done so. Cameron Boult’s relationship-based account is inspired by Scanlon’s (Reference Scanlon2008) account of moral blame. Roughly, on Boult’s account, epistemic blame arises out of our roles as members of an epistemic community who rely on each other vis-à-vis epistemic matters. Our epistemic relationships naturally give rise to corresponding intentions and expectations, and when someone violates epistemic norms and thereby undermines our expectations, we go beyond judging them to have failed epistemically – we are irked by this failure and modify our relationship accordingly. When you let me down, epistemically speaking, I trust you less. On this view, relationship modification just is what it means to epistemically blame someone.
While the details of these views differ importantly, they have in common the basic idea that epistemic norm violations can prompt a blame-like response in others which goes beyond mere negative evaluation – it involves “something more” (Boult, Reference Boult2024b: 2), like a sting or a rebuke. This “something more” seems almost built-in in the moral case. At least, no one questions the aptness of anger and resentment when moral norms are violated – you are right, we think, to be furious when your spouse cheats on you. Their wrong action has harmed you in a way that is palpable and relatable, and your rage is easily recognizable as blame.
It is precisely this “something more” that epistemic blame theorists take as a challenge to their view. Negative evaluations are a routine part of our epistemic lives, as discussed above, but our judgment that others have fallen short of some epistemic norm can come without any corresponding emotion. Rejecting beliefs as unjustified, dismissing belief-forming methods as unreliable, and repudiating paths of inquiry as fruitless can be done dispassionately, and when done with feeling, can indicate that the blame is not strictly epistemic, or so the thinking goes.
Boult, in particular, frames this as part of the puzzle of epistemic blame that he is out to solve – the stronger the rebuke, he worries, the more at home it seems in the moral or prudential realm than in the epistemic realm. The first example above illustrates this. We can attribute my anger at you for not confirming the start time of the play to the fact that I missed the opening act, and thus only indirectly at your epistemic carelessness. I would not have minded that you failed to check had you got it right, even if by accident, nor would I have minded if you had failed to check and got it wrong, if your error had landed us there early, instead of late. It is safe to say that it was not your lack of rigor or indifference to the truth that bothered me, but its practical consequences. To call this blaming “epistemic” seems like a category error, since my anger is not aimed at your violation of an epistemic norm.
And of course, the consequences of bad believing can be more serious than this. For example, given the empirical data which show that 2 doses of the MMR (measles-mumps-rubella) vaccine effectively provide 97% protection against measles, your skepticism about the vaccine seems to me unjustified, but what really raises my ire is the impact this will have on your kids, whom you have decided not to inoculate. When someone’s epistemic norm violation impacts their life or the life of someone else in a morally problematic or otherwise consequential way, blaming comes easily.
But can’t bad believing, on its own, prompt blame and its attendant reactive attitudes in others? I am less sanguine about this than Boult (Reference Boult2024b), who sees the objection that negative emotional reactions might not be fitting responses to epistemic failings as an important challenge to his account. I think being exercised by someone’s irrationality to the point of anger and resentment is a common enough response, independent of any consequences, though, again, I am not sure what epistemological conclusions to draw from this. I will return to the example of the 2020 American election, offered above, to illustrate the phenomenon of epistemic blame. Establishing that will set the stage for the matter of its epistemological importance.
2.1. Believing badly
As I characterized it, you are in possession of conclusive evidence that Donald Trump legitimately lost the 2020 Presidential election, and yet you suspend judgment on the issue. I know that you are aware of this evidence, and my reaction in this case goes beyond judging that your doxastic attitude is unjustified, indeed, I am positively rankled by this. This example was meant to show that we can be deeply perturbed – angry, frustrated, resentful – by someone’s violation of epistemic norms even if there are no obvious practical or moral consequences to it. In this case, it is the bad believing itself that purportedly causes my affective reaction.
But why? Who am I and who are you, in this scenario, such that this deep irritation makes sense? I think that question is worth exploring, and looking at various iterations of this example can help us understand the phenomenon of epistemic blame, both intuitively and theoretically.
It is easy to imagine a version of this example in which there are practical or moral consequences to your epistemic norm violation. Let’s say that you are a well-known conservative podcast host who has a significant platform which you use to espouse your political views. I am not a fan of your show, but nevertheless I was listening to it recently and I heard you discussing a widely cited report by the Associated Press (AP) which undermines any credible claim of voter fraud in the 2020 American election. However, after discussing this report you went on to question its authenticity on the basis of various conspiracy theories combined with what you called “fake news.” This left you, as you put it, with serious doubts about whether the election was legitimately decided. I got angry hearing this, given your platform and influence. Trump has recently been elected again, and I am worried about the impact of misinformation on his ongoing popularity, given your influence over large swathes of the American people.
Here, again, calling this blame “epistemic” seems off. Yes, your bad believing is the root of the problem, but my reactive attitudes are directed at its practical and moral consequences. That’s the more familiar case, but now let’s modify the example. Suppose that you are not a powerful podcast host, but instead, you are my older sister. Imagine that we are out for dinner and the subject of Trump comes up. We almost never talk politics – you generally tend to keep your views to yourself, and besides, we have starkly different political leanings. Still, despite our best efforts to avoid the subject, here we are. You tell me that you are pleased that Trump is rightfully back in office, given that the 2020 election was likely stolen from him due to massive voter fraud. I am shocked to hear you say this. Even though you are politically conservative, I know you to be a careful and rational thinker, so I can hardly believe what I am hearing. I respond by citing the various reports which have thoroughly debunked this view, including the widely cited one by the AP, which you admit to knowing about and even having read, but dismiss on the basis of various conspiracy theories, just like the conservative podcast host, calling the report and others like it “fake news.” It is hard for me to accept that you are in possession of evidence which appears to unequivocally support P, but that you are nevertheless skeptical and therefore suspend judgment about P.
I am brimming with anger and frustration. I find myself deeply distressed by your thinking on this subject, and not because of any practical or moral consequences. Sure, it is possible (it is always possible) that there are some downstream consequences to your doxastic commitments, but none that are immediately obvious to me, and anyway, that’s beside the point – those consequences, if there are any, are not the reason I am mad. I am mad because you are being irrational, you are believing in the teeth of the evidence. Conspiracy theories are highly unlikely to be true, and your use of the expression “fake news” tells me that you have bought into a politically motivated agenda used by conservatives to undermine legitimate research contained in the AP report and others like it. You seem to have a flagrant disregard for the truth! The evidence is right there in front of you, why can’t you see it?!
2.2. From evaluation to blame
A lot of people reason badly and it means nothing to me, but I do care about how my sister reasons. That is because I want her to see the world correctly, as I suppose that I see it. I want her to have good epistemic habits, and to recognize reliable forms of inference from unreliable ones. And when she fails on this front, in an inexcusable fashion, I not only judge her as doing something wrong epistemically, but I am angry and frustrated about it. I want the truth to matter to her and I want her to be able to access it. I want this for her because I care about her; her well-being matters to me. I want her to enjoy the epistemic goods which are the fruits of rationality: truth, understanding, and knowledge. As I see it, epistemic goods have intrinsic value, on top of whatever instrumental advantages they might have, and I want her to partake in this value.
When it comes to my sister, this desire is untethered from any particular outcome or action. It is the violation of the epistemic norms that is the problem. Note, I do not have this desire for everyone, but I have it for the people in my life who matter to me, and the strength of the desire varies accordingly. I want all my logic students to understand that affirming the consequence is an illegitimate path to a conclusion, but I get particularly aggrieved when one of our philosophy majors, whom I know well and like a lot, does not get it. I especially want them to understand key inferential moves, and when they do not, despite repeated explanations, I blame them for their poor reasoning, just as I blame my sister for her unjustified beliefs – with feeling. My strong desire that they reason well, combined with my belief in an instance of unexcused epistemic norm violation, brings about anger, frustration, disappointment, and various other so-called negative reactive attitudes.
2.3. Belief and desire: Brown’s account
It is routine to evaluate others from the perspective of rationality, to judge them for using unreliable methods or ignoring evidence, to assess them as having unjustified beliefs or for being in a position where they ought to have known better. And sometimes, evaluating a person for violating epistemic norms can come bundled with familiar feelings of blame which get expressed through characteristic dispositions of anger, resentment, disappointment, and frustration.
The Trump example, with me and my sister, shows why these feelings are not out of place in the epistemic realm. Why would you get angry at someone else’s violation of an epistemic norm, if there were no practical or moral consequences to the bad believing? Because you sincerely care about them and want them to be able to access truth, understanding, and knowledge. You want them to reason well and you believe that they have not reasoned well, that they are being irrational, without excuse. This blame does not arise out of moral or practical concerns nor is it aimed at moral or practical concerns. It is distinctly epistemic.
This characterization of epistemic blame does a good job of capturing our ordinary understanding of blame, where to blame someone is to have strong negative emotional reactions to their norm violations. And it is able to explain this affective response as arising out of a belief that someone has erred, coupled with a desire that they had not done so. This coupling of belief and desire sounds a lot like Brown’s understanding of epistemic blame, and I now want to turn briefly to her account, which provides some theoretical footing to support this intuitive understanding of blame.
Following Sher (Reference Sher2006), Brown (Reference Brown2017, Reference Brown2020a, Reference Brown2020b, Reference Brown, Lackey and McGlynn2025) argues that it is not just the belief that you have violated an epistemic norm that results in epistemic blame; it is that belief combined with the desire that you not violate epistemic norms which unifies the dispositions that constitute epistemic blame. The belief-desire set gives some theoretical structure to the phenomenon under consideration – that is, how one person’s epistemic failure can provoke real distress in another person and result in affective blame responses which seem at home in the epistemic realm. The strength of the desire can vary, according to Brown (following Sher), and she takes this to be a virtue of her account, and I agree. As she says, “fortunately, the account does not require that each of us, for all propositions and times, has a strong desire that ourselves and others have true beliefs” (Reference Brown, Lackey and McGlynn2025: 289), but a strong desire is necessary to explain cases where we find negative reactive attitudes, i.e. blame.
A central merit of Brown’s account is that it captures our pre-theoretical conceptions of blame as inextricably tied up with anger and other negative emotions – blaming, we think, just is getting worked up about something. If I have no stake in your bad believing, then my evaluation of it as an unexcused instance of irrationality will not be accompanied by the requisite sting that is conceptually tied to blame; and if it has the requisite sting, then that shows that I am invested in your rationality.
3. What is the point of blame?
It is not always irrational to suspend judgment about P or reject that P even if the evidence conclusively supports P. Arguably, at minimum, one has to be aware of the evidence in support of P, to have considered that evidence, and then to have formed a doxastic commitment that ignores it or goes against it. Even a strict evidentialist like Feldman (Reference Feldman2002), who would reject some of the instances of what I have been calling “bad believing” as such, would accept this one. This violation of evidential norms is a quintessential case of bad believing, and the intuitive and theoretical considerations elaborated here help to make sense of why this kind of unexcused norm violation by someone for whom we desire epistemic goods might provoke in us a distinctively epistemic form of blame. But what are we to make of this? More to the point, why is this phenomenon significant from an epistemological standpoint?
I think we should want to say here that if epistemic blame matters, from an epistemological standpoint, it is because it illuminates something important about epistemic goods and the rationality that underwrites them. We should want to say that epistemic blame matters, in other words, because it is an indication that someone is epistemically blameworthy, that they have violated epistemic norms – that they are veering away from truth, knowledge, and understanding. I think this position is implicit in Brown’s account (and in Boult’s, as I discuss in Section 4, below).
The connection between blame and blameworthiness is an orienting feature of Sher’s (Reference Sher2006) influential account of moral blame, on which Brown’s account is based. As he states: “The point of trying to understand blame is not just that blaming is among the things people find it hard not to do – so too, after all, are laughing, digesting, and being bored – but is rather that wrongdoers are often thought to deserve blame. What really interests us is not blame but blameworthiness…” (Reference Sher2006: 9; italics are mine). The belief/desire pair that constitutes moral blame indicates moral blameworthiness, on Sher’s account, if the belief that someone has erred is true, and the desire for them not to have done so is linked to the “overridingness and inescapability” (Reference Sher2006: 128) of universal, atemporal moral principles.
In the epistemic case (and arguably in the moral case; see footnote 3, below), however, it is precisely in the move from blame to blameworthiness that epistemic blame loses its distinctiveness from negative evaluation. When we evaluate a belief as unjustified, a path of inquiry as fruitless, or a source of information as unreliable, we are pointing to a violation of some epistemic norm. If we are correct in our judgment that an epistemic norm has been violated, then we have successfully picked out a failure of rationality. When we make this judgment with the feeling that arises out of a desire that someone not violate epistemic norms, we are (again) pointing to a failure of rationality, and, again, if we are right about this, then there has been a failure of rationality.
In both cases, what matters from the standpoint of rationality is whether we are correct in our judgment. That is, did the subject in fact do something wrong in believing or inquiring? Is there insufficient evidence for belief? Do the premises fail to support the conclusion? Were the methods used unreliable? Was the inquiry not worth pursing? And so on.
If there has been a failure of rationality, and it is unexcused, then the subject is blameworthy (culpable, responsible, unjustified – pick your descriptor) whether or not you, me, or anyone else believes or desires that they not be so. What’s more, while anger and resentment may indicate blameworthiness, someone can be blameworthy even if no one blames them,Footnote 3 and blaming (like believing and inquiring) is routinely done badly, and thus can fail to be a reliable indicator of blameworthiness. If this is right, then epistemic blame is neither a sufficient nor necessary condition for blameworthiness. This is worth spelling out.
Let us return to the case of my sister, Trump, and the 2020 American Presidential election. As the example goes, I care deeply about my sister’s failures as a would-be knower because I want her to experience the epistemic goods that are the fruits of rationality, and thus I get angry at her failure to believe as the evidence dictates. This example helped make intuitive sense of a distinctively epistemic form of blame, but we can imagine a version of the example where my judgment of my sister comes without much feeling. Let’s suppose that the example above featured the old me, but the new me is unflappable – to each her own, I now think. My many years in therapy have finally left me able to judge without reacting. The new me still cares deeply about my sister and continues to view her as responsible for her failure of rationality – the evidence is right there in front of her, after all – but the new me doesn’t get worked up about it. The new me thinks that the old me had trouble regulating her emotions, tending to overreact when those she cared about transgressed various norms, including epistemic ones. The new me still has the relevant belief-desire set, that is, I want my sister to access truth, more than I would others in a similar position, and I believe that she has failed on this front, without excuse, but that is her loss, the new me thinks. Although I judge her to be at fault, i.e. blameworthy, I do not blame her. I both recognize and accept that I have no control over her life. With a welcome lightness, I shrug my shoulders and move on.
I do not see why an account of epistemic blame should be troubled by this kind of case. An account of epistemic blame ought to capture the phenomenology of blame while providing a theoretical and explanatory framework which makes sense of it. Brown’s account does just that – it pinpoints the feeling of blame that is underwritten by the normative expectations that come with a frustrated belief-desire pairing. On this account, a given instance of epistemic blame can be explained by the presence of the belief-desire pair, but, as my good logic students know, this does not imply that all belief-desire pairings give rise to epistemic blame.
Boult (Reference Boult2024b: 56–60) thinks that this is a problem for Brown. He draws on an analogous case from the moral literature, raised by Angela Smith (Reference Smith2013) against Sher’s account, as a challenge to Brown’s belief-desire account. The example from Smith is that of a loving mother who believes her son to have inexcusably committed a moral wrong, but despite strongly desiring that he had not done so, she lacks the requisite reactive attitudes, that is, she does not blame him. Brown (Reference Brown2020b) also considers this example, but, following Sher, dismisses the case as involving a non-standard psychological response. Boult disagrees. He argues that this kind of response is ordinary. I think Boult is right here, and we see this not only in the moral realm, as he points out, but also in the epistemic one: “there seem to be very natural sorts of cases where an agent can desire that someone not have believed badly and yet fail to manifest dispositions characteristic of epistemic blame” (Reference Boult2024b: 60; italics in original). In his view, if frustrated desires do not lead to blame, then that is an indication “that having such a desire is not really a way of holding S to a normative expectation in the first place” (Reference Boult2024b: 58).
I don’t see that this conclusion follows, but then, I suppose that depends on what it means to hold someone to a normative expectation, in the first place. That is partly what we are trying to figure out in debates about epistemic blame. Is “blaming” the only way to do this? I don’t see why it needs to be so, even if it is one common way. In desiring that someone does not flout epistemic norms, we hold them to certain normative expectations, even if we do so without much feeling. This is not to say, along with the epistemic blame skeptic, that there is no such thing as epistemic blame, but it is to concede that in some cases, (mere) negative evaluation is sufficient to hold someone to a normative expectation.Footnote 4 An account of epistemic blame ought to acknowledge that the act of blaming is affective and emotional, and recognize that not everyone is of the blaming sort, dispositionally.
In my view, this does not detract from the robust explanatory power of the belief-desire account of blame. That account helps us understand why blame response arises in certain cases, such that if a subject has an epistemic blame response, we can be sure that the blamer both believes that the subject has erred and desires that they had not, but, as the example above shows and the discussion here explains, one can have the relevant belief-desire pair without having a blame response. Routine negative evaluation is a way of holding someone to a normative expectation, even absent the characteristic sting or force of blame.
I am not sure Brown would be happy with this concession. Her response to the absence of epistemic blame in a case where the would-be blamer believes that an epistemic norm violation has occurred is to point to a diminished desire on the part of the blamer, but as the example above illustrates, that is not the only plausible explanation. Contra Brown, there are cases where the desire/belief set holds, but the affective or emotional attitudes is blunted, for one reason or another. Maybe, like the new me, the would-be blamer has undergone many years of therapy and has finessed their ability to regulate their emotional responses, thus despite having the relevant belief-desire pair, their judgments do not give rise to blame. Or maybe, for whatever reason, our would-be blamer is emotionally dysregulated – say they grew up in a dangerous home environment where reactivity was met with a threat to their well-being, and so despite having the relevant belief-desire pair, they express no feeling. Or perhaps our would-be blamer deliberately resists blaming on conceptual grounds, having been influenced by philosophical arguments against the undermining nature of an affective form of blame (Pickard Reference Pickard2014, Reference Pickard2017); there’s no constructive point to blame, they think, so they do not blame, despite having the relevant belief-desire pair.
A person’s emotional and affective responses are shaped by a myriad of complicated environmental, psychological, and neurological facts, which means there are many reasons why someone might not express the reactive attitudes characteristic of blame, but lacking the relevant belief-desire pair need not be among them. And since someone can be blameworthy without being blamed, epistemic blame is not a necessary condition for blameworthiness. Its epistemological significance thus cannot lie there. Might epistemic blame be a sufficient condition for blameworthiness? That seems equally unlikely.
It could be that in epistemically blaming someone we get it right, that is, we blame someone only in the case that they have inexcusably violated an epistemic norm, but it could be that we get it wrong. The basic fact that would-be knowers regularly get things wrong, from the standpoint of rationality, is what is structuring the discussion here. Surely that applies in equal measure to our evaluations of would-be knowers and inquirers as it does to their intellectual pursuits. It is easy to imagine someone who is both rash with their judgments and who is the blaming sort, dispositionally. This person might think that your belief that P is unjustified, your inquiry pointless, or your source unreliable, and they might be angry and resentful about it for the reasons discussed earlier: they sincerely care about you and your attainment of various epistemic goods. But unfortunately, this person might be quick to see failures of rationality among those they care about, even when the failures are not there. They are overly hasty in this regard, and you pay the price as the target of their blame. It turns out it was not you who failed to see the probative force of evidence it was them – oops!
It should not surprise us that one person’s reactive attitudes are not a reliable guide to another’s adherence to epistemic norms. Epistemic blame is thus not a sufficient condition for blameworthiness, and since it is also not a necessary condition, its epistemological significance must lie elsewhere.
4. Epistemic blame and relationship modification
On Brown’s account, epistemic blame consists of the negatives attitudes that we commonly associate with blame and arises when we both believe that someone has failed from the standpoint of rationality and desire that they have not done so. I have defended that position here and argued that it offers a good theoretical account of epistemic blame, that is, what it consists in and the conditions under which it occurs. But I have further argued that it leaves open the important question of why epistemic blame matters from an epistemological standpoint.
Earlier, I suggested that our answer here should be that it illuminates something important about epistemic goods – truth, knowledge, understanding – and the rationality that underwrites them. We should want to say, I suggested, that the phenomenon matters because it is an indication that someone has transgressed an epistemic norm without excuse. I think that commitment is implicit in Brown’s account, but as we have seen, an instance of epistemic blame might or might not expose blameworthy failures of rationality, and blameworthy failures of rationality might or might not provoke a blame response in others.
If the epistemological significance of epistemic blame does not lie in its illumination of blameworthiness, where does it lie? As the discussion here illustrates, epistemic blame is a feature of our epistemic practices, so perhaps its significance is derived from that connection, and therefore indirectly to rationality insofar as rationality structures our social relations in the epistemic realm, at least arguably. I think this gets close to a good answer, and I will revisit this idea in the conclusion, with some thoughts on Piovarchy’s (Reference Piovarchy2021, Reference Piovarchy2025) account of epistemic agency. But first, since epistemic practices feature prominently in Boult’s influential account of epistemic blame, it is time to look at his account in more detail.
4.1. Boult’s account
Recall, for Boult, epistemic blame is interpersonal, arising out of our roles as members of an epistemic community. Taking his cue from Scanlon’s (Reference Scanlon2008, Reference Scanlon, Coates and Tognazzini2013) account of moral blame, Boult conceptualizes epistemic blame in terms of the relationships we have with one another, where “relationship” is defined in technical terms as “a set of intentions, expectations, and attitudes people have toward one another that are oriented toward their epistemic agency in different ways” (Reference Boult2024b: 74). On his account, when we judge that someone has failed from the standpoint of rationality, we modify our expectations – we intend to depend on them less, we expect less from them, we reduce trust in them, etc. For Boult, relationship modification just is what it means to epistemically blame someone: “Only when an agent modifies their epistemic intentions and attitudes in a negatively valanced type of way made fitting by this sort of judgment do they count as epistemically blaming others” (Reference Boult2024b: 80–81; italics in original).
Boult’s account captures something important about the phenomenology of our epistemic practices. We do depend on others for epistemic goods and when knowers and inquirers let us down, we do trust them less, and this seems rational – that is, it seems like this is what we ought to do, from the standpoint of rationality. Understanding modifications of our epistemic dependence on others as warranted by our judgment that they have failed, qua knower or inquirer, helps rationalize trust reductions. Importantly, for Boult, this provides the “something more” that takes epistemic blame beyond negative evaluation. The sting or rebuke, on his account, consists of the relationship modification that results when blamers are engaged or exercised into action due to the frustration of their normative expectations.
While this description of our epistemic practices gets a lot right, the move to characterize epistemic blame as consisting in relationship modification, despite it having an influential analogue in moral debates, strikes me as a misstep. Boult’s goal had been to take a well-known phenomenon like blame and delineate a distinctly epistemic version of it. To accomplish this, however, he has entirely given up on the familiar conception of blame as consisting in affective and reactive attitudes like anger, resentment, or disappointment – what Boult calls, somewhat dismissively, “reactive attitudes narrowly construed” (Reference Boult2024b: 28, 83).
Since would-be blamers, on his account, need not express the paradigmatic affective responses that comprises our common sense understanding of blame, why even call them “blamers”? Why are they not better characterized as “exercised judgers” or “relationship modifiers”? It is not clear why we should understand this account as one of epistemic blame. Boult is aware that he has made himself vulnerable to the criticism that his approach “leaves the blame out of blame” (Reference Boult2023: 822; Reference Boult2024b: 102),Footnote 5 but he does not seem particularly concerned about the “cool blame” objection. In his view, blamers are sufficiently engaged by their judgment that someone has failed from the standpoint of rationality such that they have that “something extra” that goes beyond negative evaluation, namely, relationship modification.
By situating “blame” outside of our pre-theoretical notions of blame, Boult has arguably changed the subject, but I want to bracket this concern for the time being and consider whether his account, were we to accept it as is, can tell us why epistemic blame is significant from an epistemological standpoint. Recall, I had said earlier that we should want to say that epistemic blame matters because it tells us something important about epistemic goods and the rationality that underwrites them by identifying someone as being blameworthy. At first glance, Boult’s account looks like it does just this.
If you let me down, epistemically speaking, then I epistemically blame you, i.e. I trust you less and modify my relationship with you. Perhaps, then, relationship modification is an indication that someone is blameworthy, that they have violated epistemic norms and, as such, are veering away from truth, knowledge, and understanding. But here again, as with Brown’s more traditional understanding of blame, it is in the move to blameworthiness that epistemic blame loses its distinctiveness.
On Boult’s account, relationship modification is the consequence of holding someone to a normative expectation which we judge to be culpably impaired, but from an epistemic standpoint, what matters is whether our judgment is correct. That is, what matters is whether in fact the normative expectation was culpably impaired by a violation of epistemic norms, and not whether you or I judge it to be so. Given the routine fallibility of epistemic judgments, this again is where we run into trouble.
Just like the blamer who is overly hasty with their judgment of epistemic wrongdoing, there are overly hasty relationship modifiers, people who, for whatever reason, are quick to cut others off, who have difficulty trusting, epistemically or otherwise, and who tend to modify epistemic trust relationships absent sufficient reason. They may indeed modify their trust relationships in a way made fitting by their judgments of epistemic norm violations, but these judgments miss their mark.
By the same token, there are those who have good reason to modify a relationship from the standpoint of rationality, as they have accurately judged that someone has believed without sufficient evidence, or was wrongheaded in inquiry, but, for one reason or another, they accept this epistemic shortcoming without modifying the relationship. This is common enough. Take my sister and her unwarranted beliefs about the American election, for instance. I can correctly judge that she has erred from the standpoint of rationality, but because I privilege the relationship, I do not modify my epistemic trust in her. Or maybe I am just desperate for her approval, and it’s for this reason I continue to trust in her. Either way, this kind of case is not hard to imagine, which means that there can be blameworthiness absent relationship modification and relationship modification absent blameworthiness.
Thus, relationship modification is neither a sufficient nor necessary condition for epistemic blameworthiness. If there has been a failure of rationality, and it is unexcused, then the subject is blameworthy (culpable, responsible, unjustified – once again, pick your descriptor) whether or not the failure results in relationship modification. And while relationship modification may indicate blameworthiness, someone can be blameworthy even if we do not “blame” them in the Boult sense, i.e. modify our relationship. Relationship modification, like more traditional forms of blame (and like believing and inquiring), can be done badly and thus fail to be a reliable indicator of blameworthiness.
5. Conclusion
Epistemic blame is a routine part of epistemic practices. We blame others for their epistemic norm violations, and for this reason alone the subject is worth theorizing about. Brown’s account of epistemic blame captures our pre-theoretical understanding of blame and conceptualizes blame in way that unifies the notion across various domains. Importantly, as I hope to have shown here, it also has explanatory value in helping us understand the distinctiveness of blaming someone for their bad believing and faulty inquiring. But what this account of epistemic blame cannot do is pick out whether someone has violated epistemic norms, that is, whether they are blameworthy. Neither, however, can Boult’s account. While it is a fact about our epistemic practices that we hold others to normative expectations and modify our trust in them when we judge that they have violated epistemic norms, our tendency to get this judgment right is no more assured than any given instance of inquiring or believing.
I have suggested here that the best answer to the question of why epistemic blame matters from an epistemic standpoint is that it illuminates something important about epistemic goods – truth, understanding, and knowledge – and the rationality that underwrites them. I think this commitment is implicit in both Brown’s and Boult’s accounts, but for the reasons discussed here, accounts of epistemic blame (however understood) fail to clear this bar. Blaming tendencies are shaped by a complicated mess of facts about us (environmental, psychological, biographical, neurological, the list goes on), and are as likely to be indicative of any one of these factors as they are to pinpoint genuine failures of rationality.
If epistemic blame is an inadequate metric of blameworthiness, the question of why we should care about epistemic blame, from an epistemic standpoint, remains. As the discussion here illustrates, epistemic blame is a routine feature of our epistemic practices, so perhaps its epistemic significance is derived from that connection, and therefore indirectly to rationality insofar as rationality structures our social relations in the epistemic realm, at least arguably. As some philosophers have argued (e.g. Fricker Reference Fricker2016), in the moral realm blame communicates judgments of fault, along with disappointment and anger, and this can prompt improved norm adherence by members of the moral community. Maybe epistemic blaming practices are epistemically significant for similar reasons.
Piovarchy (Reference Piovarchy2021, Reference Piovarchy2025) makes a case for this in his discussion of epistemic agency.Footnote 6 He argues that epistemic blame can help us cultivate epistemic agency through epistemic norm enforcement. As he sees it, epistemic blame motivates those who are being blamed to comply with various epistemic norms (gather more evidence, don’t generalize over a small sample size, stay focused when inquiring, etc.), and over time, the idea is that agents will internalize these norms. He argues that the enforcement of epistemic norms is particularly valuable for the well-functioning of epistemic communities, given our epistemic dependence on others,Footnote 7 though he is careful to note that the norms in question must be authoritative: “while I originally argued that we blame people to cultivate their agency, the kind of agency we want is one that is responsive to authoritative moral and epistemic considerations; that is, one which complies with moral and epistemic norms” (Reference Piovarchy2025: 11).
The cultivation of epistemic agency strikes me as a fruitful candidate for the epistemic significance of epistemic blame, even if it is somewhat indirect. Epistemically blaming someone likely does pressure them to comply with norms. But then, I suppose we ought to say here that only if the norms identified are indeed authoritative, such that our blaming practices pick out genuine instances of epistemic norm violations, i.e. blameworthiness, will the right sorts of agency be cultivated.
Open access
