1 Introduction

Epistemic trespassers are experts who pass judgment on questions in fields where they lack expertise (Cf. Ballantyne, 2019a). A particularly pernicious epistemic trespasser was Frederick Seitz (Oreskes & Conway, 2010). Seitz, a Stanford and Princeton educated American physicist, led pioneering efforts in the field of solid-state physics and helped develop the atomic bomb. Early in his career, he was a serious award-winning scientist. Yet, his late career veered away from establishment science when he dedicated his time to sowing doubt about the link between tobacco and cancer and about global warming. Possessing a platform generated by his genuine credentials, he strayed outside his area of specialization to pass judgment beyond his expertise. His pronouncements, like those of many other epistemic trespassers, engendered devastatingly consequential doubt and disbelief.

What’s wrong with epistemic trespassing? Examples of other potential trespassers include Linus Pauling (two-time Nobel Prize-winning chemist who trespassed into medical science), Richard Dawkins (evolutionary biologist trespassing into theology), Neil deGrasse Tyson (astrophysicist who has trespassed into philosophy) (Ballantyne, 2019a), and Scott Atlas (radiologist with no epidemiological expertise who in his role as advisor to the Trump administration’s Coronavirus Task Force denounced public mask-wearing in favor of spreading COVID-19 to reach herd immunity). What wrong do Seitz, Atlas, and the rest of these characters commit when they pass judgment beyond their expertise?

Concern about epistemic trespassing points us to questions about what experts owe others, questions at the intersection of social epistemology and the ethics of communication. Diagnosing the wrong of epistemic trespassing requires taking seriously the social-epistemic role played by experts. Novices rely on experts to speak the truth within their expertise because novices usually cannot evaluate experts’ claims. This dependence on experts makes novices vulnerable. Experts have an obligation arising out of this vulnerability to only pronounce from their expertise when novices could easily confuse inexpert pronouncements for ones grounded in expert authority. When experts trespass, they irresponsibly abuse their authority and neglect novice vulnerability and dependence.

Here is a roadmap for the paper. First, I clarify the notion(s) of epistemic trespassing, I raise a puzzle that should be resolved by any account of the wrong of trespassing, and I motivate the search for an answer to our question (Sect. 2). Second, I identify limitations of a seminal analysis of problems associated with epistemic trespassing in order to derive a set of desiderata that should be satisfied by accounts of what’s wrong with trespassing (Sect. 3). Then, with these desiderata in mind, I develop and defend my own account (Sects. 4 and 5).

2 Groundwork

The idea of epistemic trespassing and the goals of this normative investigation must be clarified. Epistemic trespassing occurs when experts pass judgment on questions in fields where they lack expertise. An ambiguity in “passes judgment” obscures the phenomenon, though. Does it mean forming beliefs or making assertions? Also, if trespassing involves experts passing judgments, what makes someone an expert? On the normative side, do all cases of trespassing deserve the same evaluation? Finally, why must we look beyond obvious answers to our question to find a satisfactory account? I address these issues in turn.

It may seem obvious that epistemic trespassing primarily concerns belief, as epistemic evaluations typically apply to belief. But in the literature, trespassing is frequently characterized in terms of speech rather than belief. It is easy to recognize, Ballantyne (2019a: 369) writes: “Experts drift over a highly-visible boundary line and into a domain where they lack either the relevant evidence or the skills to interpret the evidence well. But they keep talking nonetheless.” [My emphasis] He depicts the “public intellectual” or “celebrity academic” spewing soundbites to microphones and TV cameras. Initially describing his central examples, Ballantyne says nothing about belief. Instead, he (2019a: 367) writes Pauling “asserted,” Dawkins has “written and lectured,” and Tyson “said.” Later when discussing these opening examples, he (2019a: 374) reverts back to judgment-language, noting that “we find experts who issue judgments” beyond their expertise [my emphasis]. But issuing judgments sounds more like assertion than belief.

What should we make of this? Most importantly, we should not assume epistemic trespassing is only a matter of belief. However, these speech-based characterizations of trespassing notwithstanding, we should also not assume it is only a matter of assertion. Complex real-life examples of trespassing involve both belief and assertion. Nevertheless, we must treat these dimensions separately because belief-based problems with trespassing differ from assertion-based ones. In the next section, I examine an account of belief-based trespassing while my own account addresses assertion.

Implicit in the idea of “passing judgment” is that one forms beliefs or makes assertions not on the basis of someone else’s testimony, say from an expert in another field, but from one’s own evidence and skills. So, we should not say you trespass when you pass judgment beyond your expertise on the basis of expert testimony from within that field. You trespass only when you do this on your own.

Expert, rather than novice, trespassing will be our focus. What is an expert?Footnote 1 Ballantyne identifies two components: a person is an expert in a field only if they have (i) enough evidence to reliably or responsibly answer their field’s questions and (ii) enough skills to interpret this evidence. While on the right track, this omits the crucial fact that expertise is also essentially comparative (Fricker, 2006; Goldman, 2016; Hardwig, 1985). Experts have a high amount or combination of true beliefs, evidence, or skills relevant to their fields and possess more of this than most people in a relevant comparison class. My account will rely heavily on this comparative component, treating it as the key to understanding what’s wrong with trespassing.

The expertise referenced in the definition of epistemic trespassing is genuine epistemic expertise, not merely expertise marked by social credentials or reputation. Blowhards and charlatans masquerade as experts, and genuine experts’ expertise may outstrip their reputational credentials. This means appearances about whether someone is trespassing may be deceiving when genuine expertise outstrips credentials or vice versa. Whether a case of apparent trespassing truly is one will often be contested. Unfortunate as this may be, we must accept this result if we want to investigate a real phenomenon of practical concern rather than an overly idealized mere conceptual possibility.

This framework raises a puzzle for those interested in determining what’s wrong with trespassing. If epistemic trespassing occurs when an expert in one area passes judgment about any other area beyond their expertise, then significantly different cases of trespassing exist. The examples referenced so far all involve academics trespassing onto other academic fields. But imagine an auto mechanic without foreign policy expertise passing judgment about international relations. Although this counts as epistemic trespassing, it is not particularly worrisome. Anyone interested in explaining the wrong of trespassing who shares this judgment therefore must account for this asymmetry. Why are some forms of trespassing more troubling than others? What wrong do the academics commit that the mechanic does not? We could refine our definition of epistemic trespassing to exclude auto mechanic-style cases, or we can address this asymmetry head on. I will address it head on.

Granting evaluative differences between cases of trespassing, what’s wrong with it when it is wrong? More will be said since this is our main question, but making preliminary comments will motivate searching more carefully for its answer. Most obviously, trespassers lead novices astray and make messes true experts must clean up, diverting attention away from more important work (Ballantyne, 2019a: 370). Trespassing also causes problems for the trespassers themselves, Ballantyne emphasizes (2019a: 388), as their inexpert overconfidence is epistemically problematic. While these claims are true, we should search for a more satisfying answer to this question. Many things can mislead, including children, advertisements, metaphors, and Internet memes. Yet, given the epistemic role expert testimony ought to play in novices’ minds, being led astray by trespassing experts is worse. (More on this below.) Moreover, I confess to worrying much less about the problems trespassers themselves encounter from trespassing than about how trespassers impact others. Why do we want experts to avoid trespassing? If they lived reclusive existences alone in isolated cabins trespassing day and night without impacting others, if never another soul heard them overstepping their bounds, would we care? The Cartesians among us, who take the lone thinker to be the primary object of epistemic concern, certainly would. Me? I would not. My concern derives from the detrimental impact trespassing can have on others. That possibility of impact earns trespassing its unsavory status.

This does not mean that how trespassers manage their beliefs is completely orthogonal to our question. It only means that we should not explain its pertinence directly in terms of the benighted state of trespassers. Compare: we explain what’s wrong with crimes of negligence like drunk driving or leaving young children unattended in terms of potential harm to others, not the problems faced by the perpetrator. Still, if we want to understand why these acts occurred and how to prevent similar acts from occurring, it is appropriate to examine the perpetrator’s problems. Similarly, we need to distinguish (1) an account of what’s wrong with trespassing from accounts of (2) why trespassing occurs and (3) how to reduce the likelihood of its future occurrence. Trespassers’ problems pertain to (2) and (3). Looking to these problems to develop (1) seems misguided. While a trespasser’s overconfidence may explain why he wronged, it fails to explain why trespassing is wrong.

The concern for genuine experts playing intellectual custodians, mopping up after trespassers, rightly shifts our attention to potential costs trespassers impose on others. But focusing on opportunity or other costs suffered by these experts misses the mark, as experts sometimes need to clean up after true experts who have muddied the waters anyway. The wrong I intend to identify more closely addresses the nature of trespassing by considering how it might create problems for vulnerable novices. I do not claim that this is the only wrong associated with trespassing. Rather, I claim that my view will explain the wrong perpetrated by trespassers while accounting for experts’ social-epistemic role and distinguishing trespassing from other ways of being led astray.

Although forming beliefs and making assertions are both forms of trespassing, my answer to our question will address assertion. But first I critically examine an account of belief-based trespassing. Since these two accounts assess different kinds of trespassing, they cannot strictly be seen as competitors. Nevertheless, examining this account, including its limitations, will illuminate what accounts of the wrong of trespassing should accomplish. I examine this account rather than one that addresses assertion because it’s the only account in the literature I know of that attempts to answer any question about problems associated with trespassing.

3 The higher-order defeat account

The account I will consider, from Ballantyne (2019a, 2019b), describes a problem associated with belief-based trespassing. Though it was not meant to answer our question about what's wrong with trespassing, examining it will shed light on desiderata for accounts that do attempt to answer this question.

The question that receives Ballantyne’s attention is: How should experts revise their beliefs once they recognize they lack the skills or evidence necessary to form beliefs in a particular field beyond their expertise? Labeling this “the Problem of Epistemic Trespassing,” Ballantyne (2019b: 204) describes it as “the difficulty of explaining how our confident beliefs fit together with the observation we have trespassed.” This problem of individual epistemic rationality asks how an expert can rationally maintain their beliefs when they recognize that those beliefs resulted from trespassing, i.e., when they have higher-order evidence against those beliefs. They cannot, Ballantyne (2019b: 204–2055) argues. When you recognize that your beliefs concern questions you either lack evidence or skills to answer, you must reduce your confidence because this recognition provides you with a higher-order defeater. That is, unless you have a defeater-defeater that allows you to explain why maintaining these beliefs does not violate any epistemic norms. He encodes these judgments into a normative principle:

Trespassing Defeat If we are experts in one field and believe proposition p, and we recognize that (i) we lack another field’s evidence concerning p or (ii) we lack another field’s skills to evaluate the p-relevant evidence, then we have a defeater for our belief that p, unless we have reason to accept some defense. (2019b: 207)

The defense Ballantyne most carefully examines is the Transfer Defense: “I am trespassing on another field, but my own field’s skills successfully transfer to the other field.” Noting that he expects this to be the most common justification given by trespassers (2019a: 381), he illustrates it with the examples of Richard Dawkins and Neil deGrasse Tyson. Dawkins (2008: 79) denies that theologians have any expertise about cosmology that scientists like himself lack. His scientific expertise, he thinks, transfers to this domain. Tyson may also believe his scientific training has taught him the only critical thinking skills needed to evaluate philosophical questions.

Call this account the higher-order defeat account, because it primarily addresses the rational response to receiving the higher-order defeating evidence that one has trespassed. Examining the details of this account will reveal three of its limitations, which will help us develop desiderata on an account of what’s wrong with epistemic trespassing. The first limitation is that this account only applies to cases where trespassers admit they have trespassed. Second, it doesn’t capture what’s wrong with the examples motivating inquiry into trespassing. Third, it fails to connect its explanation of what’s problematic about epistemic trespassing to the nature of the phenomenon.

First, this approach only applies to cases where trespassers admit they have trespassed. The Problem of Epistemic Trespassing and Trespassing Defeat are both conditioned on the assumption that the trespasser recognizes and hence believes they have trespassed. More specifically, the epistemic trouble this account addresses stems from experts recognizing they are passing judgment on questions in fields where they lack evidence or skills (Ballantyne, 2019a: 376). But it’s possible to trespass without realizing it. Explaining why maintaining a belief recognized to have resulted from trespassing is epistemically irrational does not explain what’s wrong with trespassing. These distinct phenomena require distinct diagnoses. Another way to see this: Distinguish three possible stages of trespassing.

  • Stage 1 I trespass without realizing or believing I do.

  • Stage 2 I acquire reason to think I have trespassed.

  • Stage 3 As a result of accepting this reason, I recognize that I have trespassed.

The Problem of Epistemic Trespassing and Trespassing Defeat only address Stage 3. But Stage 1 is where the trespassing occurs. To explain what’s wrong with trespassing, we must address Stage 1.

Second, insofar as it targets Stage 3, this account neither engages with nor captures what’s wrong with the examples motivating our inquiry into epistemic trespassing. I agree with Ballantyne’s insightful observation that trespassers are likely to think their expertise transfers to the relevant field. But notice the “Transfer Defense” is a conjunction: I am trespassing on another field, but my own field’s skills transfer to this field. This is not the justification trespassers will offer. Because they accept the second conjunct, they’re unlikely to accept the first. Many trespassers are likely to deny they are trespassing in the first place. By Dawkins’s lights, he’s not trespassing; he’s at least as qualified as the alleged experts. Now, Dawkins’s confidence on this point, which may itself be cause for concern, may explain why he trespasses, if he does. To prevent trespassing, we might try to reduce potential trespassers’ confidence about the extent of their expertise. But why should we want to prevent trespassing? Answering that question requires understanding what’s wrong with trespassing even if trespassers do not believe they’re trespassing, and hence what’s wrong in the motivating cases. Because the higher-order defeat account only applies to cases where trespassers acknowledge their trespassing and it’s reasonable to think this assumption doesn’t obtain in our motivating cases, it does not apply to those cases.

This might seem nitpicky. Perhaps charitably reinterpreting statements in the defense of this account where ‘trespassing’ appears would reveal that this account applies even when experts don’t acknowledge their trespassing, in the stricter sense. Maybe the Transfer Defense is more plausibly rendered: “I am crossing into a field where my credentials are not firmly established, but my skills transfer.” Then it might seem that someone like Dawkins must admit he’s doing that when he addresses questions about God. If this is right, this account doesn’t suffer from the limitations I’m claiming it does.Footnote 2 I have two main responses. First, this looser sense of trespassing is clearly not what is under consideration here or in Ballantyne’s account; genuine epistemic expertise is the explicit target, not mere social markers of expertise. Still, maybe ‘trespassing’ is sometimes used loosely in Ballantyne’s presentation, so we should reinterpret his claims so they do not depend on experts acknowledging their trespassing. The second response is that deploying this reinterpretation strategy will not save the account from these criticisms. Its central claims that imply these limitations are not articulated in terms of trespassing. Trespassing Defeat, for instance, does not use the term ‘trespassing,’ but is explicitly conditioned on experts recognizing certain deficiencies in themselves. Insofar as Trespassing Defeat clearly articulates and fully represents the normative underpinnings of this account, we have strong reason to think the account only applies to experts who recognize their trespassing. This makes sense: Ballantyne is developing an account addressing problems of individual epistemic rationality. These problems emerge out of tensions within an individual’s perspective, not between one’s perspective and unnoticed facts.

What matters for our purposes is that any account of the problems associated with epistemic trespassing that only applies to cases wherein trespassers recognize or admit they have trespassed is significantly limited in scope. I echo Ballantyne’s insightful observation that trespassers are likely to think their expertise transfers to the fields into which they’re accused of trespassing. But this implies that epistemic principles, like Trespassing Defeat, conditioned on people acknowledging their trespassing are unlikely to apply to many cases that interest us.

The final limitation in the higher-order defeat account is that its explanation of what’s problematic about epistemic trespassing is independent of the fact that trespassing involves experts passing judgment outside their expertise. It leaves us wondering what about expertise makes trespassing problematic. The idea underlying this account is that learning you have trespassed is a higher-order defeater. This problem, though, is a generic one related to higher-order evidence, not specific to experts trespassing. Closer scrutiny of Trespassing Defeat confirms this: “expert” in its antecedent plays no role in its consequent. Without losing its force, ‘expert’ can be dropped and it can be restated:

Trespassing Defeat-R If you believe p and you recognize that (i) you lack relevant evidence concerning p or (ii) you lack appropriate skills to evaluate p-relevant evidence, then you have a defeater for your belief that p.

Trespassing Defeat-R applies to all the trespassing cases to which Trespassing Defeat applies, but makes no mention of expertise. You arguably should reduce confidence in your beliefs when you acquire reason to think you lack the skills or evidence necessary to rationally form those beliefs. But this generic principle has nothing specifically to do with experts trespassing.

In this section I’ve identified three limitations of the higher-order defeat account. The point here is not that this account fails on its own terms. Rather, these limitations help us identify desiderata on an answer to our question, what’s wrong with epistemic trespassing?

Desideratum 1 It should explain what’s wrong in the cases motivating our inquiry into the phenomenon (E.g., Seitz, Pauling, Dawkins, Tyson, Atlas).

Desideratum 2 It should apply to cases where epistemic trespassers do not acknowledge their trespassing.

Desideratum 3 The wrong identified by this account should not be independent of the fact that epistemic trespassing involves expertise.

In addition, I’ll add a desideratum derived from the puzzle presented in §2, because not all cases of trespassing are equally troubling:

Desideratum 4 It should explain the normative difference between academic and auto mechanic cases.

By identifying a significant wrong with epistemic trespassing derived from the social role of expertise, and thereby addressing the cases motivating our inquiry and tying the explanation closely to the nature of epistemic trespassing, my account will satisfy these desiderata.

4 Epistemic trespassing and the social role of expertise

Instead of focusing on problems confronted by individual trespassers, my account addresses the social side of trespassing. Epistemic trespassing worries me primarily because it can affect others. Casting doubt on the link between cancer and tobacco or on global warming when you’re not an expert on these issues, as Seitz did, strikes me as deeply wrong and worrisome precisely because it impacts others. In this section, I sketch and defend a view motivated by this concern.

This view begins with a social interpretation of passing judgment as assertion. Assertion is an interpersonal speech act. One asserts to an audience. Even the reclusive scholar who spends years crafting her magnum opus alone in her study generally intends to speak to others through this work. Moreover, even if fringe examples of non-interpersonal assertion exist, these exceptions will surely prove the rule. Most importantly, the speech acts in our motivating examples are certainly interpersonal. Seitz, Pauling, Dawkins, Tyson, and Atlas trespass by speaking to others.

There are three kinds of trespassing cases to consider, differentiated by audience composition: the audience hearing the assertion consists only of experts in the field being trespassed upon, only non-experts, or a mix. As an example of the first case, a thinker might trespass by speaking during a conference session only attended by experts. The latter two kinds of cases range from public speaking engagements to publishing books or works in academic journals accessible to novices. Trespassing by making assertions only to experts may be problematic, but often experts can spot trespassers. Novices, in contrast, usually lack the ability to reliably discriminate between trespassing and genuine expert assertion. This is the main source of the problems surrounding epistemic trespassing. In this regard, Ballantyne’s otherwise apt label “trespassing” is somewhat misleading. Not that he intended this, but the label suggests that trespassing primarily harms those whose turf has been trespassed upon. I will attend to cases where trespassers’ audiences include novices and argue the real harm is done to them.

If the interpersonal speech act of expert assertion constitutes epistemic trespassing and if epistemic trespassing is wrong, then it stands to reason that interrogating the relationship between experts, expert assertion, and its addressees would help us discover why it is wrong. How does expert assertion influence others? Relatedly, if the wrong arises from experts making assertions, we might naturally look to norms of assertion to identify that wrong. I will pursue this thought not because it provides a complete and satisfying explanation, but because it will help elucidate the relationship between the social role expert assertion plays and the wrong of trespassing.

The first challenge is to identify which norm of assertion epistemic trespassing violates. None of the usual suspects fits the bill. Consider the knowledge norm: one must assert P only if one knows P (Williamson, 2000).Footnote 3 Epistemic trespassers do not necessarily violate this norm, since they can know the truth of propositions in fields outside their expertise. Admittedly, trespassing by asserting known truths rather than falsehoods or misleading information is less worrisome. But, as Lackey (2011, 2016) argues, some assertions are epistemically proper only if they express beliefs with specific kinds of grounds. For instance, Lackey argues that “isolated second-hand knowledge”–knowledge solely based on another person’s testimony while lacking much other relevant information about this matter–constitutes a deficient ground for a medical doctor to flat-out assert that her patient has a certain disease. This is an epistemically improper assertion, she argues, because it is not based in expertise. If this is right, it suggests knowledge is not enough for epistemically proper assertion. (Similar remarks apply to justification, warrant, or truth norms of assertion.) My case will not hinge on Lackey’s specific argument from isolated second-hand knowledge; also, the cases of problematic expert assertion Lackey discusses are not cases of epistemic trespassing because they all involve experts making assertions within their domain of expertise.Footnote 4 Nevertheless, my argument heeds her more general lesson that the quality of an assertion’s grounds partly determines whether that assertion is epistemically appropriate. In particular, her conclusion that expert assertion must be grounded in expertise is independently plausible and helps diagnose what’s wrong with trespassing.

If none of the familiar norms of assertion is necessarily violated by epistemic trespassers, then we need to consider an unfamiliar norm:

Expert Norm of Assertion (ENA) When S is speaking as an expert, if P is a proposition in domain D, S’s assertion that P is epistemically proper only if (i) S is an expert in D, (ii) S accepts P, and (iii) S’s acceptance of P is grounded in S’s expertise in D.

Clarification is in order. First, acceptance includes more and less tentative attitudes. Belief counts, but less committal cognitive pro-attitudes might too. Experts need not be condemned for asserting claims that represent their best hypotheses even if they do not fully believe these claims, for instance. Articulating the principle in terms of ‘acceptance’ allows for this possibility. Second, sometimes it’s obvious when someone speaks as an expert—as when scientists act as expert witnesses in trials, doctors deliver medical diagnoses, or mechanics tell you what’s wrong with your vehicle. But sometimes it’s not: Is an expert discussing topics related to their expertise with their partner over dinner speaking as an expert? Certainly, our expectations with respect to the epistemic grounds needed for proper assertion are lower in the latter case than in the former cases (Lackey, 2011: 272–273). So, when does one speak as an expert? Other accounts that appeal to this notion leave it unanalyzed (e.g., Benton, 2016), but a variety of reasonable criteria come to mind. First approximations of sufficient conditions, probably requiring modification, include:

  • If you have been granted access to a platform because you are an expert, then when you speak from that platform you are speaking as an expert.

  • If (a) you have been given a platform to speak to audience A and (b) were you to speak about topic X from that platform, A would reasonably take you to be an expert on X, then you are speaking as an expert when you speak about X from that platform to A.

  • If you would reasonably be taken to be speaking as an expert on X by A were you to speak about X to audience B, then you are speaking as an expert when you speak about X to B.

  • If you are an expert on topic Y and you purport to be offering expert opinion on X, then in doing so you are speaking as an expert.Footnote 5

These get the right results in the witness, doctor, and mechanic cases, and they also (arguably correctly) do not necessarily imply that experts casually chatting with their partners over dinner are speaking as experts. These principles do not exhaust the circumstances in which one speaks as an expert, but they may help us understand our cases of trespassing. To fully understand and apply ENA, we would need a complete account of this notion. Indeed, its slipperiness may partly explain why trespassers sometimes feel comfortable speaking beyond their expertise: because they don’t take themselves to be speaking as experts. So, eventually, this notion must be carefully analyzed. Fortunately, though, in our motivating cases the thinkers clearly are speaking as experts, intuitively and on the above criteria, so ENA clearly applies to them.Footnote 6

When trespassers speak as experts, they violate ENA by making assertions not grounded in expertise in the domain of assertion because they lack expertise in that domain. ENA allows us to diagnose a distinctive wrong of trespassing.

Why should we think something like ENA is true? For one thing, if expert assertion based in isolated second-hand knowledge is problematic, as Lackey argues, ENA explains why. More importantly for our purposes, ENA has both intuitive and theoretical support. Consider expert testimony in trials. Experts begin their testimony by circumscribing their expertise, then they testify within that domain. I submit that they deserve criticism to the extent that their testimony outstrips their expertise.Footnote 7 Consider an example. In 2003, the well-known social psychologist Claude Steele gave an expert report on standardized testing and affirmative action in Supreme Court cases Gratz v. Bollinger and Grutter v. Bollinger. Steele begins:

My testimony is based, most generally, on an expertise that has been developed over a 25-year period of research in the areas of social psychology, the social psychology of race and race relations, and the effects of race on standardized test performance. In preparing this testimony I have consulted a broad range of knowledgeable colleagues and experts in these areas, as well as the relevant research literature. My testimony is also based on a 10-year research program that I have directed, the aim of which has been to understand the role of race and gender stereotypes in shaping test performance and the formation of academic identities. (Michigan Journal of Race & Law, 1999)

After delineating his expertise, Steele testifies that using standardized tests to evaluate minority college applicants is an unreliable practice. This testimony is completely appropriate, since Steele clearly has expertise on this matter. Now imagine he’s asked upon cross-examination whether race is real. As an intelligent African-American scholar, Steele might have well-founded opinions on this matter that qualify as knowledge. But it would be inappropriate for him to answer this question directly because it falls outside his expertise. When you’re asked to speak as an expert in a trial, you should speak only from expertise on questions on which you are an expert.Footnote 8

Thus, ENA holds in some contexts. Which contexts? There’s a compelling idea that asserting a proposition involves represents oneself as knowing that proposition (Unger, 1975). Something similar may apply to expertise. Although assertion does not generally involve representing oneself as an expert, assertion in contexts where one speaks as an expert does seem to do so. In these contexts, we expect experts, like doctors, scientists, lawyers, accountants, and academics, to pronounce only on questions within their areas of expertise. Otherwise, they misleadingly represent themselves as experts expressing opinions based in their expertise.

One might object that ENA merely reflects institutional norms, rather than deep facts about epistemic normativity. Perhaps the requirement to only speak from expertise derives from professional (e.g., medical, legal, scientific) standards. If so, why insist that ENA extends beyond these institutional contexts? Because, as Lackey (2011: 267) convincingly argues, these institutional norms are based in epistemic norms, not the other way around. The best explanation of why institutions prohibit assertions not based in expertise is that they recognize the problematic nature of such assertions in contexts where agents may be speaking as experts.

Still, claiming that ENA reflects intuitions about epistemically appropriate speech doesn’t explain why expert assertion must be based in expertise. This is a general point about norms of assertion. Defenses of these norms typically compare intuitions about the appropriateness of speakers making certain assertions with intuitions about whether those speakers meet certain epistemic conditions. Even if that form of argument is convincing, though, a deeper question about why a given norm is true must be confronted. This may seem like a bad question. Williamson (2000: 267) argues that because the knowledge norm is a constitutive rule, rather than a contingent one in need of explanation, asking why it governs assertion is pointless. Still, this doesn’t mean norms of assertion should go unexplained. Williamson admits that construing norms of speech acts as constitutive leaves a residual, meaningful explanatory question unanswered: why do we have that speech act governed by that norm? Thus, defenders of norms of assertion must eventually answer an explanatory question. If the norm is constitutive, they must explain why there is a speech act governed by that norm. If the norm isn’t constitutive, but functions more like an evaluative standard, they must explain why that norm is true.

Rather than resting content with the intuitive reaction that ENA is true, I’d like to explain why it is. The question is either why do we have expert assertion/testimony or why does ENA govern expert assertion. Start with the first question. Why does expert testimony exist? This story will be familiar to social epistemologists.Footnote 9 There is too much information in the world for any single human to learn on their own. Each of us has limited resources to devote to inquiry. But often we want someone to know information we don’t know first-hand. To solve this problem, we have adopted a cognitive division of labor. Some people learn all there is to know about cardiology, others the climate system, others law, and so on. Often, we want others to share their specialized knowledge. Someone who knows little about medicine wants to know what their symptoms signal about their health. What do they do? They ask a medical expert. Even if a novice possesses the evidence an expert would use to inform their thinking, they may still need an expert assessment of this information because they lack the appropriate training to responsibly interpret it. Expert testimony thus arises out of an unconsciously coordinated, rarely explicitly recognized, attempt to overcome individual cognitive limitations. We need not just assertion–not just individuals stating claims they think are true–but expert assertion to fill the unavoidable gaps in knowledge that arise from these limitations.

This story illuminates a central relation between novices and experts. Novices depend on experts to know what they don’t and to responsibly share this knowledge. Remember, expertise is inherently social and comparative. Experts have more evidence or skills than novices. This asymmetry—the epistemic superiority of experts to novices–has social-epistemic implications, nicely articulated by John Hardwig in his seminal paper “Epistemic Dependence.” In short, Hardwig argues that sometimes when you encounter expert testimony it is irrational to think for oneself. The epistemic superiority of experts over novices entails that experts are better positioned than novices to reliably answer the questions on which they are experts. For this reason, rationality sometimes requires deferring to experts (Harwig, 1985: 342; Fricker, 2006: 243).

The claim that novices cannot rationally refuse to defer to experts will strike some as too strong. No doubt sometimes novices can ignore experts. If you know the expert’s thinking is impaired or if they make ridiculous claims (like a meteorologist forecasting raining eggplants (Elga, 2007)), deference is not required. But novices must be careful even here. Well-established scientific facts and theories contradict ordinary experience–e.g., relativity of simultaneity–or strike some novices as wholly unlikely–e.g., evolutionary theory. Thinking for oneself in these cases may be inappropriate.Footnote 10

What matters here is not the strong claim that novices must never think for themselves when they encounter expert testimony that contradicts their own thinking. We can distinguish cases where novices have opinions regarding the topic of an expert’s testimony and those where they don’t.Footnote 11 When they lack opinion, something near complete deference to an acknowledged expert’s testimony seems appropriate (Fricker, 2006: 236; Haack, 2015: 43). That suffices to motivate my argument. My concern about epistemic trespassing derives from the fact that dependence on experts makes novices vulnerable. Novices who must defer to experts in this demanding way are vulnerable. But even when novices have an opinion, expert testimony puts substantial rational pressure on them to accept it, since from the novice’s perspective that testimony possesses stronger epistemic grounds (Fricker, 2006: 236). Even Lackey (2018: 243), who eschews expert authority and argues forcefully against certain “expert-as-authority” models of expertise, endorses a model that delivers this result: “an expert’s testimony might be weighed so heavily that it virtually guarantees my corresponding belief, either because it is just about all of the evidence I have on the matter or because the expert is so epistemically superior to me.” So, even if not all novices must always completely defer to expert testimony, it plays a weighty social epistemic role.Footnote 12 Sometimes novices must significantly defer to experts.Footnote 13

Epistemic dependence on experts makes novices vulnerable. What kind of vulnerability is this? It’s not simply vulnerability to error that arises out of fallibility or our mere epistemic dependence on others. Rather, the superiority of experts to novices implies, Hardwig (1985: 336) writes, “rational authority” over novices that undermines their “intellectual autonomy.” We don’t surrender our intellectual autonomy whenever we hear another person make an assertion. But when I, someone who knows little about physics, hear physicists make assertions concerning spacetime that don’t cohere with my understanding of the world, it is, as Hardwig writes, irrational to think for myself. Thus, dependence on expert testimony makes novices acutely vulnerable by requiring partial abandonment of their intellectual autonomy. Abandoning their intellectual autonomy leads novices to incur interrelated practical risks and intellectual costs (Fricker, 2006: 242). First, surrendering thinking often entails surrendering decision-making. Your doctor says you have cancer and you need chemotherapy. Unable to assess the evidence and options yourself, you listen and submit your body to painful treatment. Practical dependence accompanies epistemic dependence. This dependence also leaves one susceptible to practical manipulation, and all the devastating consequences that can accompany it. Second, the primary intellectual cost novices incur by deferring to experts is a reduced ability to manage and police their own belief system. Beliefs based in expert testimony are often inferentially isolated from other beliefs and information, especially defeating evidence (Fricker, 2006: 242), and novices rarely have the ability to assess those beliefs or relevant evidence on their own. Yet, thinkers grant these beliefs with authority derived from the expert testimony on which they’re based. As a result, these beliefs tend to become entrenched and difficult to dislodge because they are prioritized and given great weight in rational deliberation. This is why misleading testimony from experts is often worse than other misleading information and the harm of epistemic trespassing cannot be reduced to simply being led astray. Rarely tractable or transient, errors based in alleged expert testimony can be more recalcitrant and enduring than other kinds of errors.

Much of the literature on testimony, including expert testimony, focuses on hearers’ responsibilities, but the possibility of epistemic trespassing along with the practical and intellectual risks associated with novice dependence on experts demands shifting our attention to the responsibilities of speakers. For distinct but related reasons, theorists working on the ethics of scientific communication have identified norms that should be respected by expert scientists conveying information to nonscientific audiences. Keohane et al. (2014) defend five principles: honesty, precision, audience relevance, process transparency, and specification of uncertainty about conclusions aimed at articulating what scientists should communicate to nonscientists. Scientists should not deceive their audiences, be precise, provide clear descriptions of scientific processes, and so on. ENA, in contrast, articulates when speakers should make expert assertions and on what basis. These approaches share the goal of providing norms governing expert-to-novice speech. However, Keohane et al. (2014) are swayed by a natural temptation to articulate the communicative responsibilities of scientific experts in terms of a social contract. Following Baier (1986), we should resist justifying expert responsibilities via a social contract. When there are extensive asymmetries in dependence, vulnerability, and power, as there are between experts and novices, contracts provide a poor conceptual framework to ground these responsibilities. Instead, we should justify these responsibilities directly in novice vulnerability grounded in their dependence on experts. Experts must take care with novice vulnerability.

Williamson (2000: 268–269) too ties norms of assertion to responsibility and authority. He draws an analogy between commands and assertion. To issue a command with appropriate authority is to confer a responsibility on the person receiving the command that is discharged when it is obeyed. The knowledge account extends the analogy between commanding and asserting by claiming that making an assertion confers a responsibility on oneself to ensure the truth of its content that one discharges by having the requisite knowledge.

Expert testimony is both an assertion and a sort of command. The epistemic authority behind genuine expert testimony is analogous to the authority behind a command. Just as someone issuing a command confers a responsibility on the person receiving it, when someone delivers expert testimony to a novice, they confer a responsibility on the novice to believe what’s asserted. Failing to adopt this belief means failing to discharge this responsibility. But because expert testimony confers this responsibility on others, experts also confer responsibilities on themselves when they speak as experts, namely, the responsibility to possess and speak from expertise. When experts violate ENA, they impose an undue burden of responsibility on the listening novice and fail to discharge their own responsibilities.

We can finally explain the wrong of epistemic trespassing. Novices listening to experts are often rationally required to accept what experts assert. Such acceptance entails incurring serious practical and intellectual risks. Novices often lack the ability to judge where expertise begins and ends. So even when experts speak beyond their expertise, refusing to defer will often be irrational. When experts trespass, they irresponsibly abuse their authority and neglect the responsibilities they have toward epistemic dependents. As in any case of dependence, those charged with caring for the well-being of dependents have stringent responsibilities, including the responsibilities to avoid (i) abusing power asymmetries arising from dependence and (ii) neglecting dependent vulnerabilities. By failing to discharge these responsibilities, epistemic trespassers irresponsibly impose significant practical and intellectual risks on novices.

On this account, trespassing is an epistemic crime of negligence. Crucially, crimes of negligence cannot be defended by claiming that no one was harmed by the negligent behavior. Doing no damage does not absolve drunk drivers of responsibility. Risking harm to others is enough to have done wrong. Rules prohibiting negligence are aimed at discouraging reckless disregard for others. This is how we should understand a prohibition against trespassing. Trespassers disregard novices and risk harming them by trespassing. A particular act of trespassing cannot be defended by claiming that it harmed no one. Risking harm is enough to have done wrong. The prohibition against trespassing aims to discourage experts from disregarding vulnerable novices.Footnote 14

We can also now finally address the academic/auto mechanic asymmetry. An auto mechanic’s trespassing onto foreign policy seems less troubling than the trespassing perpetrated by academics like Seitz. Imagine: While retrieving your car from the shop, your mechanic says you need to replace the timing belt. Then, cable news playing in the background, he starts opining on what the government must do to improve relations with Iran. Has he irresponsibly abused his authority and neglected his responsibilities toward epistemic dependents? No, but what’s the difference?

In this context, the content of his assertion makes it patently clear that he is no longer speaking as an expert, so he does not violate ENA. When there’s no reason to think the mechanic has foreign policy expertise, once he makes the policy assertion, any epistemic authority supporting his advice about your car clearly no longer applies. For extreme comparisons, imagine a parent with no military background first telling their child to go to their room and then telling their friend to invade North Korea or telling a fully autonomous adult friend to go to their room. In both cases, minimal contextual knowledge suffices to indicate that the person issuing the commands simply lacks the necessary authority. The same should be said of epistemic trespassing. Sometimes it’s perfectly obvious, even if only because of an assertion’s content, that the trespasser is not speaking as an expert even when they were a moment before. Once this is obvious in the mechanic example, there’s no risk of the mechanic’s genuine epistemic authority illegitimately imposing pressure on novices to defer about foreign policy. However, as we have seen, boundaries of expertise can be unclear and contested by alleged experts. In these cases, when experts trespass into nearby territory, the prohibition against trespassing derived from ENA applies because novices may reasonably interpret them to be speaking as experts.

5 Further defense

In this section, I further defend my account. This account satisfies Desiderata 1–4. By explaining what’s wrong with epistemic trespassing independently of whether experts acknowledge they’ve trespassed, it captures what’s wrong with trespassing, including in the motivating examples. Moreover, it diagnoses a problem with trespassing that prioritizes the social dimension of expertise, and distinguishes more worrisome from less worrisome cases of trespassing. Experts have responsibilities to others, in virtue of their expertise, that they neglect when they make assertions on questions that listening novices will be unable to discern as beyond their expertise.

Objection 1 It may seem that my account fails to satisfy Desideratum 1–2. This would be troubling, since one motivation for seeking an alternative to the higher-order defeat account is that it fails to satisfy these desiderata. This was not the only motivation. The higher-order defeat account misses the mark by focusing on problems for individual trespassers rather than novices potentially impacted by trespassing. Nevertheless, my argument would be significantly weakened if my account failed to satisfy these desiderata. So, it’s worth developing and answering this objection.

Desiderata 1–2 say, roughly, that accounts of what’s wrong with trespassing should apply to cases, like the ones motivating our inquiry, where trespassers fail to recognize their trespassing. However, my account may appear not to apply to those cases. After all, expert norms of assertions say experts should avoid asserting beyond their expertise. But how can you avoid this if you don’t know the limits of your expertise? In other words, can’t you avoid trespassing only by knowing the limits of your expertise? Moreover, I claim that experts abuse their authority when they trespass. But doesn’t this count as abuse only if the trespasser recognizes their trespassing? If this line of reasoning is correct, then perhaps my account and Ballantyne’s are not so different after all since they both insist that experts need to know their limits.Footnote 15

Three strands in this objection must be untangled. Each requires a different response. The first concerns how experts can conform to expert norms of assertions without knowing the limits of their expertise. I accept that reliably avoiding trespassing requires experts to know their limits, but I deny that this puts my account on par with the higher-order defeat account vis-à-vis Desiderata 1–2. That account is silent when experts fail to recognize their trespassing. It only says if experts recognize their trespassing, then they acquire a defeater. What if they don’t recognize their trespassing? The higher-order defeat account says nothing about these cases. In contrast, my account applies to these cases because it says if you trespass when speaking as an expert, you have done something wrong (which entails if you do this but don’t realize it, you have done something wrong). But, the objector might reply, how can you avoid speaking beyond your expertise without recognizing your limits? First, admitting that one has trespassed is different from recognizing one’s limits. Second, the assumption behind this question agrees with my account: failing to recognize one’s limits is part of trespassing’s wrong. Experts have a duty derived from novice dependence to know and not exceed their limits. When they trespass, they fail to discharge that duty whether or not they acknowledge their trespassing.

The second strand of the objection worries that trespassing cannot count as abuse of authority if one doesn’t recognize one’s trespassing. This seems like a mistake. People can unintentionally abuse their authority precisely because they are ignorant of its limits or do not realize they have exceeded those limits. Law enforcement officers unwittingly straying outside of their jurisdiction may unintentionally abuse their authority. Heads of state may abuse their authority without realizing it by issuing executive orders that exceed their state-granted powers. Experts, too, can abuse their authority without realizing it. Recall, on my account, trespassing is not merely an abuse of authority; it is neglectful abuse. Expert authority carries with it an obligation to attend to novices’ interests. Those with authority have an obligation to know its limits and to only exercise it within those limits. Failing to meet this epistemic obligation and then asserting beyond one’s expertise amounts to neglecting vulnerable epistemic dependents.

The third strand is that my account and Ballantyne’s end up looking similar since we both insist that experts must know their limits. I won’t rehearse all the differences between our accounts. But I will emphasize that agreement on this point is clearly consistent with disagreement on many others. We agree that preventing trespassing is a worthwhile aim achievable only when experts know their limits, while disagreeing about which flaws primarily explain this. Different ailments can have the same cure.

Objection 2 Another worry about my account can be derived from Ballantyne’s discussion of the physicist Erwin Schrodinger’s jaunt outside of physics when he published a book on the nature of life. Aware that he may be accused of trespassing, Schrodinger (1967: 1) gives voice to the objection that scientists must avoid trespassing as a matter of “noblesse oblige.” According to Ballantyne (2019a: 388), noblesse oblige is the idea that privileged people should act kindly and generously towards less privileged others. Ballantyne interprets Schrodinger as saying that scientists must avoid trespassing as a “matter of kindness” towards novices and experts in other fields. I agree with an appropriate characterization of the sentiment underlying this concept. But Ballantyne’s characterization may seem to diminish its importance. We should not characterize the responsibility experts have to novices as a mere kindness or generosity. The prohibition against epistemic trespassing, codified in ENA, is a duty derived from experts’ superior position in the social-epistemic order. Better than the nobility analogy is the command analogy. Someone with authority to issue commands has a responsibility to avoid issuing unauthoritative commands to those who will reasonably think they must obey. This isn’t mere kindness or generosity. It’s a matter of avoiding abusing authority and avoiding neglecting the interests of those who depend on you. The same goes for trespassing. Avoiding trespassing is necessary to avoid abusing authority and neglecting vulnerable dependents. It’s a matter of respect and what experts owe others.

Interestingly, the significance of this kind of account shines most brightly in cases where experts acknowledge their trespassing. I’ve ignored these cases to evaluate epistemic trespassing in the absence of acknowledgement. But consider these other cases. The expert knows they lack expertise, but they deliver alleged expert testimony anyway. This person not only abuses listeners’ vulnerability vis-à-vis first-order facts, he also exploits their ignorance about the limits of his expertise. He knows he’s not an expert. But they don’t. So he keeps talking, exploiting this ignorance. This case even more clearly illustrates how abusive, neglectful, and exploitative the trespasser’s behavior is. Ballantyne claims Schrodinger didn’t go far enough with noblesse oblige because he omitted the fact that experts are unkind to themselves. I agree with Ballantyne that a strong dose of epistemic modesty and self-knowledge is a helpful antidote to trespassing. And I’m enough of a pragmatist (or cynic) to endorse the strategic aim of helping experts realize that self-interest sometimes favors avoiding trespassing. But if we are explaining why experts shouldn’t trespass, especially knowingly, the emphasis on the individual trespasser and how unkindly he treats himself seems deeply misguided.

Objection 3 I say experts shouldn’t make assertions beyond their expertise when speaking as experts, but why not say, instead, that experts should cancel any implicature that they are speaking as experts when speaking outside their expertise? Rather than remaining silent, Tyson could say “I’m not a philosopher, but philosophy is useless.” In response, I admit that identifying the limits of your expertise before you trespass is preferable to trespassing without doing this. But altogether avoiding making these assertions when speaking as an expert is preferable still. I worry that the “conversational score” (Lewis, 1979) will not shift as easily as this objection assumes when the boundaries of expertise are unclear. The risk of harm is high enough that it’s probably worth staying silent. If the novice listener misses the expert’s delimitation of their expertise, the expert’s assertions may still sound like expert testimony. Even if the novice doesn’t miss it, they may still defer because they perceive residual authority behind the assertions. Experts should avoid asserting beyond their expertise when their assertions can be easily misconstrued as expert testimony. Too much can go wrong. Of course, this prohibition must always be weighed against the costs of remaining silent. But, all else equal, the prohibition is a good one.

Objection 4: Finally, one might argue that because the wrong I’ve identified is moral not epistemic, I’ve missed the epistemic target. I don’t find this objection compelling. Suppose I grant it. My goal was to identify what’s wrong with epistemic trespassing. Why think that wrong must be epistemic? Nevertheless, I don’t need to grant it. Speakers who violate epistemic norms like ENA, as trespassers do, fail epistemically. Moreover, an implication of the picture of epistemic dependence I have sketched is that moral and epistemic concerns are not easily disentangled. As Hardwig (1991: 708) writes, “many epistemic claims must meet ethical standards. If they cannot pass the ethical muster, they fail epistemologically.” That doesn’t guarantee that moral failures always translate into epistemic failures. Still, the distinction between these failures becomes less sharp when you prioritize trust and dependence in epistemic theorizing. Epistemic trespassing’s wrong may be both moral and epistemic.

6 Conclusion

Epistemic trespassing is wrong not merely because it can lead people astray or because it leaves a mess that genuine experts must clean up, and certainly not because it harms trespassers themselves. It is wrong because it is a neglectful abuse of authority. When experts speak as experts, novices typically must abandon their intellectual autonomy, surrendering more or less direct access to their minds, opening themselves up to epistemic and practical harms. Such vulnerability gives rise to expert responsibilities, like the responsibility to avoid haphazardly wielding one’s authority by speaking as an expert when one lacks that status. Epistemic trespassers neglect this responsibility. With their inexpert pronouncements, they risk illegitimately taking up an unearned position of authority in novices’ minds. Like other crimes of negligence, by risking this harm, whether or not someone is harmed, epistemic trespassers do wrong.