Inquiring for yourself for others

Abstract

Why should you inquire for yourself as a novice in a domain of inquiry when, for most questions within most domains, there are established experts to consult instead? In the face of this question, recent discussants of “autonomous-yet-novice” inquiry have sought to defend its epistemic value for the inquirer. Here I argue that autonomous-yet-novice inquiry can also be epistemically beneficial for agents other than the inquirer herself. Paradigm cases are those in which one agent improves her zetetic skills or virtues through an encounter or interaction with a more skillful or virtuous autonomous-yet-novice inquirer.

1. A puzzle about autonomous-yet-novice inquiry

Inquiring for oneself is often touted as a mark of epistemic value and courage, perhaps even a duty in the life of a well-informed citizen. But there is something puzzling about this, given our widespread epistemic dependencies on other people. As John Matheson puts it, “[t]he puzzle is in determining why it is epistemically valuable to think for yourself, since doing so will almost always not be the best available route to the answer to your question” (2022, p. 2). If you pick a domain of inquiry at random, many people will be more expert than you in that domain. Indeed, in most domains, you will not only be outclassed by many others – you will be an utter novice.

This might seem like an idle point whenever we cannot access expert testimony, but it is increasingly common in the modern era that we can access it. It can also be challenging, no doubt, for novices to identify the relevant experts (Millgram 2015; Levy 2022). But to the extent that this challenge is tractable, ¹ we clearly benefit from frequent expert testimony in our epistemic lives. ² The puzzle described by Matheson can be restated this way: why, amidst so many experts, should you engage in autonomous-yet-novice inquiry?

The puzzle dissolves if such inquiry is not really epistemically permitted, but this is “wildly counterintuitive” (Levy 2022, p. 5). A different thought is that autonomous-yet-novice inquiry is permitted to the extent that one is becoming an expert in a domain, but this point has a heavily limited scope; we are finite cognitive agents and hence cannot even attempt to become experts in most domains. Yet another thought is that novices can justifiably inquire for themselves by conducting “shallow research” (Ibid.). This is research in which one is thinking through an available range of expert testimony, sampling conclusions that have already passed through the crucible of expert scrutiny. In shallow research, there is no tension between deference to experts and one’s own zetetic efforts because “shallow research is guided by deference” (op cit., p. 6). However, the familiar injunction to “think for yourself!” or “do your own research!” (Ballantyne et al. 2022) seems to call for autonomous-yet-novice inquiry of a “deep” sort, whereby one does not simply wade through the shallows of expert testimony and, instead, delves into a domain’s raw evidential data. It is this sort of inquiry that I refer to as autonomous-yet-novice.

Matheson offers his own take on the present puzzle. He admits that one will rarely be justified in autonomous-yet-novice deep inquiry for the sake of knowledge, since the best path to knowledge will typically be expert testimony. However, autonomous-yet-novice inquiry may be a permissible or even obligatory route to further epistemic goods, such as:

1. (A) Enabling one to understand a truth, whereby one not only knows P but also understands the epistemic grounds for P (op cit., p. 9) ³

2. (B) Equipping one with the ability to “manage incoming defeaters” to P, this being a major upshot of understanding P over and above merely knowing P (op cit., p. 10)

3. (C) Enabling one to understand debates within a domain of inquiry despite potentially failing to acquire understanding or knowledge of the truths of that domain (op cit., p. 12)

4. (D) Providing answers to questions that one had not initially aimed to answer (op cit., p. 13)

5. (E) Generating “some novel way to approach some other question” or “some insight that aids us in answering” a question, other than the question one initially set out to answer (Ibid.)

6. (F) Providing an opportunity for one to develop virtues like epistemic humility (Ibid.)

Pursuing these sources of epistemic value as an autonomous-yet-novice inquirer is compatible with admitting that the best way to reliably acquire knowledge in a domain is often (even always) expert testimony (op cit., pp. 14–15). Notice, however, that these are sources of epistemic value for a single individual who gains understanding, manages incoming defeaters, acquires unanticipated information, or cultivates epistemic virtue through her autonomous-yet-novice inquiry. I flag this observation because the social-epistemic value of autonomous-yet-novice inquiry has received comparatively little attention thus far.

Matheson does acknowledge Bertrand Russell’s claim that an epistemically virtuous citizenry is socially valuable (1912/2016). What Russell and Matheson seem to mean is that the healthiest epistemic communities will be those with a high proportion of epistemically virtuous members, and so, autonomous-yet-novice inquiry, as a means of cultivating epistemic virtue, will simultaneously move us toward a social-epistemic good. ⁴ Additionally, Neil Levy suggests that marginalized groups must sometimes engage in autonomous-yet-novice inquiry to amplify their voices amidst non-marginalized experts (Levy 2022, p. 355). Beyond these few remarks, though, we hear little else about the social-epistemic value of autonomous-yet-novice inquiry.

One might get the impression that there is simply not much else to say here. For instance, Matheson wonders:

…how autonomous deliberation by a novice could have epistemic value for some larger group, particularly when there is already expert opinion to defer to on the matter…Even if it is important that there be some sort of a check on the relevant experts, it is harder to see that any one individual, particularly a novice, will provide such a benefit to society. (Matheson 2022, p. 8)

This concern is fair. However, focusing strictly on a societal scale obscures the possibility that autonomous-yet-novice inquiry might yield epistemic value at a smaller social scale, one as small as pairs of individuals. This is, indeed, the scale at which I will set my sights. My thesis is this: in suitable conditions, autonomous-yet-novice inquirers can provide epistemic value to others who encounter or interact with them. In a nutshell, this is because many agents have underdeveloped zetetic skills or virtues and can improve them through encounters or interactions with autonomous-yet-novice inquirers.

In §2, I will introduce various conceptual clarifications regarding our subject matter. I will then list three conditions under which social-epistemic benefits can be effectively actualized in small-scale social situations involving autonomous-yet-novice inquiry. Across §3–5, I will motivate these conditions. In §6–7, I will address some residual issues. In §8, I will conclude.

2. Conceptual clarifications and key conditions

I have advertised a view according to which an agent’s “encounters or interactions” with an autonomous-yet-novice inquirer can provide epistemic benefits for the former. Eventually, we will need to cash out these notions of encounter and interaction. To anticipate some pertinent questions here: does one agent benefit from another agent’s autonomous-yet-novice inquiry by observing that inquiry, or by receiving an explication from the autonomous-yet-novice inquirer, or by means of some sort of training from them, or through some sort of zetetic collaboration? I postpone discussion of these possibilities until §6 because my initial aim is to articulate the epistemic value for one agent’s encounter with another agent’s autonomous-yet-novice inquiry regardless of exactly which epistemic channel or channels transmit this value.

There are other clarifications that I cannot reasonably postpone. Among them, some remarks about the operative notion of autonomous inquiry are in order, after which I must also clarify the notion of autonomous-yet-novice inquiry. ⁵

For my purposes, autonomous inquiry occurs when an agent pursues deep inquiry within a domain, D, without assistance from experts with respect to the zetetic tasks of asking and answering D-constitutive questions. Such inquiry will qualify as deep rather than shallow insofar as the inquirer is not simply selecting between different expert opinions concerning D and is, instead, “getting the relevant first-order reasons for oneself, and evaluating them for oneself” (Matheson Ibid.). If experts happen to make one aware of potential first-order reasons for a belief, an autonomous inquirer can happily examine these reasons. What is forbidden is deference with respect to those reasons: one cannot give special weight to these reasons just because they come from an expert. Because of this, it also follows that the questions one takes up in autonomous inquiry are first-order as opposed to higher-order. They are questions about the subject matter of D itself rather than questions about whose testimony is to be trusted within D (op cit., p. 4).

Now, while shallow inquiry does not fall under my stipulative use of the term ‘autonomous inquiry’, indexing autonomous inquiry to particular domains does allow me to grant that an autonomous (hence deep) inquirer within D₁ may have engaged in shallow inquiry in D_2…n. This accords with contemporary insights from social epistemology, where it is now widely agreed that nobody is an epistemic island. If autonomous inquiry ever happens, it only sometimes happens (Matheson 2022, p. 3). Autonomous inquiry is compatible with the fact that inquirers frequently (if not always) rely on background knowledge when inquiring into a new domain or a new question within it. Specifically, if an agent inquires into D and relies on background knowledge B which was not acquired in autonomous inquiry, they still qualify as inquiring into D autonomously insofar as they are not assisted by any other epistemic agents in applying B to epistemic questions and answers constitutive of inquiring within D. Autonomous inquiry does not require Cartesian heroics: no global suspension of background belief is called for.

Given that autonomous inquiry does not require Cartesian heroics, is it nevertheless a solitary endeavor? I have hinted at a more interactive, collaborative possibility above. The idea of autonomous-yet-novice inquiry with which I am working forbids working with experts, in a sense, but not with other novices. One insight of the social and zetetic turns in epistemology is that agents who make good decisions about how to divide epistemic labor, and who bring their critical faculties to bear on one another’s contributions to joint zetetic projects, can be characterized as autonomous inquirers despite not inquiring solipsistically (cf. Grasswick 2018; Vega-Encabo 2021). Following Nathan King (2021, p. 88), Matheson writes that “epistemic autonomy requires thinking for yourself, not by yourself” (2024, p. 2). I accept this claim. For my purposes, then, what matters is that novice inquirers can collaborate and yet still be counted as autonomous when doing so. I return to this point in §6.

Crucially, I regard autonomous inquiry as skillful, whether conducted by novices or experts. This may sound strange when considering the novice case, but it need not. All that needs to be accommodated is some distinction between the zetetic skills possessed by novices and experts. On my view, what separates the novice from the expert is that the former possesses only domain-general zetetic skills, whereas the latter possesses domain-general and domain-specific zetetic skills. I also accept a similar distinction regarding virtue. However, I wish to focus on zetetic skills at this time, setting virtues aside for a moment, since they will be discussed less than zetetic skills in what follows.

Take domain-general zetetic skills first. Some of these are, we might say, investigative: they are general skills in the art of asking questions. For instance, one basic domain-general zetetic skill is the ability to step back from a chain of reasoning and ask, of one of its components, “is there evidence for this proposition?” Thinkers with poor domain-general investigative skills generally mismanage the frequency with which they pose questions of this form to themselves. Relatively more advanced domain-general zetetic skills include the ability to intelligently ask “is this evidence for P strong compared to the available counter-evidence?”, “how might I evaluate the quality of two different sources of evidence?”, and “should I double-check my answer to this question?” There are probably also meta-skills here, such as those that enable one to ask oneself when one is better off seeking new questions rather than meddling with old ones or suspending one’s inquiry until more potential evidence comes to light. Thinkers with poor domain-general investigative skills ask any number of these questions at inopportune times or fail to ask them at opportune times.

Other domain-general zetetic skills are, we might say, inferential: they concern knowing how to answer questions or to recognize potentially good answers. One can answer questions when one knows how to identify evidence as evidence, to compare propositional justifications, or to evaluate evidential sufficiency and so on. To be sure, one often needs knowledge of the specifics of a domain to assess evidential claims within it. However, there are still zetetic skills that extend across domains irrespective of different domains’ particular evidential standards and types. These skills are grounded in a facility for navigating general epistemic, zetetic, and logical norms. ⁶

I do not propose that all zetetic events are exercises of skill. For instance, inferences themselves are not skillful exercises. Drawing an inference is a psychological event that one cannot help but undergo when one judges one’s evidence to sufficiently favor a conclusion (Boghossian 2014). It exhibits no opportunity for modulation or control in the manner of skills (cf. Pavese forthcoming). Nevertheless, prior to drawing an inference during inquiry, there are all manners of activities that exhibit control via deliberate attention and careful scrutiny, such as deliberate examinations of evidence that eventually culminate in one’s drawing an inference. These are the sorts of skillful cognitive undertakings described by Galen Strawson as “catalytic” of inference (2003, p. 231).

The presence, absence, or strength of an agent’s domain-general zetetic skills might be gleaned by noticing an inquirer’s vulnerability to epistemic manipulation across unrelated domains. For example, C. Thi Nguyen (2021) observes that we often rely on feelings of clarity when considering claims and theories. These are eureka moments that lead us to stop our inquiry; they signal to us that we have thought about something enough. However, this clarity heuristic can be exploited. An unscrupulous manipulator might offer you a theory that maximizes feelings of clarity, even though it is an otherwise bad theory that further inquiry would (perhaps easily) have exposed. If an agent frequently over-relies on her clarity heuristic across unrelated domains, this is a clue that she is deficient in at least one domain-general zetetic skill, whatever it may be.

Now consider domain-specific zetetic skills. These are constitutive of one’s expertise (or, weaker, above-novice competence) in particular domains of inquiry, given that nobody is an expert in all domains. They are skills that enable one to seek or evaluate epistemic resources in contexts where some degree of insider knowledge of the contents and epistemic norms of a domain is a prerequisite for expertise (or, again, above-novice competence).

For example, an agent might be highly skilled at a strategy game that requires domain-specific competence to win consistently. In this game, knowing which questions to ask and which epistemic standards to take most seriously depends on one’s knowledge of the game’s mechanics and core concepts, such as those that carve out specific play patterns or metagame strategies. Suppose you are playing a game and an outside observer thinks that you are close to victory because you have 99 out of 100 possible points. You, on the other hand, know that the appearances can mislead because there are random-number-generating components of the game that could, at any moment, drastically reduce your score or supply an opponent with a massive score boost. Alternatively, you might know that, at a certain high level of play, your opponents will try to curb your access to strategic information and snatch victory from you at the last moment, whereas outsiders may not think about such possibilities at all. Your ability to navigate these factors, while inquiring about your next move, constitutes a set of domain-specific zetetic skills that the outsider lacks.

In what follows, I will frequently appeal to the distinction between domain-general and domain-specific zetetic skills. As we will see, it will allow us to characterize the notion of autonomous-yet-novice inquiry, and to understand the conditions under which autonomous-yet-novice inquiry can generate social-epistemic benefits. These conditions are threefold:

1. (1) The autonomous-yet-novice inquiry should be exercised via domain-general zetetic virtues or skills, the exercise of which can be appreciated by a less domain-generally zetetically virtuous or skillful agent.

2. (2) The autonomous-yet-novice inquiry should be genuinely engaged, i.e., undertaken in such a way as to potentially generate new epistemic goods for the inquirer.

3. (3) The autonomous-yet-novice inquiry should not be too difficult for meaningful progress to be made, nor can this progress be too difficult for any involved parties to comprehend.

Naturally, (1)–(3) invite further clarifications.

First, my use of ‘should’ in (1)–(3) is fundamentally a matter of epistemic prudence. Across §3–5 I will argue that autonomous-yet-novice inquiry in (1)–(3) creates significant epistemic opportunities for agents to improve their zetetic skills or virtues through encounters or interactions with an autonomous-yet-novice inquirer. I focus on these conditions because I take them to strongly facilitate the transmission of epistemic value between such agents, though I do not rule out the possibility of discovering further conditions that also do so.

I say that the operative ‘should’ is fundamentally rather than entirely prudential because, even though epistemic prudence norms for benefitting others through autonomous-yet-novice inquiry are my key focus, there are probably other epistemic norms that bear on autonomous-yet-novice inquiry as well. Specifically, there may be deontic epistemic norms that govern autonomous-yet-novice inquiry. Such norms might be negative or positive. Negative deontic epistemic norms undermine permission to engage in autonomous-yet-novice inquiry. I cannot ignore the possibility of such norms in this paper, since such norms may trump the prudential epistemic norms that are my key focus. I take up this issue in §7. ⁷

On the other hand, positive deontic epistemic norms will dictate obligations that we have to pursue autonomous-yet-novice inquiry (whether for our own benefit or for the benefit of others). Pursuing this project might be a matter of examining our epistemic obligations to serve as epistemic exemplars for others (Croce & Pritchard 2022, §4). However, this is not my project. Rather, I am addressing myself to a debate about the epistemic value of inquiring in an autonomous-yet-novice way. If my arguments succeed, they will show that autonomous-yet-novice inquiry has more social-epistemic value than hitherto appreciated in the literature.

Now, even though my project is not chiefly an exercise in epistemic deontology, I do not deny that autonomous-yet-novice inquiry may be on better epistemic footing if one exhibits intellectual or epistemic virtues in doing so. This is why my conditions (1)–(3) refer to the autonomous-yet-novice inquirer’s skill or virtue. Indeed, some say that autonomous inquiry is constitutively virtuous because epistemic autonomy a virtue (Matheson 2023).

As for the skill-virtue distinction itself, it has sometimes been argued that virtues are skills (Annas 1995), at least in the epistemic domain (Sosa 2007). Another view is that epistemic virtues are closely associated with epistemic skills (Zagzebski 2012). From where I sit, there are good reasons to follow Zagzebski and hence draw a separation between skill and virtue. One compelling observation here is that skills can be exercised without wisdom whereas virtues cannot be exercised without wisdom (McDowell 1979; Foot 2002, chapter 1). Similarly, one might think that virtuous inquiry must be properly motivated (Zagzebski Ibid.), whereas skillful inquiry need not be (Croce & Pritchard 2022, p. 585). Finally, there is Phillipa Foot’s observation that we are happy to call an agent skillful at ϕ-ing even if she deliberately fails to ϕ, whereas we are less happy to say that an agent is virtuous with respect to ϕ if she deliberately fails to ϕ (Foot 2002, chapter 1). ⁸ As aforementioned, my primary focus in this paper is zetetic skill, though I will occasionally make important points about zetetic virtue.

I now turn to the task of motivating conditions (1)–(3), in order to argue that autonomous-yet-novice inquiry can produce epistemic benefits not only for the initial inquirer (as Matheson argues), or one’s society (as Russell, Matheson, and Levy argue), but also for others to whom one stands in more immediate social-epistemic relations.

3. Autonomous-yet-novice inquiry (for others)

Consider (1) again:

1. (1) The autonomous-yet-novice inquiry should be exercised via domain-general zetetic virtues or skills, the exercise of which can be appreciated by a less domain-generally zetetically virtuous or skillful agent

As aforementioned, I believe that even a novice inquirer is domain-generally zetetically skillful. An agent lacking any zetetic skills whatsoever would scarcely count as an autonomous inquirer, or even as an inquirer at all. However, an agent’s lacking domain-specific zetetic skills in a domain, D, is not a necessary barrier to counting her as a potentially autonomous inquirer; she might still be an autonomous-yet-novice inquirer relative to D. She has a general capacity for asking questions that could further her inquiry, and has a general grip on how to think about differences in evidential quality, but lacks the capacity (at the outset of her inquiry, at least) to do so in ways characteristic of expert performances in D. There may be certain domains, D_a…n, to which one stands as too novice to make meaningful progress (see §5 below), such that one cannot even be an autonomous-yet-novice inquirer with respect to them, but this kind of failure must be ascertained on a case-by-case basis.

You might wonder how general these zetetic skills are. Are they universal, such that any domain-generally skillful inquirer must possess each of them? Maybe so, but it need not follow that they exist in equal measure for each agent. For instance, some inquirers might stumble into a new domain of inquiry, D, and be better at using their domain-general zetetic skills to acquire the more particular zetetic skills constitutive of good zetetic performances in D (more on these in a moment). Other inquirers might methodically plan out how they are going to approach D before actually attempting a zetetic performance in D. The former inquirers might be worse at the preparatory organizational tasks undertaken by the latter, whereas the latter might be worse at the more improvisational approach taken by the former, even if both inquirers still must possess all domain-general zetetic skills to some degree if they are to count as inquirers at all. ⁹

This point about differing degrees of domain-general zetetic skill between different agents is crucial for my purposes. Without there being such a difference, it is hard to understand how one agent could stand to learn much from another’s exercise of such skills in autonomous-yet-novice inquiry. Fortunately, as a general matter, skills admit of degrees. This point is unaffected by the possibility that domain-general zetetic skills are so general as to be (partly) constitutive of one’s epistemic agency. ¹⁰

Now, it is uncontroversial, I hope, that we often learn best from those who are more skillful than us at a task, which is why (1) articulates an asymmetry between the domain-general zetetic skills of different agents. To be sure, if two agents are equal in domain-general zetetic skill, it could happen that an agent’s autonomous-yet-novice zetetic conduct reminds a second agent to reflect on her own zetetic skills, thus sedimenting them further. This would indeed be an epistemic benefit to the second agent. My point, however, is that, all else being equal, less epistemic value is conferred here than what can be derived from interactions between asymmetrically domain-generally zetetically skillful agents. For, in the situation of asymmetrical domain-general zetetic skill, the less skillful agent could be both reminded to reflect on her own zetetic skills and come to recognize ways in which the autonomous-yet-novice inquirer applies those skills more effectively than herself.

With this said, agents may sometimes learn poorly from inquirers whose zetetic skills vastly outpace their own. One case of an inquirer’s being excessively zetetically superior to another agent is when the first agent is an expert with respect to D and the second agent is not. The non-expert is liable to get lost given that she entirely lacks the domain-specific zetetic skills possessed by the expert. But a similar lesson may apply even when we are focusing on pairs of autonomous-yet-novice inquirers, hence, when only domain-general skills are operative. This is because there may be cases in which an autonomous-yet-novice-inquirer has incredibly well-developed domain-general zetetic skills, so much so that a second agent is bewildered by the inquirer’s exercise of them. This suggests that (1) could be formulated slightly more strictly as:

1. (1*) The inquirer should have domain-general zetetic virtues or skills, the exercise of which can be appreciated by a less (but not too much less) domain-generally zetetically virtuous or skillful agent

With (1*) in view as a slightly refined condition for epistemically beneficial encounters with autonomous-yet-novice inquirers, I conclude my discussion for this section.

4. Genuine autonomous-yet-novice inquiry (for others)

I now turn to my second proposed condition. To reiterate:

1. (2) The autonomous-yet-novice inquiry should be genuinely engaged, i.e., undertaken in such a way as to potentially generate new epistemic goods for the inquirer

To make sense of (2), I will explain why it is better to engage in “genuine” autonomous-yet-novice than mere “ersatz” inquiry, at least if one’s inquiry is to confer epistemic benefits on others.

In ersatz inquiry, one merely “goes through the motions” of inquiring; one does not really take oneself to face an “open question” (Woodard, p. 4). For Elise Woodard, we can fix on this idea by not only distinguishing between genuine and ersatz inquiry but also between genuine and ersatz double-checking, where double-checking amounts to inquiry in which an inquirer already (a) has a “belief-like attitude” toward a proposition P and (b) inquires into that proposition despite being aware that one has this attitude toward P. Woodard’s example is of an agent, Eliza, who double-checks that her movie ticket purchase was successful just to appease an anxious friend Sandra, rather than to learn for herself whether it was successful (op cit., p. 3–4). Eliza does not acquire anything of epistemic value for herself. Contrariwise, if Eliza were double-checking whether she bought the tickets to make sure that she had done so, even while believing that she did buy them, this would be genuine double-checking and, hence, genuine inquiry. ¹¹

I intend condition (2) in such a way as to screen out ersatz inquiry, but why? What difference should it make from one agent’s point of view if they encounter or interact with another agent who is engaging in ersatz autonomous-yet-novice inquiry? Imagine, for instance, that you are being taught to play a strategy game and your friend undertakes an inquiry simply to demonstrate the sort of epistemic activity that you must undertake when you play, and not because the friend stands to learn anything new about the game. They engage in ersatz inquiry for your sake. Why is this not a source of epistemic value for you?

I do not deny that the preceding case can confer some epistemic value for you. However, it is arguably not autonomous-yet-novice inquiry that your friend has hereby performed, for it rather seems that experts (or at least above-novices) are those who can conduct ersatz inquiry in the first place, at least in many domains. It is not the sort of inquiry that is paradigmatically available to autonomous-yet-novice inquirers.

Moreover, and even if my first response is wrong, I contend that it is more beneficial for one agent to encounter a second agent’s autonomous-yet-novice agent’s inquiry when it is genuine rather than ersatz (hence my focusing on (2) rather than some other similar condition). This is because the inquirer is hereby challenged, and skills are typically more saliently exercised when they are challenged (more about this in §6). This is admittedly a delicate issue. If too many obstacles confront an agent’s zetetic activity, then she may buckle under their weight and fail to exercise her zetetic skills at all. But if she confronts no obstacles whatsoever, then she may enact her skills in such a trivial way that some of their finer features are not exercised, at least not saliently so. This will make it harder for other agents to appreciate one’s zetetically skillful exercises, and hence to epistemically benefit from them.

Here is a good opportunity to mention something about zetetic virtue. Plausibly, (autonomous-yet-novice) inquiry requires more epistemic courage when it is genuine than when it is ersatz. Similarly, the challenge presented by such inquiry may prompt the inquirer to reflect on her epistemic limitations more readily and to try, within reason, to apportion her efforts accordingly (Ballantyne et al. 2022, p. 12). This is a matter of epistemic humility. If epistemically virtuous people ever serve as exemplars for the rest of us, this must be partly because we can recognize their epistemic courage and humility in one way or another. Inquiry is acontext where this lesson applies. Thus, to the extent that autonomous-yet-novice inquiry is virtuous, other agents who encounter or interact with the inquirer may be bolstered in the development of their own epistemic virtues. ¹²

It might be objected that being a novice inquirer is irrelevant here. For, even if an inquirer is an expert in domain D, one might think that the expert can simply abstain from exercising her domain-specific zetetic skills in D. The expert might do this precisely to help another agent realize that progress can be made in D without relying on domain-specific zetetic skills. Arguably, however, an agent who has domain-specific zetetic skills but does not exercise them in D is a functional novice in D. The expert is proceeding as a novice does and so, for all intents and purposes, others will benefit from being made aware of the domain-general skills invoked in that autonomous-yet-functionally-novice performance.

5. Meaningful progress in autonomous-yet-novice inquiry (for others)

Finally, we turn to:

1. (3) The autonomous-yet-novice inquiry should not be too difficult for meaningful progress to be made, nor can this progress be too difficult for any involved parties to comprehend.

My use of the term “progress” is not particularly technical. I simply mean this: if a novice inquires into an excessively challenging domain, such that she cannot make any progress on her zetetic goals (be they knowledge, understanding, or something else), then she is not likely to exercise any skillful zetetic behavior in that domain. In virtue of this, others will not stand to benefit from encountering or interacting with that agent’s inquiry.

Two open questions are: whether there are many domains into which a given inquirer cannot make meaningful progress, and which domains these are. I, personally, am barred from making all kinds of progress when inquiring about biochemistry, but it may be possible to execute domain-general zetetic skills in the interest of learning some of this domain’s more interesting concepts. This might be a slow-going process, but it would not be one in which I was universally barred from making any meaningful progress. In fact, it may be difficult to conceptualize a domain into which a domain-generally skillful inquirer could never make any meaningful progress given infinite time, but we must recognize that opportunities for zetetic progress are time-constrained in our actual lives.

Let us now take stock and, in doing so, refine our picture. An autonomous, novice inquirer is domain-specifically unskilled with respect to a given domain, D, and yet domain-generally skillful as an inquirer nevertheless. She has general skills for asking and answering questions in accordance with epistemic, zetetic, and logical norms, and has some general idea of how to navigate the generic normative dimensions of different answers, but she is still a novice with respect to D because she has not yet figured out how to navigate its domain-distinctive contents and norms. In such cases, she is in a challenging position. She might need to carefully evaluate how to begin her inquiry, thus exercising her domain-general investigative skills or virtues, perhaps with a view to inquiring about how best to acquire the domain-specific zetetic skills or virtues characteristic of successful zetetic performances in D. Second, she might set out on the wrong foot or make a wrong turn during these challenges, thus needing to change the course of her inquiry. Meeting either type of challenge involves exercising domain-general zetetic skills or virtues. In one way or another, others can benefit from considering how the first agent navigates these challenges.

6. The epistemology of zetetic skill transmission

Across §3–5 I have focused on the value of an epistemic relationship between (at minimum) two agents: an autonomous-yet-novice inquirer and another person who in some way encounters or interacts with the first agent during autonomous-yet-novice inquiry. So far, I have left these notions of ‘encounter’ and ‘interaction’ largely unanalyzed, spare for the epistemic value that I take to arise from them. Some readers will now feel that we are owed further analysis. Our central question thus becomes: how exactly does the second person take epistemic advantage of the first person’s autonomous-yet-novice zetetic conduct, and in such a way as to cultivate greater domain-general zetetic skill or virtue? In taking up this question, I will not settle on a single answer. Instead, I will set out multiple possibilities that strike me as applicable to different contexts.

Two further questions will help us frame our discussion. First, must we understand the encounter as one between an autonomous-yet-novice inquirer and another agent who merely encounters the first agent’s zetetic activity? Or, second, can we understand the encounter in a more participatory way? While both possibilities might be characterized as generating a social-epistemic benefit, the latter is more ‘thickly’ social-epistemic than the former. We might put it this way: in ‘thin’ social-epistemic situations, one agent, A, can benefit epistemically from B’s undertaking of some autonomous-yet-novice inquiry I, without A’s being a participant in I, whereas in ‘thick’ social-epistemic tasks like joint inquiry, A benefits epistemically from undertaking I with B. By frequently using the disjunctive locution of one’s ‘encounter or interaction’ with an autonomous-yet-novice inquirer up to this point, I have deliberately made space for both thin and thick social-epistemic situations.

Let us take thin social-epistemic situations first. To rephrase, these are situations where one agent engages in autonomous-yet-novice inquiry and a second agent merely encounters this. Now, it is easy to beg the question in favor of a particular epistemology if one’s language is not chosen carefully. For instance, we could refer to this encounter as a matter of one agent’s observing another agent’s autonomous-yet-novice inquiry. However, if one interprets the language of observation too crudely, it will suggest that one agent benefits from another agent’s autonomous-yet-novice zetetic performance by merely perceiving the skills exercised therein. Skills, however, are arguably not perceivable themselves. A more plausible move would be to focus on the perception of skillful performances, but what do skillful zetetic performances look like from a third-person point of view?

This challenging question suggests to me that autonomous-yet-novice inquiry in thin social-epistemic situations ought to involve some vocalization of one’s zetetic considerations and moves: various explicitly vocalized questions or notes as one goes through the process. These can be taken as cues through which observers begin to recognize how certain questions and other zetetic performances (like evaluating pieces of evidence) are indexed to certain contexts at certain stages of a zetetic process. If the arguments of §4 are on the right track, the inquirer will be better at vocalizing her process when she is engaging in genuine rather than ersatz inquiry, since the challenges of successful inquiry will be more salient to her when she is not simply going through the motions. Either way, all of this will be of potential benefit to would-be observers. Indeed, this observational process may not be purely observational: it may also involve inquiry as to what the more domain-generally skillful autonomous-yet-novice inquirer is doing when she vocalizes her various zetetic steps and considerations.

We might also consider a hybrid epistemology: part observational, part inferential. However, because of how thinly social this all is (the one agent can undertake such observations and inferences without any direct interaction or participation in the autonomous-yet-novice inquirer’s zetetic processes), readers may wonder whether we have gone off track. After all, in the epistemological literature on pedagogy, it is common to talk of training as a central means by which one agent learns from another, and it is not immediately obvious how epistemological notions like observation and inference suffice to characterize the seemingly more interactive (hence thickly social-epistemic) notion of training.

Training is not a monolith, however, and indeed there seem to be some forms of training that are still relatively thinly social-epistemic. For example, one conception of training is purely instructional, and one sort of instruction is what I call brute zetetic instruction. In situations of brute zetetic instruction, you sit me down and talk to me about the various ways in which I might be a more domain-generally scrupulous question-asker and answerer. This is interactive to some degree. Will Small is skeptical, however, about whether skill can be transmitted from one agent to another on a purely instructional basis (2014). I suspect that this skepticism is better directed at the transmission of motor skills where propositionally encoded instructions are unlikely to play a sufficiently impactful role. When it comes to zetetic skills, things may be easier to manage, since these are partly a matter of recognizing the force of zetetic norms and acquiring a propositional, reflective appreciation of them.

Now, while I regard the brute zetetic instruction account of zetetic training as coherent, it is not likely universalizable, nor is it obviously a dominant model for my purposes. The reason is that the ontogeny of zetetic skill simply does not seem to come solely from sharing zetetic and epistemic instructions with a largely passive trainee. Thus, a thicker picture is possible, wherein the trainee herself contributes to the zetetic process. One such alternative is backseat zetetic guidance. In such a case, a trainee might inquire into D while a second – more skillful – agent steps in to provide backseat zetetic guidance at opportune moments, dropping hints here and there as to how the trainee might successfully move their inquiry forward. Arguably, this is a more compelling picture from an ontogenetic point of view (Croce & Pritchard 2022; Small 2014).

However, backseat zetetic guidance is not as social as things could get, and presumably does not perfectly dovetail with the ontogeny of zetetic skill-acquisition. Rather, the thickest social-epistemic scenario of interest to me involves both agents engaging their domain-general zetetic skills to the best of their abilities with respect to a domain D, where neither agent is engaging in backseat zetetic guidance on the other’s behalf, and yet one agent is still more domain-generally zetetically skillful than the other, such that the one party stands to benefit epistemically not by merely encountering the former’s zetetic conduct, but by doing so whilst contributing in her own way to a shared zetetic project. We can call this joint inquiry.

Joint inquiry is clearly something in which we engage, and involves autonomous epistemic capacities on the part of all participants (Grasswick 2018). It is also clearly thickly social-epistemic. However, I think it is myopic to analyze the arguments of §3–5 solely through the lens of joint inquiry. More plausible, I think, is that the road to a learner’s acquiring and refining her domain-general zetetic skills is long and winding, with only some stages manifesting the learner’s capacities for joint inquiry. Think, by analogy, of how a father might oscillate back and forth between showing his child how to knead some dough, commenting on the child’s own kneading of the dough, holding the child’s hands as the child kneads the dough, and kneading the dough while the child also does so. These are all methods for improving the child’s kneading skill, ranging from least interactive to most interactive. In zetetic skill-acquisition, things may be similar: one domain-generally skillful inquirer may help a less domain-generally skillful inquirer to deepen her skills through a range of epistemic channels, not all of which we characterize as thickly social, but all of which have some sort of social-epistemic benefit simply in virtue of involving a transmission of epistemic goods from one agent to another. ¹³

One can see, then, that I have not settled on a single epistemology concerning how exactly an agent learns from encountering another agent’s domain-generally zetetically skillful (i.e., autonomous-yet-novice) performances. This is deliberate, for I take there to be a plurality of ways in which to do so – some thinly social-epistemic and some more thickly so. Those impressed by the depth of the social-epistemic turn in contemporary epistemology would presumably invite us to focus primarily on thicker social-epistemic contexts, such as joint inquiry, since such contexts were formerly obscured by the dominance of ruggedly individualistic epistemological paradigms in philosophy. However, I take it as a benefit of my discussions here that more passive epistemic relationships between agents are still potential grounds for growth on the part of more novice inquirers.

7. Concerns and objections

I now round out my discussion by considering a few more concerns for my arguments in this paper.

7.1. Autonomous-yet-novice inquiry for others: Deep or shallow?

In §1, I noted the distinction between shallow and deep research. To reiterate, shallow research engages with second-order evidence, i.e., sorting through expert opinions on first-order evidence, whereas deep research requires foregoing deference to expert testimony to sift through the first-order evidence for oneself. Throughout §3–5 I focused on the latter as the central undertaking of autonomous-yet-novice inquirers. But are there differences in epistemic value for a second agent when she encounters or interacts with the first’s autonomous-yet-novice inquiry? The question matters because one possible answer would threaten to show that there is comparatively less value in observing an agent’s autonomous-yet-novice deep inquiry.

My first reply repurposes a virtue-epistemic suggestion from §4: deep inquiry requires more courage and presents greater opportunities for epistemic humility than what can be found in shallow inquiry. It thus ensures greater opportunity for manifesting these virtues in a manner that can be appreciated by other participants to or observers of such inquiry.

This is also a point at which we can helpfully draw on Matheson’s work once more. In §1 we saw that it is possible to recognize sources of epistemic value in autonomous-yet-novice deep inquiry, even if there is just one agent involved. To recall, even if a single agent’s best route to knowledge is to defer to expert testimony, or even to engage in shallow autonomous-yet-novice inquiry, there are other epistemic goods to be gained if they go deep: understanding, effective management of incoming defeaters, and so on. If these desirable epistemic ends are reasonable for a single novice to pursue, then whatever zetetic skills or virtues are exercised in this pursuit can also, potentially, be exercised in the company of others, and hence to the epistemic benefit of others.

7.2. Obligations to avoid autonomous-yet-novice inquiry?

It is possible that one agent’s autonomous-yet-novice inquiry will not be appreciated in the right ways by other agents, or that others may pick up on the wrong things. A big-picture worry for the arguments of this paper is that the risks of autonomous-yet-novice inquiry may be so significant as to warrant a blanket prohibition on this form of zetetic activity. If such a prohibition exists, this will mean that the prudential epistemic norms favoring autonomous-yet-novice inquiry (for one’s own sake or for others’ sake) are trumped by deontic epistemic norms against it.

What argument could establish such a prohibition? ¹⁴ One argument might be that autonomous-yet-novice inquiry is prohibited to the extent that expert testimony is universally superior to novice testimony. However, I never argued that autonomous-yet-novice inquiry yields value for others due to its potential testimonial upshots. The value pertains to the skillful or virtuous behavior that can be emulated as a result – it is the zetetic process, not the product, that yields the epistemic fruit in which I have been interested. I do not dispute that the autonomous-yet-novice inquirer is obligated to attend to the expert’s testimony over the novice’s testimony, but I have already argued attending to the expert’s zetetic process is not always the better choice (§3, §5). Moreover, epistemically valuable access to experts’ zetetic processes is arguably scarcer than access to many novices’ zetetic processes, precisely because it is difficult to recognize the domain-specific zetetic skills of experts when one lacks these skills (cf. Millgram 2015).

Another argument for a blanket prohibition on autonomous-yet-novice inquiry is that we bear an epistemic obligation to care about the quality of our society’s research norms and hence to defer to others instead of inquiring as an autonomous novice. This could be a damning objection if the autonomous-yet-novice inquirer were chiefly motivated by the thirst for knowledge. But if the inquirer is motivated by epistemic goods that do not plausibly allow for deference, such as understanding or the capacity to engage with potential defeaters for one’s knowledge (again, see §1), then deference is not what is obligated, and hence autonomous-yet-novice inquiry may still be permitted. ¹⁵

A final objection is that we might adopt other agents’ zetetic vices by attending to or participating in their novice inquiry, such that we are obligated to avoid such attention or interaction. Recall, however, that condition (1) acknowledges a role for virtue as well:

1. (1) The inquirer should have domain-general zetetic virtues or skills, the exercise of which can be observed by a less zetetically virtuous or skillful agent

To more firmly block the present objection, perhaps we should reconstrue the disjunction ‘zetetic virtues or skills’ as the conjunction ‘zetetic virtues and skills’. If an autonomous-yet-novice inquirer is not just domain-generally zetetically skillful but also zetetically virtuous (Matheson 2023), then the risk of ‘vice-transmission’ is mitigated. Of course, care will still need to be taken in order to discriminate between intellectually virtuous and vicious inquirers, but similar lessons are true beyond the zetetic realm: we must always take care to distinguish the virtuous from the vicious. Unless we want to throw out the very idea of improving ourselves by emulating virtuous individuals, we should not endorse a blanket prohibition against emulating virtuous, autonomous-yet-novice inquirers (cf. Croce & Pritchard 2022). ¹⁶

8. Conclusion

Should we engage in autonomous-yet-novice inquiry? If the target epistemic good is knowledge, then perhaps we should defer to the experts instead. If the target epistemic good is something other than knowledge, then autonomous-yet-novice inquiry may serve one well (Matheson 2022). Not only that, it may serve others well when we engage in autonomous-yet-novice inquiry. This is because doing so provides opportunities for emulation, discussion, or interaction which can enable others to improve their zetetic virtues and skills – precisely those domain-general zetetic virtues and skills that underwrite autonomous-yet-novice zetetic performances. I conclude that there is social-epistemic value to be found in autonomous-yet-novice inquiry at an interpersonal scale, hence not just for individual inquirers or the larger societies to which they belong.