1. Introduction
Outsourcing beliefs is necessary. If we were only to believe the things that we know through individual investigation – through sources of justification such as perception and reason – then our catalog of beliefs would be, as Robert Audi puts it, “at best impoverished” (Audi (Reference Audi2011), 150). Many of the beliefs that we acquire as young children are based solely on the testimony of others. When my sons were young, they regularly asked me to read What Do People Do All Day? – a book in which Richard Scarry cleverly names and describes everyday occupations. Whether they were accepting the testimony of me or Richard Scarry, their beliefs about occupations were formed on the basis of testimony.
Forming beliefs on the basis of testimony is a lifelong endeavor. While education ideally involves the critical engagement and exploration of the most important questions humans ever ask (What is the meaning of life? Does God exist? What is a human person? What is human flourishing?), it also certainly involves the transfer of information from credentialed experts to their students. When Tom Flint taught me modal logic, I accepted the axioms on the basis of his testimony.
Moreover, many of the beliefs that I require to carry out my day-to-day life are based on the testimony of others. When the mechanic tells me that my Honda needs a new timing belt at 100,000 miles, I form the belief that my car needs an appointment for this service. When my doctor diagnoses me with a sinus infection, I take the prescribed antibiotic. I have no idea what a timing belt is or how my doctor diagnosed the infection, but I form the relevant beliefs on the basis of their expert testimony, and I am inclined to think that I am rational in doing so.Footnote 1
In Bad Beliefs, Neil Levy appeals to empirical studies that purport to demonstrate that human flourishing requires “our capacity to engage in distributed cognition.” I will leave it to the reader to evaluate the social science data. I think the intuitive examples listed above are sufficient to support this relatively uncontroversial claim. In the spirit of Aristotle, Levy describes human beings as rational social animals (Levy (Reference Levy2022), 61). He adds the social to emphasize the way in which we must outsource our beliefs. Failing to do so would impact our flourishing – the sure result of an impoverished belief system. But Levy emphasizes that the outsourcing of beliefs to others is not only necessary for flourishing, but it is rational, which is just to say it is epistemically good.
So what’s the problem? When we form beliefs based on the testimony of others, which I agree that we must do, our justification for those beliefs is not derived from examining the direct evidence for those beliefs. Instead, our justification is derived from testimony. In many cases, the testifier will have examined the direct evidence for her beliefs; her justification for the beliefs is based on this evidence. But as the receiver of her testimony, my justification for the beliefs is her testimony. Which brings us to the problem – whose testimony should I trust? To whom would it be rational to outsource my beliefs?
One approach to solving this problem originates in virtue epistemology. A prominent proponent of this approach is Linda Zagzebski, who applies the virtue tradition in ethics to belief formation. In Virtues of the Mind, Zagzebski suggests that good, rational belief formation is a species of good actions. She notes that just as there are virtues in other domains, there are intellectual virtues aimed at knowledge and reliably successful at achieving this aim. In this paper, I will explore the role (if any) that intellectual virtues might play in rational belief formation on the basis of testimony. I will begin by considering Zagzebski’s proposed intellectual virtue of being able to recognize reliable authority. I will argue that this quality, which is surely an excellence, is better categorized as a skill than a virtue. Then I will explore whether other intellectual virtues contribute to one’s ability to assess the reliability of a testifier. I consider two ways in which they might contribute: (1) by enabling one to directly assess the reliability of the claims of the testifier and (2) by enabling one to indirectly assess the reliability of the testifier. With respect to (1), I will agree with Neil Levy and argue that employing the virtues to directly assess the claims of the testifier requires like expertise to the testifier (which is impossible to have in every domain). Moreover, to employ the virtues to directly assess the claims of the testifier is to justify those claims on the basis of evidence, not testimony. With respect to (2), I will argue that intellectual virtues do enable one to identify second-order features of testifiers that indicate reliability. Against my contention, Levy argues that evaluating these second-order features of testifiers is impossible to do in our polluted epistemic environment. In response, I will suggest that the intellectual virtues have the added benefit of helping us identify and escape pollutants in our epistemic environment. Thus, the role of intellectual virtues in rational belief outsourcing is aimed primarily at identifying second-order features of testifiers and good social structures for belief formation. Given Levy’s observation that we are rational social animals that must engage in distributed cognition, this role for intellectual virtues is of great importance.
2. On the virtue of being able to recognize reliable authority
One way in which the intellectual virtues might support rational belief outsourcing is if there is a virtue of being able to recognize reliable authority. Zagzebski includes such a virtue in her initial list of candidates for intellectual virtues.
Intellectual Virtues:
-
• The ability to recognize the salient facts; sensitivity to detail
-
• Open-mindedness in collecting and appraising evidence
-
• Fairness in evaluating the arguments of others
-
• Intellectual humility
-
• Intellectual perseverance, diligence, care, and thoroughness
-
• Adaptability of intellect
-
• The detective’s virtues: thinking of coherent explanation of the facts
-
• Being able to recognize reliable authority
-
• Insight into persons, problems, theories
-
• The teaching virtues: the social virtues of being communicative, including intellectual candor and knowing your audience and how they respond (Zagzebski (Reference Zagzebski1996), 114, emphasis mine)
While undoubtedly it would be an epistemic good to be able to recognize reliable authority, is it the sort of quality that we would consider a virtue? To determine if this quality is a virtue or something else, I will begin with Zagzebski’s Aristotelian account of a virtue in general, and of intellectual virtue in particular.
2.1. Zagzebski on virtue
Zagzebski defines a virtue as “a deep and enduring acquired excellence of a person, involving a characteristic motivation to produce a certain desired end and reliable success in bringing about that end” (Zagzebski (Reference Zagzebski1996), 137). Across the vast literature on virtue, there is little disagreement that virtue is an excellence. The lists of candidate virtues vary widely; they range from traditional moral excellences such as courage to Aristotle’s social excellences such as wit (Aristotle (1999) 27) and Hume’s practical excellences such as discretion (Hume (Reference Hume1983), 53). But the notion of a virtue as an excellence is common. Moreover, by “deep,” I take Zagzebski to be highlighting the habitual and enduring nature of the virtues. Once acquired, they are not easily lost; the exercise of the virtues in the relevant contexts arises out of habit for the possessor of the virtues.
Moreover, Aristotle describes both moral and intellectual virtues as acquired excellences in contrast with natural dispositions. “Virtue, then, is of two sorts, virtue of thought and virtue of character. Virtue of thought arises and grows mostly from teaching; that is why it needs experience and time. Virtue of character…results from habit” (Aristotle (1999), 18). Zagzebski notes, it is important to construe a virtue as acquired if we are to be held morally responsible for our character. It seems unreasonable to praise or blame a person for traits that they possess naturally, which are outside of their control (Zagzebski (Reference Zagzebski1996), 102–103). While the acquisition of a virtue is not entirely under an individual’s control, they are character traits developed through habituation. We engage in practices that cultivate virtues (or vices).
On Zagzebski’s definition, a virtue is an acquired excellence with two central components: motivation and success. According to Zagzebski, a motive is “an emotion or feeling that initiates and directs actions towards an end” (Zagzebski (Reference Zagzebski1996), 131). For example, an open-minded person has the following motive: an emotion or feeling such as delight in discovering new truths that initiates and directs her toward the end of considering diverse perspectives. When motives such as these operate persistently as opposed to for brief intervals of time, they are what Zagzebski calls motivations (Zagzebski (Reference Zagzebski1996), 132). Virtues involve motivations – persistent emotions that initiate and direct actions towards an end. So the virtue of open-mindedness involves the persistent disposition to delight in discovering new truths that initiates and directs one toward the end of discovering diverse perspectives. While the intellectual virtue of open-mindedness has a particular end that is associated with this virtue, namely, discovering diverse perspectives, Zagzebski notes that all the intellectual virtues ultimately aim at the end of knowledge (Zagzebski (Reference Zagzebski1996), 167).
In addition to a motivation component, virtue also includes a success component. “Virtue possession requires reliable success in attaining the ends of the motivational component of virtue” (Zagzebski (Reference Zagzebski1996), 134). Returning to the virtue of open-mindedness, the person with this virtue aims at discovering diverse perspectives and is reliably successful in achieving this aim. Since the intellectual virtues ultimately aim at knowledge in addition to the specific aims of the virtue, possession of intellectual virtues produces reliable success in attaining knowledge. Thus, those possessing the virtue of open-mindedness reliably achieve success in discovering diverse perspectives, which in turn enables the possessor to more reliably achieve success in attaining knowledge.
2.2. Virtues related to testimony
Above, I noted that Zagzebski identifies being able to recognize reliable authority as an intellectual virtue. In the several lists of intellectual virtues given throughout the history of philosophy, this virtue does not make a frequent appearance. Open-mindedness or fairness or intellectual diligence, care, and thoroughness are far more common. This uncommon intellectual virtue is not the only virtue related to testimony that makes an appearance in contemporary epistemology. Robert Audi suggests a virtue of trust. “The absence or laxity of filtering beliefs yields credulity; the presence of excessively rigorous ones yields skepticism. Intellectual virtue—and epistemic responsibility conceived as a kind of virtue—are attained when we achieve the ‘mean’ between excessive credulity and unwarranted skepticism” (Audi (Reference Audi2011), 152). While this is a candidate virtue that is relevant to forming beliefs on the basis of testimony, it is distinct from being able to recognize reliable authority. One could cultivate a virtue of trust and exercise this virtue in a particular situation but exercise it without simultaneously exercising the virtue of recognizing reliable authority. For example, a patient might lack the virtue of recognizing reliable authority; he could not distinguish a reliable doctor from an unreliable doctor. Nonetheless, if his doctor is in fact a reliable authority, the patient shows proper trust in his doctor when he is not overly lax or stringent in filtering the beliefs he receives from the doctor. Thus, trust is a distinct virtue from the ability to recognize reliable authority, though both clearly relate to the formation of beliefs based on testimony.
Though I have argued that trust is a distinct virtue from recognizing reliable authority, it does seem that trust is only a virtue when trust is in fact placed in reliable authority (whether or not this is recognized). In fact, the vices on the extremes may constitute a virtuous response depending on the reliability of the authority. In the face of unreliable authority, skepticism is the virtuous response though it would be a warranted skepticism. Moreover, in the face of a reliable authority, something closer to credulity is the virtuous response; there is no reason to filter the beliefs one is willing to accept from a reliable authority.Footnote 2 Thus, the virtue of trust seems to require the virtue (if there is one) of being able to recognize reliable authority if we are to exercise it wisely.
Another virtue related to testimony that is suggested by Zagzebski is the mean between the extremes of being too reliant on authority and too intellectually independent (Zagzebski (Reference Zagzebski1996), 97). She does not name this virtue; let’s call it proper intellectual dependence. This intellectual virtue acknowledges both our need to outsource some of our beliefs and the intellectual immaturity of outsourcing all of them. While most of the beliefs of a young child might be outsourced, we will deem a university student who has not begun to investigate any claims for herself as intellectually stunted. Whether proper intellectual dependence should count as a virtue is outside the scope of my investigation. One thing that is clear is that this virtue is distinct from the virtue of being able to recognize reliable authority, though both relate to the formation of beliefs based on testimony.
2.3. On the virtue of being able to recognize reliable authority
Given Zagzebski’s account of virtue as an acquired excellence consisting of both a motivation and success component, what might the virtue of being able to recognize reliable authority look like? Surely being able to recognize reliable authority would be an excellence. Moreover, if being able to recognize reliable authority is an acquired characteristic as opposed to a natural inclination or talent, then there must be some practices that cultivate it (more on this later). The motivation behind this virtue might be construed as the persistent disposition to acquire new knowledge beyond one’s expertise that initiates and directs one toward the end of carefully recognizing reliable testifiers. Furthermore, the possessor of this virtue will reliably achieve this goal and, ultimately, the goal of all intellectual virtues, which is knowledge.
So is there a virtue of being able to recognize reliable authority? I will argue that while such a quality would be an epistemic good, it is not properly called a virtue. First, I will suggest that being able to recognize authority is closer to Aristotle’s concept of a skill than a virtue. Second, I will argue that the means by which we cultivate being able to recognize reliable authority is distinct from the means by which we cultivate a virtue. Third, I will argue that while being able to recognize reliable authority will reliably achieve the goal of virtue and the goal of knowledge, it does so tautologously.
In Nicomachean Ethics, Aristotle compares and contrasts virtues and techne (skills or crafts). Contemporary virtue theorists have offered several distinctions between virtues and skills.Footnote 3 Phillipa Foot offers the following distinction: “A virtue is not, like a skill or an art, a mere capacity; it must actually engage the will” (Foot (Reference Foot1978), 8). A skill does not need to be exercised in order for it to be possessed. But if a virtue is possessed and the circumstances arise in which it would be properly exercised, it must be exercised, or it is not genuinely possessed. Gilbert Meilander offers the following illustration of this distinction: “If I deliberately miss a baseball pitched to me, it does not show that I lack the skill to hit it. But if while playing baseball I deliberately treat the opposing team unjustly, this does indicate that I lack a certain virtue” (Meilander (Reference Meilaender1984), 9). Hitting the baseball is a skill that we can choose not to exercise and still have the skill; acting justly is not.Footnote 4 Likewise, Sarah Broadie notes that a virtue is done from a “firm and unchangeable state” (Broadie (Reference Broadie1991), 89). This distinguishes virtues from skills, according to Broadie, for “it says nothing against a person’s skill if he fails to exercise it in the face of distractions or with someone begging him not to” but, presumably, it does say something against his virtue in this case as it would not then reflect a firm and unchangeable state (Broadie (Reference Broadie1991), 89).
Is being able to recognize reliable authority more like a skill or a virtue in this regard? Should a person have the quality of being able to recognize reliable authority and find herself in a circumstance in which a judgment about a testifier is necessary but choose not to make the judgment either because she is distracted or being coerced not to make such a judgment, it does not impugn her ability to recognize reliable authority. She may choose not to use the ability in this circumstance, but she still has the capacity to recognize reliable authority. If such a quality were a virtue, a failure to exercise it in the relevant circumstances would indicate a lack of the quality. Because failure to exercise it does not indicate a lack of the ability to recognize reliable authority, I argue that this quality is closer to Aristotle’s notion of a skill than a virtue.
James Wallace distinguishes between a skill and a virtue in that the former is the mastery of a technically difficult action, whereas a virtue does not require a technically difficult action. “The actions that a skill is a capacity to do well are things that are difficult to do well—at the very least, difficult at first for most people” (Wallace (Reference Wallace1978), 44). Virtues do involve difficult actions, but the difficulty arises because we have “contrary inclinations” (Wallace (Reference Wallace1978), 46). Consider the skill of performing a mathematical calculation. It is technically difficult; we must learn the technique from a teacher. But it is not technically difficult to do a courageous action. The difficulty arises because fear inclines us to avoid situations that require courage. Likewise, it is not technically difficult to perform actions associated with intellectual virtues. Anyone can be open-minded though we resist open-mindedness because we are inclined to avoid those who disagree with us.
Given this further distinction between skills as technically difficult and virtues as difficult because of our “contrary inclinations,” is the ability to recognize reliable authority better understood as a skill or a virtue? The ability to recognize reliable authority is not merely a superpower that we acquire through practice, a reliability detector of sorts that applies across domains. Instead, the ability to recognize reliable authority requires assessing the expertise of the testifier or other second-order qualities of the testifier such as attention to detail. This is technically difficult, especially when the domain of expertise of the testifier is outside of our own. Given that our need to form beliefs based on testimony is due to our inability to be subject matter experts about everything, we will find it technically difficult to assess expertise in any situation in which forming beliefs based on testimony is required. Moreover, I do not have a contrary inclination to being able to recognize reliable authority – it seems greatly to my advantage to be able to do so.
A final relevant distinction between a skill and a virtue is made explicitly in Zagzebski, though there are hints of the distinction in Wallace. Zagazebski writes, “On all accounts, a vice is the contrary of a virtue, not its contradictory, but a skill has no contrary” (Zagzebski (Reference Zagzebski1996), 112). Thus, while cowardice and foolhardiness are the contraries of courage, they are not contradictory to courage. However, in the case of a skill, there is just its contradictory. One either has the skill of performing a particular mathematical calculation or lacks the skill; there are no contrary states to having such a skill. Wallace puts this distinction as follows: “Skills are simply capacities to do things that are technically difficult, and the “opposite” of a skill is only its absence” (Wallace (Reference Wallace1978), 50).
In applying this distinction between virtues and skills to the ability to recognize reliable authority, it again seems to fall into the skill category. There is no excess or deficiency associated with the ability to recognize reliable authority. One either has this quality or lacks it – the having of it is certainly an excellence, but it seems that it is the sort of excellence that we associate with skill and not virtue.
I have canvassed three distinctions between a virtue and a skill; there are additional proposed distinctions, but these three should suffice to establish my claim that the ability to recognize reliable authority is properly categorized as a skill, not a virtue. But for the unconvinced, there are additional problems with categorizing the ability to recognize reliable authority as a virtue. A second problem is that it is not acquired in the way in which we acquire a virtue. According to Aristotle, virtues are acquired through doing the actions that a person with the relevant virtue would do. Thus, the virtue of justice is cultivated by doing just actions; the virtue of bravery is cultivated by doing brave actions. Likewise, intellectual virtues such as fair-mindedness are acquired through doing the actions that a fair-minded person would do.
In order to cultivate virtues in this way, we must (1) be able to identify the person with the virtue and (2) be able to perform the actions of the person with the virtue without yet having the corresponding virtue. For example, if I want to cultivate courage, I must be able to identify the courageous person, and I must also be able to perform courageous actions without yet being courageous. This seems possible. I can identify that the person who rescues a fellow soldier who is injured on the battlefield is displaying courage. And I can go and do likewise even if I don’t yet do it courageously. Can the same be said for recognizing reliable authority? If I want to cultivate the alleged virtue of being able to recognize reliable authority, I must be able to identify the person with this virtue and perform the actions of this person without yet having the virtue. But this seems impossible. How would I identify the person with this virtue? Everyone around me is performing the action of judging the reliability of testifiers. How would I recognize the person who is doing it well? It seems that to be able to identify this person, I must already have the ability to recognize reliable authorities because the person who is judging the reliability of testifiers well is the one who is correctly identifying reliable authorities. Thus, it is impossible to identify a person with the virtue of being able to recognize reliable authorities without having the virtue myself. Moreover, it is also unclear how I could practice the action of judging the reliability of testifiers well without yet having the virtue. I can practice judging the reliability of testifiers, but to practice doing it well, I must already know which testifiers are reliable.
One might reply that I have misidentified the action of the person with the virtue of recognizing reliable authority. Instead of judging the reliability of testifiers, a person with this virtue is carefully scrutinizing the authority of testifiers. While everyone around me is judging the reliability of testifiers, the one’s engaging in this kind of scrutiny are doing it well. On this understanding of the action, I can identify the person with the virtue of being able to recognize reliable authority without already having the virtue myself. However, if this is the action of the person with the virtue of recognizing reliable authority, I do not think that I can perform the action without yet having the corresponding virtue. It seems that the doing of this practice just is having the ability to recognize reliable authority. Unlike a virtue, I do not have to do it in a certain way in order to have the quality of being able to recognize reliable authority. Given that cultivating virtues requires both the ability to identify the person with the virtue and the ability to perform the actions of the person with the virtue without yet having the corresponding virtue, one cannot cultivate the alleged virtue of being able to recognize reliable authority.
A third problem for categorizing the quality of being able to recognize reliable authority as a virtue is that it reliably achieves success tautologously. If virtues are supposed to create the conditions under which we best achieve our aims, then the quality of being able to recognize reliable authority does this too quickly. As I noted above, a virtue is reliably successful at achieving both the aim of the virtue itself and the ultimate aim of virtue. So the intellectual virtue of open-mindedness reliably achieves success in discovering diverse perspectives, which in turn enables the possessor to more reliably achieve success in attaining knowledge. In this case, the virtue does not, by definition, guarantee that knowledge is achieved. Instead, the virtue enables its possessor to discover diverse perspectives, and this sort of discovery is a part of the conditions out of which knowledge is achieved. However, in the case of the ability to recognize reliable authority, the possessor of this virtue seems to automatically achieve the end of recognizing reliable testifiers. Assuming that reliable testifiers are good sources of knowledge (what else would make them reliable testifiers?), success in achieving the goal of knowledge is built into the alleged virtue. Just as we would not consider knowledge itself a virtue which aims at knowledge, so we should not consider being able to recognize reliable authority as a virtue which aims at knowledge since knowledge necessarily results from this quality.
3. Intellectual virtues and assessing the reliability of testifiers
If, as I have argued above, there is no virtue of being able to recognize reliable authority, then we must look for another way in which the intellectual virtues could support rational belief outsourcing. I will now turn to a second possibility: could the other intellectual virtues be valuable in assessing the reliability of testifiers?
In her 2012 monograph entitled Epistemic Authority, Zagzebski offers an account of trusting authority that is grounded in self-trust. She argues that I rightly trust my epistemic faculties when I conscientiously form beliefs. This self-trust is the basis for trusting an authority. If I conscientiously form the belief that another person believes p who has the same qualities that I trust in myself, I will have a prima facie reason for believing p. And when I conscientiously judge that I am more likely to form a true belief and avoid a false belief if I believe what another person believes about p, then their belief has authority for me. In this case, my reasons for believing p are content independent; moreover, the fact that the authority believes p preempts any other reasons I had for or against p.
While much could and has been said about her account of authority, for our purposes, the relevant question is whether or not she has identified an intellectual virtue that is valuable in assessing the reliability of testifiers or, even more broadly, authorities. Should we understand her notion of conscientiousness as an intellectual virtue? According to Zagzebski, epistemic conscientiousness is “the quality of using our faculties to the best of our ability in order to get the truth” (Zagzebski (Reference Zagzebski2012), 48). For Zagzebski, we rightly trust evidence for our beliefs when we trust that we have evaluated that evidence conscientiously. She writes, “The identification of evidence, the identification of the way to handle and evaluate evidence, and the resolution of conflicting evidence all depend upon the more basic property of epistemic conscientiousness” (Zagzebski (Reference Zagzebski2012), 49). This self-reflective conscientiousness certainly sounds like a virtue that forms the basis upon which we trust our evaluation of evidence. But will it help us evaluate the reliability of testifiers?
In addition to conscientiousness, Zagzebski’s list of intellectual virtues includes other virtues that might be relevant to assessing the reliability of the testifier. Assuming, as I have argued, that being able to recognize reliable authority is not properly categorized as a virtue, the other virtues in her list focus on characteristics that would put us as individuals in a good position to evaluate evidence for a claim. In this way, they are akin to her virtue of conscientiousness. In fact, they may be species of conscientiousness. For example, sensitivity to detail is important when I am considering the facts in support of a claim. Moreover, fairness in evaluating the arguments of others focuses on the individual evaluating evidence for a claim as it is presented in the arguments of others. The same can be said for open-mindedness, intellectual humility, intellectual perseverance, diligence, care, and thoroughness; all of these intellectual virtues are excellences that enable individuals to evaluate the evidence for a claim that individuals are investigating and to do it well. Levy notes that the recommendations of virtue epistemologists “appear to aim to bring us each to inculcate the virtues in ourselves and then, guided by our intellectual excellences, to tackle hard problems largely on our own” (Levy (Reference Levy2022), 91). On the surface, this seems correct; the intellectual virtues mentioned above, including conscientiousness, are focused on individuals evaluating evidence. But could they also be excellences that enable individuals to assess the reliability of testifiers?
In Vices of the Mind, Quassim Cassam suggests that the answer to this question is yes. He argues that intellectual vices and vicious thinking are central to the explanation of why people believe in conspiracies. To the extent that conspiracies are believed on the basis of the testimony of others, he seems to be arguing that intellectual vice contributes to bad beliefs on the basis of testimony. On the flip side, this suggests that intellectual virtues could contribute to forming beliefs on the basis of testimony well. Cassam cites empirical studies that suggest that those who believe conspiracy theories exhibit a “conspiracy mentality,” a candidate for an intellectual vice (Cassam (Reference Cassam2019), 69). They tend to believe in multiple conspiracies and even conflicting conspiracy theories. However, Cassam points out that conspiracy thinking is not always vicious – some conspiracies actually occur, and detecting them requires some degree of conspiracy thinking. Here, he cites the uncovering of the Watergate conspiracy. Moreover, anti-conspiracy thinking is not always virtuous – the inability to recognize a conspiracy when it actually occurred due to a bias against conspiracy theories is not truth conducive. Here he cites the case of the Birmingham Six. Though one might initially be inclined to think that identifying when conspiracy thinking is virtuous and when it is vicious depends on what the world is like, Cassam disagrees. He argues that conspiracy thinking is neither virtuous nor vicious; it is vicious when it is accompanied by intellectual vices such as gullibility, closed-mindedness, and illogic. So what makes conspiracy theories about 9/11 vicious is that they are accompanied by vices such as those mentioned above; likewise, the Birmingham Six anti-conspiracy thinking is vicious because it is accompanied by those same vices. Cassam’s argument implies that when these vices accompany any sort of beliefs based on testimony, conspiracy or not, our thinking will fail to be truth conducive. Thus, developing the corresponding virtues would enable our beliefs based on testimony to be truth conducive.
Let’s suppose Cassam is right. Imagine that I have the virtues of attentiveness, open-mindedness, and logic.Footnote 5 I approach the claims presented by the testifier with these virtues. How might these virtues aid me in forming beliefs based on testimony? There seem to be two options: (1) I could individually investigate the beliefs based on testimony employing these excellences to determine if the beliefs are true and thus the testifier is reliable, or (2) I could use the excellences to evaluate the second-order features of the testifier that make the testifier reliable. Let’s consider both options.
3.1. The intellectual virtues and directly assessing claims of the testifier
In option (1), the virtues enable me to perform an independent investigation of the claims of the testifier. Surely, intellectual virtues are valuable in supporting a minimal test for plausibility, such as testing the claims of the testifier for consistency with other well-established beliefs in my noetic structure.Footnote 6 However, employing the intellectual virtues to accomplish a full-scale independent investigation of the belief encounters familiar problems given the truism that we cannot acquire the expertise to perform such an investigation on every belief we must form. Levy focuses on two beliefs that most of us form on the basis of testimony: beliefs about the reality of climate change and beliefs about the necessary precautions in the COVID-19 pandemic. For most of us, it is outside of our capabilities to acquire the expertise required to determine the truth of these claims. When experts in the field of climate science disagree, I, for one, am a lifetime of study from being able to responsibly weigh in on the disagreement. I’m not even sure that I have the intellectual capacity for science that is required to investigate the claims for myself. Even outside of these highly technical scientific claims that would be difficult to investigate, Levy argues that historical claims likewise require expertise to evaluate. The intellectual virtues are of little help without the required expertise. And even if these virtues do enable me to do the investigation well, note that the beliefs I would then form are no longer formed on the basis of testimony. They are now formed on the basis of an examination of the evidence. So setting aside our inability to acquire the relevant expertise, option (1) does not explain how virtues are relevant to beliefs based on testimony. At best, there is a minimal role for intellectual virtues in establishing the initial plausibility of the claims of the testifier.
3.2. The intellectual virtues and assessing second-order features of a testifier
But what about option (2)? Could the virtues enable me to evaluate the second-order features of the testifier that make the testifier reliable and to do this evaluation well? In order to evaluate option (2), I will begin by explaining the second-order features of a reliable testifier. Then I will consider two obstacles to evaluating these second-order features and how virtues might enable us to overcome these obstacles.
Given that we are all limited in our ability to directly assess the expertise of our testifiers, many have suggested that we evaluate reliability by assessing second-order features of the testifiers. Which second-order features are relevant? Lists abound!Footnote 7 Alvin Goldman offers the following set: (1) the ability to argue for one’s views and critique rivals, (2) agreement from other experts, (3) appraisals by meta-experts (including credentials), (4) evidence of objectivity, and (5) evidence of past track records (Goldman (Reference Goldman2001), 93). Assessing these second-order features in the testifier does not require the relevant expertise in the domain to which the alleged expert testifies.
Elizabeth Anderson provides an overlapping set of second-order criteria for lay assessment of scientific claims: (1) expertise (the testifier’s access to evidence and ability to evaluate it), (2) honesty (the testifier’s disposition to report what they believe and not mislead), and (3) epistemic responsibility (the testifier’s responsiveness to others in their expert community) (Anderson (Reference Anderson2011), 145–46). According to Anderson, an ordinary person can evaluate the claims regarding climate change by evaluating purported experts using these second-order criteria (Anderson (Reference Anderson2011), 153).
In Levy’s discussion of his own skepticism toward COVID-19 experts (written in the early days of the pandemic), he cites criteria for determining when expert testimony is reliable. “A scientific consensus is reliable when it has been stress-tested, by all the disciplines relevant to the topic, for an extended period of time. Only under these conditions is the consensus reliable” (Levy (Reference Levy2022), 109). Note that these criteria do not require us to acquire the relevant expertise. They require us to evaluate the level of prolonged agreement among experts who have been challenged. So even Levy seems to suggest a set of second-order criteria by which we could assess the reliability of testifiers.
Many philosophers have sought to specify the relevant second-order criteria. Levy suggests that a sampling of the lists converge on the following features (quite similar to Goldman’s list): credentials, track record, argumentative capacity, agreement with the consensus, and intellectual honesty (Levy (Reference Levy2022), p. 111). These features should suffice to help us consider whether or not intellectual virtues will help us accomplish a second-order assessment of the reliability of a testifier.
3.2.1. Obstacle #1: Epistemically polluted environments
The focus on second-order features of testifiers avoids Levy’s concern that the intellectual virtues are unhelpful without the expertise of the testifier. Prima facie, intellectual virtues such as open-mindedness, humility, and attention to detail seem well-suited to contribute to our ability to assess the relevant second-order features of testifiers. Nonetheless, Levy argues that this prima facie judgment is mistaken; intellectual virtues will not enable us to identify these second-order features of testifiers in the epistemically polluted environment in which we live (Levy (Reference Levy2022), 112). Moreover, we are well aware that we are in this epistemically polluted environment, so we do not trust alleged authorities, even when they are trustworthy. What are the pollutants that are prevalent in our epistemic environment? Levy offers several examples (Levy (Reference Levy2022), 112–17). First, there is extensive mimicry – testifiers who intentionally mimic the features of reliable authorities but lack the genuine traits. Second, there are fake journals that publish the work of the mimics to legitimize it without going through a proper peer review process. Third, legitimate certifying institutions (universities, peer-reviewed journals, medical boards) have financial and ideological interests that can conflict with their interests in certifying experts. Thus, universities have a financial interest in inflating the credentials of their faculty; journals have a bias toward certain kinds of research. Fourth, the research projects that are pursued are influenced by what industries will fund, what grants are available, and what researchers believe will be published. The existence of these epistemic pollutants undermines our ability to identify second-order features of testifiers. Moreover, our awareness of these pollutants makes us deeply skeptical of legitimate, reliable authorities.Footnote 8
Are the intellectual virtues of any use to us in this polluted epistemic environment? I have argued that the intellectual virtues are of limited help when directly assessing the claims made by testifiers since we lack the relevant expertise. However, the intellectual virtues might enable us to identify whether the social epistemic environment in which we are forming beliefs has these pollutants that make it difficult for us to indirectly identify reliable testifiers through characteristic features like credentials, track record, argumentative capacity, agreement with consensus, and intellectual honesty.
Consider Zagzebski’s proposed intellectual virtue of the ability to recognize salient facts. Such a person would be sensitive to detail. While a person with this virtue might not have the expertise to assess the medical evidence for smoking causing cancer, she might recognize the salient fact that those against this claim have a financial interest in promoting cigarette consumption. Or maybe her developed sensitivity to detail will enable her to recognize that the journal that published the study is created and funded by a cigarette company (to use a fictional example). In this way, the virtue enables her to recognize the presence of a pollutant in her epistemic environment.
As a second example, consider what Zagzebski refers to as the detective’s virtues: thinking of coherent explanations of the facts. In 2013, administrators from Harvard College reported that the median grade at Harvard is an A-, and the most common grade is an A.Footnote 9 Moreover, grade inflation is thought to be on the rise at the institution according to a 2022 article in The Harvard Crimson.Footnote 10 Surely, if there is a well-respected credentialing institution, it is Harvard College. So how should we understand this data? The explanation that Harvard would like us to believe is that they admit the best and brightest students and these students reliably perform at the highest levels throughout their time at Harvard. But the person with the detective virtue will consider other coherent explanations of the facts. She might note that Harvard has a financial interest in maintaining this reputation. Thus, her intellectual virtue might alert her to the possibility of an epistemic pollutant in her environment.
I agree with Levy that we form our beliefs based on testimony in polluted epistemic environments. For this reason, I think the cultivation of the intellectual virtues is of the utmost importance; these virtues enable us to recognize the pollutants present in our epistemic environments that impede the use of ordinary criteria for making good judgments about reliable testifiers.
3.2.2. Obstacle #2: Echo chambers and epistemic bubbles
A polluted epistemic environment is not the only structural obstacle to identifying these second-order features of testifiers. It is well-recognized that we form our beliefs in echo chambers and epistemic bubbles, which limits our ability to assess our testifiers for argumentative capacity, agreement with consensus, and intellectual honesty. In his paper, “Echo Chambers and Epistemic Bubbles,” C. Thi Nguyen makes a distinction between two social epistemic structures: epistemic bubbles and echo chambers. According to Nguyen, “An epistemic bubble is a social epistemic structure which has inadequate coverage through a process of exclusion by omission” (Nguyen (Reference Nguyen2020), 143). It is a truncated as opposed to a polluted epistemic environment.Footnote 11 Nguyen suggests that epistemic bubbles arise by accident; it is a side effect of our tendencies to seek out like-minded sources. He argues that while epistemic bubbles are problematic, they are fragile. We can burst an epistemic bubble by exposure to contrary arguments.
According to Nguyen, echo chambers are a more insidious social epistemic structure. He defines an echo chamber as “an epistemic community which creates a significant disparity of trust between members and non-members” (Nguyen (Reference Nguyen2020), 146). Membership in an echo chamber requires commitment to a core set of beliefs, which include beliefs about which sources of information should be trusted. In an echo chamber, the credentials of the members are exaggerated, and nonmembers are discredited. Thus, mere exposure to the contrary arguments of nonmembers of the echo chamber will not enable one to escape the echo chamber as membership in the echo chamber involves believing that these nonmember sources are unreliable. In this way, echo chambers are both truncated and polluted epistemic environments.
In his description of the origins of echo chambers, Nguyen suggests that they often originate not in the accidental way that epistemic bubbles originate but by the intentional actions of bad actors. He writes, “Echo chambers are excellent tools to maintain, reinforce, and expand power through epistemic control. Thus, it is likely (though not necessary) that echo chambers are set up intentionally, or at least maintained for this functionality.” The sort of echo chamber that he seems to have in mind involves the intentional, epistemically illegitimate, and malicious discrediting of outsiders. According to Nguyen, echo chambers usually involve a leader who knowingly discredits nonmembers to maintain epistemic control of members, not because the nonmembers are genuinely epistemically unreliable. Nguyen’s case in point: Rush Limbaugh. He argues (alongside Jamieson and Cappella, Reference Jamieson and Cappella2008) that Rush Limbaugh intentionally created an echo chamber by illegitimately discrediting mainstream media (Nguyen (Reference Nguyen2020), 142).
In epistemic bubbles and echo chambers, we have either accidentally or intentionally limited access to testifiers. When it comes to belief formation on the basis of testimony, especially in the case of belief formation where the beliefs in question are complex, controversial, and not easily settled, environments that limit access to competing testimony are not truth conducive. How might the virtues help us to recognize and escape such environments?
Consider the intellectual virtue of open-mindedness. Jason Baehr offers the following account of open-mindedness: “An open-minded person is characteristically willing and (within limits) able to transcend a default cognitive standpoint in order to take up or take seriously the merits of a distinct cognitive standpoint” (Baehr (Reference Baehr2011), 266). While aspects of this analysis require clarification, it does seem to capture the intuitive sense in which we use the term open-minded. Above, I have characterized the open-minded person as one who reliably achieves success in discovering diverse perspectives, which in turn enables the possessor to reliably achieve success in attaining knowledge. Since the person with the intellectual virtue of open-mindedness will seek diverse perspectives, including perspectives that she does not currently share, she seems particularly well-suited for recognizing and escaping epistemic social structures that limit access to contravening opinions.Footnote 12 In other words, she will both recognize and desire to escape an epistemic bubble and an echo chamber. She will take the limited testifiers in the epistemic bubble to be an undesirable belief-forming environment, and she will understand the discrediting of outside sources by the echo chamber to be problematic. The intellectual virtue of open-mindedness will contribute to her ability to identify and escape this problematic epistemic social structure. Once outside of the echo chamber, she will be better positioned to assess the second-order features of testifiers.
Consider a second intellectual virtue from Zagzebski’s list: intellectual perseverance, diligence, care, and thoroughness. Certainly, this intellectual virtue could be directed at evaluating the evidence for a claim. But it could also be directed at forming beliefs on the basis of testimony. Perseverance, diligence, care, and thoroughness are qualities that manifest themselves in broad consideration of the evidence. The evidence in this case is expert testimony. When forming a belief about complex, controversial, and not easily settled issues, it seems that the person with this virtue will consider a wide range of expert testimony. If she finds herself in an environment that limits access to testifiers, as do epistemic bubbles and echo chambers, she will be unable to exercise this virtue. She will then take these epistemic environments to be problematic. Thus, again, we have a case where the possession of an intellectual virtue contributes to a person’s ability to identify and escape problematic epistemic social structures. Once in a healthier epistemic social structure, she will be able to better identify which testifiers display the second-order features associated with reliability.
One intellectual virtue that seems particularly necessary for identifying and escaping problematic social structures is what Zagzebski referred to as intellectual humility.Footnote 13 In the recent literature on echo chambers, authors rarely (if ever) cite examples of a truncated or polluted environment that the authors inhabit. I am struck by the ease with which we identify these epistemic social structures in our ideological opponents but struggle to critique our own epistemic environments. I am not immune from this criticism! It seems to me that a healthy dose of the virtue of intellectual humility would be of great benefit in enabling us to recognize and escape the epistemic bubbles and echo chambers within which each one of us likely forms some of our beliefs based on testimony. Such an escape is necessary if we are to accurately assess the second-order features of testifiers.
According to Levy, “[I]t’s simply false that epistemic virtues and their responsible application enable the person reliably to track truths. To the extent she succeeds, it is her embedding in appropriate epistemic and social networks that enables her success” (Levy (Reference Levy2022), 125). I agree with Levy that truth conducive belief formation, especially when the source of the belief is testimony, requires being embedded in appropriate epistemic environments. It is precisely for this reason that I argue we do need to cultivate intellectual virtues; it is the intellectual virtues that help us recognize and escape insidious epistemic environments.Footnote 14 Moreover, the intellectual virtues will enable us to identify second-order features of testifiers in healthier epistemic environments.
4. Conclusion
We are rational social animals; we must form beliefs on the basis of the testimony of reliable authorities within our social epistemic structures. Moreover, it is rational to do so. I have argued that recognizing these reliable authorities is not itself an intellectual virtue nor will intellectual virtues enable us to assess the reliability of testifiers by directly assessing their claims without the relevant expertise. Nonetheless, I argue that intellectual virtues do enable us to recognize when we are in a social epistemic environment that is not conducive to assessing the second-order features of testifiers to determine reliability. Therefore, cultivating intellectual virtues is an important step toward outsourcing our beliefs well. Levy proposes that we manipulate epistemic environments through political policy and epistemic nudging. In contrast, my proposal that we cultivate intellectual virtue respects the autonomy of rational agents by empowering us to identify and escape insidious social epistemic environments instead of placing the power to construct epistemic environments in the hands of politicians or those with the ability to nudge.Footnote 15
Open access
