2.1. Three possible sources of epistemic individualism

There are (at least) three elements of theories of virtue that exhibit a strong element of individualism. They are (1) the individual as the subject of virtue, (2) the nature of the connection between the subject’s virtuous faculty or trait and the virtuous outcome, and (3) the attribution of the value produced by or inherent in the exhibition of the virtue. Call these various kinds of individualism “subject individualism,” “faculty/trait individualism,” and “value individualism.”

Let us first discuss subject individualism, primarily in order to leave it behind. Traditionally, it is individual persons who are taken to be virtuous or not. It is their character traits or faculties that are candidates for being virtues in the first place. In the Aristotelian tradition, virtues are important mainly because manifesting virtue is what constitutes a flourishing (individual) life. And, of course, in epistemology more broadly, it has been individuals (and their parts and properties) who are at the center of normative theorizing. But there is a strand of social epistemology that we might call “collectivist epistemology” that challenges these longstanding assumptions. Works by Margaret Gilbert, Reza Lahroodi, and Jeroen de Ridder, among others,Footnote ² for example, explore the idea that supra-individual entities (e.g., aggregates, collectives, etc.) can be epistemic agents and have properties that are epistemically evaluable in much the same way that we have always taken individuals to be. If this is so, then it might well be possible for such supra-individual entities to have epistemic virtues as well, as Lahroodi (Reference Lahroodi2007) and Palermos (Reference Palermos2022) have explored, for example.

This may be, as Lahroodi claims, “the most obvious way in which epistemic virtues might be social.” (2007, p. 281) It is, however, also one of the most radical and demanding ways to pursue the socialization of virtue epistemology. Moreover, it wears its departure from traditional background accounts of the virtues on its sleeve, as it were. It is unlikely that proponents of collectivist virtue theory will fail to recognize the likelihood of a mismatch between standard background theories of virtue and the needs of a theory that allows collective entities to be virtuous. Since we are focusing on what we take to be the more hidden problems of socializing epistemic virtues, we will say no more about subject individualism here.

An often overlooked fact about theories of virtue is that, in order to be compelling, they need to have accounts of both the nature of virtue and of the value of virtue. Virtues are typically taken to be valuable in all cases, much as knowledge is. Hence, a condition of adequacy on a theory of epistemic virtue is that it can account for that value. When we look carefully at standard accounts of epistemic virtue, it turns out that the individualism that makes it hard for such theories to be a part of social epistemology can stem from either the nature component or the value component. Hence the need to distinguish the two components in order to diagnose the respective problems and the need to identify two additional kinds of virtue individualism.

Faculty/trait individualism is the idea that when one achieves a virtuous outcome of some kind, one does it “by oneself,” so to speak. Traditionally, virtues have been defined as faculties or traits of individuals, and a virtuous act is one that is produced in the right way by an individual. But social epistemologists frequently stress the deep epistemic interdependence amongst epistemic agents, and the upshot is that many (most? all?) of our epistemic achievements are shared achievements. But if our epistemic achievements are often shared, then what of individual epistemic virtue? If we rarely achieve epistemic goods by the operation of our own faculties or traits alone, is epistemic virtue even possible?

Value individualism, on the other hand, is the standard claim that virtuous accomplishments are, in addition to being done “by oneself,” are also in some sense “for oneself.” The value produced by or inherent in virtues and the resultant virtuous acts is traditionally understood to adhere to the virtuous agent herself. But it has become common for social epistemologists, following Kawall (Reference Kawall2002, Reference Kawall2020) for instance, to point out that producing epistemic goods for others is crucial for the healthy operation of our epistemic community and, hence, should count as epistemically virtuous as well.

This is just a very brief sketch of the “mismatches” raised by these two kinds of individualism that are inherent in traditional background theories of virtue that are often assumed by social epistemologists. We will go into much greater detail in later sections devoted to them. The specifics of each mismatch depend upon the particular background theory one assumes, so we must get clear on these first.

Virtue epistemologies are typically divided into two types – virtue reliabilism and virtue responsibilism. There is a lively debate about the extent to which these approaches are genuinely distinct or exclusive, but for our purposes we need to treat them separately. For whether they are exclusive or not, the canonical schemas for each type of virtue are different and face different challenges when one tries to “socialize” the virtues.

2.2. Virtue reliabilism

For a virtue reliabilist, an epistemic virtue is a mechanism that systematically produces some epistemic goods. For example, in John Greco’s early account:

Or as Anne Baril explains Ernest Sosa’s view:

The paradigm cases of such virtues are cognitive and perceptual faculties, like the various modes of perception, memory, etc. We can generalize these definitions to include diverse sorts of epistemic goods,Footnote ³ and so propose a general structure for the reliabilist notion of epistemic virtue:

The definition strongly suggests an instrumentalist reading of the value of an epistemic virtue. That is, an epistemic virtue is valuable insofar as it produces valuable epistemic goods. Its value is derivative of the epistemic goods it produces.Footnote ⁴ And indeed, this is the usual way the value of reliabilist epistemic virtues is understood.Footnote ⁵

Together EpVRel and ValueRel constitute the general account of virtue reliabilism. And it is in this sense that we shall further refer to this position.Footnote ⁷

Now that we have this general characterization, we can begin to assess the source of the two kinds of individualism we are interested in. First, faculty/trait individualism is plausibly, though not quite explicitly, inherent in the definition of virtue in terms of an “ability.” Abilities are had by individuals and a virtue is an ability that is sufficient to reliably produce epistemic goods “on its own.” This, at least, is the most natural reading of such a definition. The main point for our purposes is that trying to account for epistemic virtues that involve, for example, interdependence among various individuals in a virtue reliabilist framework cannot simply take the framework as it is traditionally developed and ordinarily understood. Serious work addressing the nature of such a virtue, and defending and motivating the existence of such virtues is required.

Another reason to take this interpretation as the default is that most virtue reliabilists are developing accounts of epistemic virtue primarily to provide a theory of knowledge. Indeed, most such theories are best seen as developments and refinements “from earlier externalist epistemologies such as simple process reliabilism.” (Turri, Alfano, & Greco, Reference Turri, Alfano, Greco and Zalta2021) Few process reliabilists consider the possibility that knowledge for S might be the result of a combination of processes in multiple different individuals.Footnote ⁸

Interestingly, though, there seems to be nothing in the general account of reliabilist virtue itself that precludes counting as virtues those abilities that produce knowledge, say, in others. But, as Anne Baril explains, it is almost never even considered as a possibility:

One notable exception to this tendency is Jason Kawall’s articles on other-regarding virtues, which we will discuss in a later section. For now, the point to take away is that, while neither the definition of epistemic virtue nor the account of the goods that make them valuable precludes counting as a virtue an ability that produces epistemic goods in a third party, the context of their theoretical development is such that it is almost never theorized about.

Similarly, value individualism seems to be both inherent in the most natural reading of virtue reliabilism and not explicitly encoded. Again, the most natural interpretation of standard virtue reliabilism is that the value of the virtuous faculty adheres to the agent who bears the faculty. Nobody we are aware of (other than possibly those working on collectivist versions of virtue theory) defends a view according to which a virtuous ability is shared somehow by more than one individual. Again, the point is not that no sense can be made of reliabilist virtues in a context of frequent interdependence; the point is that doing so requires making some serious revisions to one’s background theory of virtue.

For many, part of the appeal of virtue reliabilism is that the justification-producing mechanisms are good old-fashioned cognitive faculties like our senses, powers of inference, etc. There is a continuity from the empiricist strand of our philosophical history that this taps into. It may be that social virtue reliabilists will need to explain why, say, being a reliable testifier is the same kind of valuable epistemic faculty as, say, good vision. So while neither EpVRel nor ValueRel explicitly commit them to value or faculty/trait individualism, given the way traditional virtue reliabilist theories are usually introduced, developed, and defended, anyone wishing to develop a genuinely social version of the theory has some serious work to do.

2.3. Virtue responsibilism

Classical virtue responsibilism proposes a different account of both the nature and the value of virtue. Montmarquet’s early version of virtue responsibilism (Montmarquet Reference Montmarquet1993, p. 30) took a radical step of defining virtue in a way that did not imply reliable success. The role of virtue consists in giving the right motivation for a truth-desiring person. Epistemic virtues are qualities that a truth-desiring person should want to have, and they are virtues regardless of whether they subsequently lead to the acquisition of epistemic goods (which for Montmarquet are true beliefs). However, it is Linda Zagzebski who developed the now-canonical (though not the only) version of virtue responsibilism. It is a version of her position that we will adopt here as paradigmatic of virtue responsibilism.Footnote ⁹

Notice that there are several distinct elements to Zagzebski’s definition of an epistemic virtue. It is a “deep and enduring acquired excellence.” This means that virtues are character traits which we acquire through habituation. Furthermore, the trait must involve “a characteristic motivation to produce” epistemic goods. And, finally, the trait must prove to be a reliable way to achieve the goods toward which it aims.

This broader account of the nature of virtue is further complemented with a distinct account of the value of virtue. As in the case of virtue reliabilism, the Zagzebskian virtue responsibilist requires epistemic virtues to be reliably productive of epistemic goods. Hence, responsibilist virtues have the same kind of instrumental value as do reliabilist virtues. However, the addition of a motivation component provides an additional vector of epistemic value. It could thus appear that the value of epistemic virtues would be bidimensional as well. In what concerns reliable success, it would be reduced to ValueRel, and in what concerns a proper motivation, it would relate to some kind of intrinsic personal value. However, Zagzebski argues that ValueRel cannot stand on its own for the same reasons that Kvanvig (Reference Kvanvig2003) framed in his discussion of the Meno Problem (Zagzebski, Reference Zagzebski2003). Thus, she proposes that the value of virtues be ultimately derived from the value of the good motivation.Footnote ¹⁰

There remains, of course, the question of whence the value of proper motivation. Zagzebski considers two different ways to understand the value of proper motivation, one in terms of its contribution to a flourishing life, and the other in terms of the intrinsic value of the motivations themselves (Zagzebski, Reference Zagzebski1996, pp. 197–211). She ultimately chooses the latter as her account, but it is worth pointing out that this is both idiosyncratic to Zagzebski and also somewhat at the service of her theistic metaphysical commitments. It is much more common to take the more traditional Aristotelian view that virtues are valuable because of their contribution to flourishing.Footnote ¹¹ As we are here using Zagzebski’s view simply as a canonical example, we will take the other course and assume the flourishing account of the value of epistemic virtues. Thus, for virtue responsibilism, the value of an epistemic virtue may be explained in the following manner:

As before, let us briefly take stock of how value individualism and faculty/trait individualism are inherent to the virtue responsibilist framework. The trait individualism is both explicit and theoretically deep. Virtues simply are “deep and enduring” traits of the individual. Moreover, what makes virtues valuable is their contribution to the virtuous person’s life of eudaimonia. My virtues are good because they contribute to my becoming a eudaemon. It is not straightforward how my production of epistemic goods for others makes me a more excellent human – at least on traditional Aristotelian or neo-Aristotelian grounds. Contributing positively to someone else’s life might be good in various ways and for various reasons, but it does not count as virtuous (unless the contribution to others’ good somehow also contributes to your own, which is often how virtue ethicists defend the morality of other-regarding acts).

So the value individualism is also both quite explicit and theoretically deep. When we say that these commitments to the two kinds of individualism are “theoretically deep,” we mean that they are not coincidental or accidental. As this section demonstrates, those commitments are vital to the central metaphysical and normative features of the view.

2.4. Two challenges

These general considerations go some way toward making our case that virtue epistemologists cannot take their preferred background theory of virtue for granted when developing accounts of more social epistemic virtues. If virtue epistemology is to thrive in the contemporary debate, it must adapt its accounts of virtue and of virtue’s value to be consistent with the social nature of knowledge (and other epistemic goods). In the next two sections, we will look more deeply at the challenge that value individualism and faculty/trait reliabilism pose to virtue epistemology. We will do this by looking more closely at two ways one might wish to socialize epistemic virtues and showing what kinds of deep modifications to theories of virtue would be required to do so.

One way of socializing epistemic virtues that has been proposed is that there are epistemic virtues, the products of which are primarily or exclusively good for someone other than the virtue bearer. Clearly, this runs afoul of value individualism. If an epistemic virtue theory is to accommodate such virtues, it will have to broaden its theory of value accordingly. We will discuss this issue further in Section 3.

Another example is the division of epistemic labor that seems common in the achievement of a wide variety of epistemic goods. Our achievement of knowledge, say, often will depend on contributions from other epistemic agents, especially in cases of testimony. Here the issue is not so much that one person’s virtue is directed toward another person’s good. Rather, the feature of interest is that our own virtue is often not sufficient to bring about our desired epistemic ends and so it is less than obvious that it can give us a sort of credit that would transform a mere true belief (possibly a lucky one) into knowledge.

This degree of epistemic interdependence seems to violate at least the spirit of faculty/trait individualism. But given the social nature of human inquirers and knowers, this is more of a feature than a bug. Hence, a fully social virtue epistemology should be able to account for our having genuine epistemic virtues that, perhaps even more often than not, do not suffice to produce the epistemic goods toward which they are aimed.

We are not aware of anyone who addresses this issue head-on, but there are some obviously analogous issues with the problem of shared credit for knowledge when it is the result of multiple epistemic agents working together. So, in Section 4 we will look at an attempt to make sense of these instances of shared epistemic credit for guidance in how to make sense of epistemic goods brought about by multiple agents’ virtues.

It is worth pointing out at this stage that we are not defending any particular version of virtue epistemology. We are arguing that socializing virtues requires a hard look and quite likely revision of the background framework of virtue that is usually employed in epistemology. And we are offering some suggestions about ways forward for both virtue reliabilism and virtue responsibilism. But fully developing any of those strategies would be the work of another paper.

3.1. The relational self and epistemic interdependence

One way to get what’s good for other people into the value of a virtue would be to get those people appropriately connected to the virtuous person herself. But not just any relationship will do. It will need to be such that a virtue’s being good for the other person still somehow contributes to the virtuous person’s eudaimonic life. Some recent work by Heidi Grasswick (building on the work of Lorraine Code and Annette Baier, among others) might help here. Grasswick develops an account of constitutive relational autonomy that she thinks can play a central role in responsibilist virtue epistemology.

Autonomy is relevant here because of its centrality in a responsibilist virtue epistemology. Responsibilism requires robust epistemic agency on the part of the virtuous person. And according to Grasswick, epistemic autonomy “… is central to the idea of epistemic agency itself” (Grasswick, Reference Grasswick and Battaly2018, p. 196). Autonomy is a relatively familiar notion in the history of Western philosophy, but it has often had connotations of “rugged self-reliance.” Much contemporary work on epistemic autonomy repudiates that conception of autonomy and acknowledges that epistemic autonomy is consistent with epistemic dependence on others. (Matheson & Lougheed, Reference Matheson and Lougheed2022) Still, Grasswick’s conception of relational autonomy goes far beyond this acknowledgment.

In short, Grasswick argues that not only are individuals reliant on the care and nurturance of others to develop from helpless infants into autonomous epistemic agents, but the very capacity of epistemic agency is inherently social as well. In her terms, our epistemic autonomy is not only developmentally relational, but constitutively relational as well. The exercise of our epistemic agency, which we must do to develop and exhibit epistemic virtue, at least sometimes requires interacting with others: “On the constitutive model, our epistemic dependence on each other runs deep: much of what allows us to be successful as knowers, and many of the very activities of inquiry, involve being responsive to other knowers. Reasoned inquiry is dialogical” (Grasswick, Reference Grasswick and Battaly2018, p. 201). There are many complex and fascinating details here, but the upshot for our purposes is that the epistemically autonomous agent, on this view, cannot engage in (much) inquiry without being in epistemic relation with others.

This might provide the connection necessary to solve the problem at hand for responsibilist virtue epistemology. There are at least two different ways this might go. One way would be taking Grasswick’s view to imply that the self is, at least to some extent and in some relevant ways, plural.Footnote ¹⁴ Hence, what is epistemically good for (at least some) others would also be epistemically good for the agent herself because the exercise of her epistemic agency depends constitutively on (at least some) others. This would require, at a minimum, defending such an account of the self. But additionally, it would require showing that such an account is social enough to make sense of the full range of generous virtues. That will be easier or harder depending on how widely shared our autonomy turns out to be. This may also turn out to be a pivot to collectivist epistemology.

Another way this might go is to keep a fairly traditional concept of the individual self, but accommodate generous virtues along Grasswickian lines by arguing that the shared autonomy results in shared responsibility for producing the epistemic goods, and so to shared credit for the achievement. This is actually a strategy that we consider in a slightly different guide in Section 4, so we will say no more about it here.

Obviously, there are many complexities and details of both of these strategies that we have not addressed here, and we are not defending Grasswick’s view as a way to resolve value individualism. We mean only to be highlighting the significance of the modifications needed to reconcile the existence of social epistemic virtues with their theory.

3.2. Epistemic generosity and personal flourishing

There is at least one other possibility that gets to a similar resolution without modifying our conceptions of central features like autonomy or agency. Instead, we will consider a modification to the idea of what the epistemic part of a life of eudaimonia looks like. Any account of what a flourishing human life looks like will depend, as it did for Aristotle, on a prior conception of something like “human nature.” What it takes for a human to flourish depends crucially on what being human is necessarily like. We have already noted that humans are the kind of creature for whom epistemic success is largely a communal endeavor. Even without assuming a relational view of our epistemic autonomy, it is clear that most of what we know we learned from others, including our language and reasoning skills. Having epistemic success is, presumably, key to one’s flourishing in general. Hence our individual flourishing depends on getting a lot of what we need from others. So, once again, when we help others epistemically, we are improving our own epistemic lot as well. Our individual flourishing is enhanced when other people are better off epistemically.

But perhaps a slightly more robust approach would be to understand human nature as more inherently social, such that being dependent on a broader epistemic community (and contributing to its quality) is not so much a contingent strategy for achieving epistemic goods but rather a deep feature of being human at all. On such a construal of human nature, producing epistemic goods for others would straightforwardly contribute to a human’s flourishing because it would be an important part of being an excellent human.

Both of these approaches require a revision to the generally assumed account of human flourishing and its associated account of human nature rather than a revision of the theory of virtue itself (though that might also be necessary). It is worth pointing out that any of the strategies discussed in this section might also be useful for other popular virtues whose value can be hard to account for, but aren’t obviously or directly other-regarding either. The proposed direction, taking into account the influence of epistemic virtues in shaping the social environment that supports agents in acquiring knowledge, may help to explain such problematic social virtues as open-mindedness, trust, and empathy. For example, such authors as Battaly (Reference Battaly2018), Carter & Gordon (Reference Carter and Gordon2014), and Levy (Reference Levy2021) have argued that open-mindedness is not universally valuable the way virtues are supposed to be.Footnote ¹⁵ It might be the case that in non-ideal conditions, one can get worse off being open-minded, and a healthy dose of close-mindedness might just be what truth-seeking agent should adopt. Likewise, trust and empathy might easily be considered moral virtues, but their epistemic import might not be as straightforward as one wished. Both the above strategies for accounting for generous virtues might help explain the value of those epistemic virtues in terms of their promoting human flourishing in an extended, more social, sense.Footnote ¹⁶ No doubt open-mindedness, trust, and empathy strengthen the epistemic community one belongs to, and in that way can be valuable in all the ways discussed above.

Obviously, from these brief sketches, it is far from clear that any of these strategies is ultimately defensible or would solve the problem at hand. Our point in this section is, in part, to suggest some ways forward for the development of a responsibilist social virtue epistemology. But it has also been to highlight the kinds of serious philosophical work that such a project entails, whatever strategy one ends up pursuing. Work that is rarely, if ever, done.

In the next section, we turn to the second dimension of epistemic individualism and the related, problematic social epistemic phenomenon – the shared production of epistemic goods. We will do so by way of Adam Green’s account of shared epistemic credit. As with the views just discussed, the idea is not that Green’s solution is the only one of interest or definitive. But because of his explicit and consistent building on Greco’s (implicitly: individualistic) traditional virtue epistemology, it provides a good case study for our inquiry into how virtue epistemology can accommodate social epistemic phenomena.