![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
At the intersection of social and virtue epistemology lies the important, yet so far entirely neglected, project of articulating the social dimensions of epistemic virtues. Perhaps the most obvious way in which epistemic virtues might be social is that they may be possessed by social collectives. We often speak of groups as if they could instantiate epistemic virtues. It is tempting to think of these expressions as ascribing virtues not to the groups themselves, but to their members. Adapting Margaret Gilbert’s arguments against individualist accounts of collective beliefs, I show that individualist accounts of group virtues are either too weak or too strong. I then formulate a non‐individualist account modeled after Gilbert’s influential account of collective beliefs. A crucial disanalogy between collective traits and beliefs, I argue, makes the success of this model unlikely. I conclude with some questions with which the future work on collective epistemic virtues should engage.
Acknowledgements
In writing this paper, the author incurred a special debt to Margaret Gilbert for her richly rewarding writings on collective phenomena. The author is deeply grateful to Frederick Schmitt, who introduced him to social epistemology and metaphysics in his graduate seminars and provided many insightful comments on two versions of this paper. The author is greatly indebted to Kay Mathiesen for her patience and for helpful comments that led to several improvements in the paper. Finally, the author would like to thank the Graduate School in the University of Northern Iowa for an eight‐week Summer Fellowship in 2005 that supported work on this paper.
Notes
[1] I borrow the term “collective epistemology” from Gilbert (Citation2004, 96).
[2] For critical commentary on Zagzebski’s virtue epistemology, see the contributions to the symposium on the Virtues of the Mind (Citation1996) in Philosophy and Phenomenological Research volume 60 (2000, 180–209) and Baehr (Citation2006a). For critical evaluations of virtue and the agent‐centered epistemologies of Montmarquet (Citation1993) and Greco (Citation2000), see Lahroodi (Citation2006) and Lahroodi and Schmitt (Citation2003), respectively.
[3] In Lahroodi (Citation2007), I proposed “theory of epistemic virtue” as a label to designate this expansion of epistemology. I now prefer the term “character epistemology,” recommended by Baehr (Citation2006a, 508). Baehr (Citation2006a, section 4.3) usefully lists some of the main philosophical problems that define the field of character epistemology.
[4] To simplify, I drop “vice” from now on, but intend what I say about virtues to apply, where appropriate, to vices as well.
[5] The terminology introduced in the last three paragraphs is adapted from Gilbert (Citation1989, 237, 243; Citation2002, 36). The term “summative” was first introduced by Quinton (Citation1975, 9). Gilbert (Citation1989, 241) uses the term in a broader sense. My formulation of simple summativism for collective traits is adapted from Gilbert’s definition of simple summativism for collective beliefs.
[6] For careful discussions of the divide between virtue reliabilists and responsibilists, see Greco (Citation2002) and Baehr (Citation2006b).
[7] Open‐mindedness and curiosity are cognitive traits of character. Generosity and benevolence are not. Although some of their manifestations are cognitive—say, generous or benevolent thoughts—it is not true that all their manifestations are cognitive. Although sketchy, this definition is good enough for my purpose here.
[8] A simplifying assumption in this paper is the binary notion of trait possession (a person either has or lacks a trait), rather than a graded one (a trait can be possessed to varying degrees). For useful discussions of character traits, see Brandt (Citation1970), Butler (Citation1988), Miller (Citation2003), and Wiggins (Citation1974).
[9] See note 23 below.
[10] See note 25 below.
[11] In attributing a trait to an established social group of a specific kind, we may say that, for instance, this jury or that family is open‐minded or that, equivalently, the members of the jury or the family (collectively or as a group) are open‐minded. On other occasions we ascribe it to a number of people whom we do not necessarily think of as an established social group: Dick and Jane are open‐minded. And sometimes we attribute it to mere populations of individuals with common properties. See Gilbert (Citation2004, 96) for an analogous point about the attribution of cognitive states to collectives.
[12] See Schmitt (Citation1994b, p. 272; Citation2003b, 3).
[13] See Gilbert (Citation1989, 238). The theses about collective traits are adapted from Gilbert’s parallel theses concerning collective beliefs. This section also adapts for collective traits the points made by Gilbert on the motivations for summativism (Citation1989, 238) and simple summativism (Citation1989, 242) for collective beliefs.
[14] A third consideration that can motivate simple summativism is conceptual individualism, a view about the correct analysis of group talk. According to conceptual individualism, talk of groups can be reduced to talk of individuals; when we attribute a property to a group, what we really do is attribute a non‐social property to individuals and composites of these. (See Schmitt [Citation2003, 7–9] for further discussion of conceptual individualism.) Now, conceptual individualists cannot analyze a group predicate like “well‐coordinated” into well‐coordinatedness of group individuals for the simple reason that it cannot be meaningfully attributed to them; other predicates of individuals need to be brought into the analysis. But they may be tempted into thinking that group predicates like “diligent” or “open‐minded” can be analyzed in terms of the same predicates in the simple fashion that simple summativism recommends.
[15] Faced with the question of what collective virtue statements mean or what collective virtues are, it is best to begin with an explicit formulation of our method. The method adopted in this paper is to look at the contexts in which we make such statements and then determine whether the verdicts delivered by the version of summativism under consideration as to whether such statements are true or not match the judgments we would normally make in these contexts. In the absence of relevant empirical data, I make the defeasible assumption that my considered judgments about the truth or falsity of such statements in the given contexts are fairly representative. Some recent wok in epistemology (for example, Bishop and Trout Citation2004; Weinberg, Nichols and Stich Citation2001) has cast doubt on conceptual analysis, as this method is often referred to, for reasons to which I am sympathetic. What I claim for this method is fairly modest, however; at best a description of what, according to our everyday understandings, a collective virtue is.
[16] This is Gilbert’s first counter‐example to simple summativism for collective beliefs. Suppose an anthropologist asserts that the Zuni tribe believes that the north is the region of force and destruction. According to simple summativism for collective beliefs, it is a sufficient condition for the tribe having this belief that all or most members of the tribe have this belief. But imagine a situation, the secrecy situation, in which “each member of the tribe believes that the north is the region of force and destruction, but at the same time is afraid to tell anyone else that he believes this: he is afraid that the others will mock him, having no reason to think that they believe it” (Gilbert Citation1989, 257) In such a situation, simple summativism entails that the anthropologist’s assertion is true, but we tend to judge it as false. For more counter‐examples along these lines, see Gilbert (Citation1989, 258).
[17] For an example of a group not aware of its properties, see Graham (Citation2002, 72–5). Thanks to Kay Mathiesen for drawing my attention to this.
[18] Global or broad traits are those dispositional character traits that exhibit a certain degree of cross‐situational consistency. In contrast, individuals who possess a local or narrow trait will perform the trait‐relevant behavior only in a narrowly defined set of circumstances. Generosity, for instance, is a global trait, as generous individuals are expected to exhibit it in a wide variety of generosity‐eliciting situations. But generosity to one’s friends, with one’s time, and when in a good mood, all count as local traits. Inconsiderateness is a global trait, but inconsiderateness when in a hurry and inconsiderateness towards strangers are local traits. The classification of traits to global and local has prominently featured in recent discussions of situationist ethics. For a very influential example, see Doris (Citation1998).
[19] For a rigorous discussion of the phenomenon of masking of disposition, see Fara (Citation2005, 48–61).
[20] While example 2* would be a coherent example on a positive answer, it is not clear if it would be also effective. It appears difficult to tell a compelling story about how a traffic lights committee can lack global open‐mindedness in a way that renders example 2* interestingly different from example 1*.
[21] It is difficult to say if there are realistic scenarios according to which the group is globally open‐minded when in fact not a single member is. Note that example 2* can be similarly modified to yield example 2**: The editorial board of a newspaper happens to have the exact same members as the volunteer committee to sweep the neighborhood streets in preparation for the annual summer festival. All or most of the members are globally open‐minded as is the board itself, but the sweeping committee is not. The idea is that the editorial board is in charge of making decisions on the publication of enormous number of reports, opinion pieces and letters from readers that bear on a very wide range of topics. This fact qualifies it for global open and closed‐mindedness. The sweeping committee, on the other hand, not being in charge of even minimally complex cognitive tasks, is not globally open‐minded. Of course, it is neither globally closed‐minded—but it is the lack of global open‐mindedness that makes this a counter‐example to simple summativism for global traits.
[22] It is an intriguing question whether there are limits to the divergence we are talking about here. There may be extreme cases where a divergence cannot obtain. If a group is totally pervasive, then there is a question whether it is really possible for all the members to be open‐minded while the group is not. There is some plausibility to the idea that a society is a group and society might be the most important case of a pervasive group for purposes of social theory. It would be interesting to know whether a divergence between the group’s trait and the sum of the members’ traits is possible for the case of society. Can there be, say, an open‐minded society even when most individuals in the society are not open‐minded, or conversely? Maybe this is logically possible, but it is not easy to get a picture of it.
[23] It has seemed to many that a number of well‐known experimental results in social psychology, effectively summarized in Ross and Nisbett (Citation1991), have the implication that people in general exhibit far less cross‐situational consistency than is necessary for trait possession. A growing body of works in philosophy examines the implications of these results and the alleged absence of cross‐situational consistency for virtue ethics and character evaluation. Despite the wide range of positions defended in this literature, one can find, I believe, broad endorsement of the following two claims. First, the majority of people do not possess global traits. Second, people do possess a great variety of local traits. The acceptance of the two claims is most clearly evident in Harman (Citation1999, 316), Doris (Citation2002, 25, 66), Miller (Citation2003, 379, 382), Vranas (Citation2005, 15–16) and Wielenberg (Citation2006, 9). Harman (Citation1999, 316–18) views global traits as the only dispositions legitimately called character traits. Harman (Citation2000, 225) also defends the claim, stronger than the first claim above, that no one possesses a global trait. Kamtekur (Citation2004, 469) and Sreenivasan (Citation2002, 56) endorse the second claim. They hint that the experimental results prove the first claim only if “trait” is understood in the personologists’ sense. They elaborate on a richer conception of character that, they believe, is left unscathed by the experimental results. Sabini and Silver (Citation2005, 542) again seem to endorse the second claim, but conclude (Citation2005, 560–2) that the experimental results prove the first claim for only a small group of traits.
[24] The special subset account, a non‐summative yet correlative account holds that a group G has the trait T if, and only if, a special subgroup within G has trait T. The special subgroup account, although not a summative account, is consistent with summativism. Tuomela (Citation1993, Citation1995) has proposed an analogous special subset account for organizational beliefs. According to Tuomela (Citation1995, 314–16), an organization believes p only if all the operative members (those within the organization with the power to make decisions) accept p as the group belief. Gilbert (Citation1989, 239–40) formulates and rejects a version of special subset account for collective beliefs.
[25] The condition expressed here is merely a necessary condition. Some virtue epistemologists (for example, Zagzebski Citation1996, 137) think that such a cognitive trait can be a virtue if it satisfies further conditions such as “depth,” stability,” and “being acquired.” There are a number of candidates for epistemic ends, prominent among them being acquiring a preponderance of true beliefs (the standard view), understanding (Montmarquet Citation1993, 33; Zagzebski Citation1996, 167), and acquiring new interesting truths at the frontiers of knowledge (Zagzebski Citation1996, 182). Zagzebski (Citation1996, 167) considers “cognitive contact with reality” as the ultimate aim of epistemic virtues, construed so as to include truth, knowledge, and understanding.
[26] A similar classification can be constructed for theories of epistemic vices. My formulation of this classification is indebted to Driver (Citation2001). Montmarquet (Citation1993), Zagzebski (Citation1996), and Driver (Citation2001) have defended versions of internalism, mixed theories and externalism, respectively.
[27] For Gilbert’s views on convention, see Gilbert (Citation1989, chapter VI).
[28] This is so because internalism is motivated by a desire to tie the notion of virtue to notions of epistemic responsibility and praiseworthiness and blameworthiness. Montmarquet’s version of internalism, for example, requires that the possessor of a virtue have some awareness of what the exercise of the trait involves and that its exercise, “given the general conditions that appear to obtain in the world,” is virtuous; that is, is conducive to truth (Montmarquet Citation1993, 30). In contrast, externalists will be content if the trait is in fact truth‐conducive, “even though in particular instances we may have hardly a clue as to whether the particular quality or qualities we are exercising are liable to be so” (Montmarquet Citation1993, 28–9). This would be unacceptable to Montmarquet because, in such circumstances, we can hardly be held responsible for exercising the virtue or deserve praise for doing so. For critical discussion of virtue and agent‐centered epistemologies of internalism in Montmarquet (Citation1993), see Lahroodi (Citation2006).
[29] The counter‐examples I wielded at the summative account of collective virtues do not seem to owe their effectiveness to the virtues being specifically epistemic. Hence, the general points about collective epistemic traits and virtues I made in this paper are likely to generalize to their moral analogues as well.