1. Introduction

This paper is an attempt to clarify the ways in which social theories of science should conceive of individual epistemic agents. To do this, I examine here one of the most important recent developments of a social theory of science, Helen Longino's “contextual empiricism” (Reference Longino1990, 215), or “critical contextual empiricism” (Reference Longino2002, 208). This theory consists of two primary parts. The first is a rejection of the ideal of value neutrality; more specifically, Longino argues that “contextual values,” or values traditionally thought to be external to scientific reasoning, inevitably influence the epistemic evaluation of research. Second, she argues that, despite the value-laden character of research, the goals of objectivity and epistemic acceptability are still attainable—but only if the research in question is produced within certain kinds of communities, namely, communities that approximate an ideal, free marketplace of ideas, such as that described in the writings of John Stuart Mill.Footnote ¹ Thus, on Longino's account, the social and the epistemic are reconciled by regarding certain social interactions as necessary features of the process of epistemic justification.

Longino's theory of science is one of the most sophisticated recent attempts to explicate the role of the social in scientific knowledge, and I find significant parts of her theory—especially her criticisms of the ideal of value neutrality—to be very persuasive. At the same time, the role that the social plays within her theory is ambiguous. While Longino argues that scientific knowledge is necessarily social, other strands within her writings point in a rather different, and more individualistic, direction. In particular, I will argue that Longino's notion of an ideal scientific community, a notion inspired by Mill's political liberalism, presupposes a conception of the individual that ultimately undermines her own claim that scientific knowledge is necessarily social. Following this, I will draw upon certain aspects of the work of Thomas Kuhn and Imre Lakatos in order to highlight an alternative conception of the individual, one not only more plausible than that presupposed by the ideals of Longino and Mill but also better suited to a truly social theory of science.

2. Longino on the Social Character of Scientific Knowledge

Longino's discussion of the social character of scientific knowledge begins with her rejection of the ideal of value neutrality—or, more specifically—with her advocacy of an inevitable role that contextual values play in the evaluation of scientific research. To examine the ideal of value neutrality is not one of the aims of this paper;Footnote ² nevertheless, it will be necessary to discuss briefly Longino's rejection of this ideal, as it plays a crucial role in her argument for the social character of scientific knowledge.

Longino begins her rejection of the ideal of value neutrality by stating the relatively uncontroversial point that there is a logical gap between data and hypotheses and that this gap is inevitably filled by background assumptions, which usually take the form of a particular theoretical framework (Reference Longino2002, 125). More controversially, she argues that these background assumptions are dependent on the context in which they arise. While a theoretical framework or set of evaluative criteria is required for a given set of data to count as evidence for a particular hypothesis, no set of data dictates which theoretical framework or set of evaluative criteria should be employed; in principle, a variety of different frameworks could be adopted. It is at this stage, Longino argues, that contextual values play an inevitable role. The kinds of contextual values that play a role will depend upon the science in question, as well as upon the historical situation in which the research takes place. In some situations, background metaphysical or theological beliefs constrain the kinds of theoretical frameworks open to scientists; in other situations—especially those concerning areas of science that touch on controversial social concerns—values that are morally and politically laden play an inevitable role in the choice of one theoretical framework over another. The crucial point, however, is that given that contextual values do play this inevitable role in scientific reasoning, the ideal of value neutrality should be rejected.

If we grant Longino's argument that evidential reasoning in science inevitably involves background assumptions that are context-dependent—that is, that depend upon the aims, interests, and historical locations of the communities doing the research—then a crucial epistemological question arises: what becomes of our belief that scientific knowledge and objectivity are possible? Can we reject the ideal of value neutrality and still retain these traditional epistemic ideals? Longino argues that we can—but only on the condition that we view scientific knowledge as social. That is, it is only when scientific research is produced within a community that the subjective preferences of individuals can be transformed into objective beliefs, frameworks, methods, or criteria. Scientific knowledge is thus necessarily social.

Yet it is clear that not just any social processes are productive of knowledge. There are many ways of organizing scientific research that are detrimental to scientific progress; the organization of genetics research in the Soviet Union is just one well-known example. Only certain kinds of communities—communities that facilitate certain kinds of critical interaction—are productive of scientific knowledge, and it is an essential part of Longino's theory of science to spell out precisely which kinds of communities are thus productive. Her method of doing this is through an examination of our intuitive understanding of fundamental epistemological concepts, and the result of this examination is an articulation of an ideal scientific community.Footnote ³ An ideal community, she argues, is one that approximates a free marketplace of ideas, as described in the writings of Mill. More specifically, an ideal community possesses four characteristics: public venues for the criticism of research, uptake of criticism, publicly recognized standards for the evaluation of research, and a tempered equality of intellectual authority (Longino Reference Longino2002, 129–134).

The first characteristic, public venues for the criticism of research, creates the formal conditions under which research—including evidence, methods, assumptions, and reasoning—can be criticized from a diversity of perspectives. Following Mill, criticism from a diversity of perspectives “not only spurs evaluation and reevaluation of hypotheses, but also leads to better appreciation of their grounds and of their consequences” (Longino Reference Longino2002, 129). The second requirement, uptake of criticism, is intended to ensure that criticism is not ignored but, rather, considered carefully and acted upon. The third requirement, publicly recognized standards for the evaluation of research, is meant to ensure that criticism is relevant to a position—that is, that it appeals to some standards held by the person(s) maintaining the position—as well as to ensure that what counts as relevant is not merely a function of the arbitrary whims of individuals but, rather, of a set of shared standards to which the community is bound. Finally, the fourth requirement, tempered equality of intellectual authority, ensures that individuals coming from different perspectives can interact with one another on a relatively equal playing field.

There are a couple of points regarding Longino's discussion of the social character of scientific knowledge that need to be emphasized. The first is one that has already been made but that merits repeating: it is the social character of scientific inquiry, and only the social character of scientific inquiry, that allows scientific knowledge to be produced. She writes, for example, of “the necessity of critical interaction for the integrity of knowledge” (Reference Longino2002, 4; my emphasis). Furthermore, in summarizing her account of scientific knowledge, she writes: “I have argued that cognitive practices are social and that the located character of empirical subjects and the underdetermination problem together mean that scientific justification must be social” (Reference Longino2002, 134; my emphasis). Thus, a necessary condition for the development of scientific knowledge is that research be scrutinized within a community—in particular, a community organized so as to approximate her ideal.

Second, it is important to emphasize the fundamental role that Mill's political philosophy plays within Longino's account of scientific knowledge. It is clear that her articulation of an ideal scientific community is inspired in important respects by Mill's political liberalism, but the connection between the two is not merely one of vague influence. Rather, Longino's epistemology is logically embedded within the framework of Mill's political philosophy. Once again, Longino argues that scientific research is epistemically acceptable only if it is scrutinized in communities that approximate a Millian marketplace of ideas. In other words, the existence of scientific communities that are organized along the lines of Mill's political liberalism is, for Longino, a necessary condition of the development of scientific knowledge. Because of this, Mill's political philosophy is nothing less than an essential part of her epistemology of science.

The fact that Longino's epistemology of science is embedded within the framework of Mill's political liberalism has important implications for her theory of science. More specifically, I will argue in the next section that the conception of individual epistemic agents that is implicit within both Mill's political liberalism and Longino's characterization of an ideal community ultimately undermines Longino's own claim that scientific knowledge is necessarily social.

3. On the Ambiguously Social Character of Longino's Theory of Science

The primary argument of this section is that Longino's adoption of Mill's political philosophy creates an unworkable tension in her account of the social character of scientific knowledge. Longino argues, on the one hand, that scientific knowledge is necessarily social, but her employment of Mill's political philosophy, on the other hand, suggests something very different. It suggests that while certain social arrangements are helpful ways of producing knowledge, they are not, strictly speaking, necessary.

To see this, let us consider Longino's claim that an ideal community satisfies the norm of the uptake of criticism. It is clear that the development of scientific knowledge requires, at least in some general sense, responsiveness to criticism. If scientific communities, or the individuals making up those communities, were to ignore all criticism, the result would be intellectual stagnation. But how is it, precisely, that we are to understand the norm of uptake? In order to answer this question, it is helpful to begin by noting that there are at least two levels at which this norm could operate: the individual level and the social level. In the first case—let's call this the individualistic interpretation of uptake—we would say that a community is characterized by the norm of uptake if (and only if) that community consists primarily of individuals who exhibit this norm. In the second case—let's call this the social interpretation of uptake—we would say that a community is characterized by the norm of uptake if that community manages to be responsive to criticism, or to avoid intellectual stagnation, without it necessarily being the case that all, or even the majority, of the individuals within that community are thus responsive. When Longino demands that communities exhibit the norm of uptake, which of these two interpretations does she have in mind?

Unfortunately, she does not say, and the neglect to do so creates some confusion in interpreting her ideal. However, there are strong reasons for believing that she favors the individualistic interpretation. For example, she tends to explain what the norm of uptake requires in terms of the actions and attitudes of individuals—for instance, when she writes, “What is required is that community members pay attention to the critical discussion taking place and that the assumptions that govern their group activities remain logically sensitive to it” (Longino Reference Longino1990, 78). Furthermore, Longino is very critical of views that make individual knowledge derivative upon social knowledge—for example, views in which individuals “know in virtue of their membership in their communities” (Reference Longino2002, 146)—which in turn leads her to emphasize the importance of individual intellectual autonomy: “Individuals have doxastic autonomy in the sense that what they believe is not constrained by the beliefs of their community, although their belief in any given instance may be the result of particular interactions” (Reference Longino2002, 154). This emphasis upon individual intellectual autonomy coheres very well with the demand that individuals listen and respond carefully to the criticisms that they receive.

If we assume, then, that the norm of uptake is meant to apply to individuals, we must now ask what individuals must do in order to satisfy the norm. What is required in order to “pay sufficient attention” to critical discussion? Given that Longino embeds her epistemology of science within the framework of Mill's political liberalism, we might look to Mill for help in answering this question. Consider, for example, Mill's discussion in On Liberty of the justification of beliefs:

The interpretation of uptake that emerges from this passage, I would suggest, is the following: communities satisfy the requirement of uptake if and only if they consist primarily of individuals who are genuinely responsive to criticism, and individuals are genuinely responsive to criticism if and only if they are willing to question any of their beliefs by considering any criticism from any source. In other words, communities satisfy the norm of uptake if and only if they consist primarily of individuals who are open to everything.

One might be skeptical of the claim that Longino's ideal scientific community places such stringent requirements upon individual epistemic agents. Further confirmation that it does comes in the form of Longino's response to what one might call the “creationism objection.” The creationism objection is an objection to Longino's ideal community according to which the community of creationist scientists meets the four criteria of her ideal. If this were the case, it would obviously represent a reductio for Longino's view. Given this, it is important for her to be able to show that at least one of these criteria is not satisfied by the creationists. To this end, Longino argues that creationist scientists do not satisfy her norm of the uptake of criticism. “A practice of genuinely open criticism and discussion,” she writes, “requires an openness to all perspectives: no claim or belief can be held immune to criticism” (Longino Reference Longino2002, 159; my emphasis). Creationist scientists, she argues, are not open to all perspectives in the way that is demanded by the norm of uptake. The important point for our purposes, however, is that Longino's interpretation of uptake requires that individual epistemic agents be open to everything.

This conception of the individual has been the subject of much criticism within moral philosophy and political philosophy, from pragmatists such as John Dewey to communitarians such as Michael Sandel (see, e.g., Dewey Reference Dewey[1935] 1987; Sandel Reference Sandel1984). The criticisms that come from these two quarters are the same: the individual who is open to everything—or, in Sandel's terminology, “the unencumbered self”—is a view of the self that falsely assumes that individuals have complete freedom to choose their own aims, values, and conceptions of the good life. Unencumbered selves are not intrinsically socially located; they have no commitments to aims, values, or conceptions of the good life that are given to them by virtue of the fact that they are socialized within certain contexts. Rather, they stand apart from all such “attachments” and are open to, and capable of, choosing from among them and revising them at will. On this conception of the individual, social arrangements and institutions are, to quote Dewey, “things that operate from without, not entering in any significant way into the internal make-up and growth of individuals” (Dewey Reference Dewey[1935] 1987, 30). Both Dewey and Sandel, of course, reject this notion of the individual, and I am highly sympathetic to their arguments. I will return to this issue in the next section. For the present, however, the important point is to note the striking similarities between the unencumbered self—the self that, in some sense, stands apart from the social context in which it exists—and the conception of the individual that lies at the heart of both Mill's and Longino's ideals.

Given this discussion, we are now in a position to see the tension that exists within Longino's theory of science. As argued, Longino accepts the individualistic interpretation of uptake, according to which a community satisfies this norm if and only if all (or most of) the individual members of that community are open to everything, in the sense discussed above. Yet if we grant that an ideal community is one that is composed of individual epistemic agents who are open to everything, then why is it that the development of scientific knowledge requires, in principle, a community? In this regard, it is important to recall Longino's primary reason for asserting the necessarily social character of scientific knowledge in the first place—namely, that critical discursive interaction is the only way, in principle, to transform the subjective preferences of individuals into objectively maintained beliefs. But if individuals, in the ideal, are completely open-minded—if they are capable of questioning all their beliefs, of examining evenhandedly all potential weaknesses in their views, and of adjudicating between opposing beliefs, methodologies, evaluative criteria, and so on, in a fair and evenhanded manner—then it is false that a community is necessary for the justification of beliefs. For such individuals are perfectly capable, at least in principle, of evaluating lines of argument in an objective fashion—and of doing so on their own, qua individuals.

Of course, I am not suggesting that, within Mill's and Longino's ideals, social processes are not important, epistemically speaking. As Mill emphasizes, the “wise man” hears what can be said about his views from “persons of every variety of opinion” and “every character of mind.” Given the fallibility of individual epistemic agents, and given that individuals are highly limited in their physical and cognitive capabilities, it is highly beneficial, epistemically speaking, for individuals to live within diverse communities. But “highly beneficial” is a far cry from “necessary.” Individuals within Mill's free marketplace of ideas, being the completely open-minded individuals that they are, are capable of adjudicating between their own positions and alternative views in an evenhanded and hence objective fashion and, therefore, capable of transforming subjective preferences into objectively maintained beliefs—and they are capable of doing so on their own. Thus, it follows from Longino's interpretation of uptake that, while certain social arrangements are highly beneficial for the development of scientific knowledge, they are not, in principle, necessary.

I have argued thus far that Longino's employment of Mill's political philosophy creates an unworkable tension in her account of the social character of scientific knowledge. On the one hand, Longino argues persuasively that the ideal of value neutrality should be rejected; as a consequence of this, she argues, scientific knowledge is necessarily social. At the same time, she embeds her epistemology of science—and her conception of an ideal community—within the framework of Mill's political liberalism, which ultimately undermines her own claim that scientific knowledge is necessarily social. At this point, then, we have a choice. Should we accept the general outlines of Longino's ideal scientific community, including the conception of individuals as, in principle, open to everything and thereby give up the claim that scientific knowledge is necessarily social? Or should we affirm the necessarily social character of scientific knowledge and accordingly reject her conception of an ideal community as approximating a free marketplace of ideas? While I believe that we should take the latter course, a thorough justification of this belief is beyond the scope of this paper. In the next section, however, I will spell out an alternative conception of individual epistemic agents—individuals as advocates, not unencumbered selves—and I will adduce some prima facie reasons for believing that this conception is more realistic than, and epistemically superior to, the conception of the individual presupposed by the ideals of Mill and Longino.

4. Advocates, Not Unencumbered Selves

As a means to elucidating an alternative conception of the individual, it will help to return to Sandel's rejection of the unencumbered self. In a passage in which he argues that unencumbered selves lack character and moral depth, Sandel writes:

On this view, again, individuals are intrinsically socially located, in that they have certain aims, values, and conceptions of the good life that are given to them by virtue of the fact that they are socialized within particular contexts—for example, within a particular historical period, country, or family. This is not to say that individuals are incapable of revising any of their aims or values or that they cannot critically reflect upon the contexts in which these aims and values are formed. It is, however, to say that there are some aims, values, and conceptions of the good life that are very deeply held, even to the point that they are viewed as essential by the individuals holding them.

I would argue that the conception of individual moral agents defended by Sandel is a good model for a conception of individual epistemic agents in science. To see this, consider the theories of science developed by Kuhn (Reference Kuhn1962) and Lakatos (Reference Lakatos, Lakatos and Musgrave1970)—theories that are obviously very different from one another in many ways but that also agree, at least in outline, upon a conception of individual epistemic agents. According to both Kuhn and Lakatos, the aims, criteria for theory evaluation, experimental techniques, and so on are, at least in part, given to individual scientists by virtue of the fact that they are socialized within particular communities, disciplines, or subdisciplines. On these views, individual scientists are committed to their respective paradigms or research programs; this means, among other things, that they are extremely reluctant to question the foundational elements of their approaches—whether this be the model problems and solutions that partially characterize a paradigm or the hard core of a research program. The result is that individual scientists are viewed more along the lines of advocates for particular approaches, paradigms, or research programs than as individuals who are open to everything.

This particular conception of individual epistemic agents is clearly more realistic than the one presupposed by Mill and Longino; if the history of science teaches us anything, it is that individual scientists are not open to everything. In addition to its descriptive power, however, there are also strong epistemic reasons for adopting this conception of the individual. To see this, consider a community that consists of two Lakatosian research programs and suppose that the scientists working within each of these programs criticize the tenets of the other program, including the hard core of that program. Suppose furthermore that the scientists who receive these criticisms are simply uninterested in following them up; they see that as a waste of time and resources and a diversion from the more important task of working out their program. Is there anything necessarily problematic about this, epistemically speaking? According to Longino's individualistic interpretation of uptake, there is: individuals are not being open to all perspectives. But surely this can’t be right. For suppose that the criticisms put forward by the scientists working within one of these research programs turn out, at the end of the day, to be either misguided or insignificant and that the criticisms put forward by the scientists working within the other program are highly trenchant. Suppose furthermore that as a result of the problems with the first research program, it degenerates and dies out, while the second makes dramatic progress. In this situation, there is nothing epistemically problematic about the individual scientists not taking up the criticisms of the members of the other research program; scientific progress was made, even though the individual scientists of each research program failed to take up the criticisms of the other.

Moreover, not only is it not necessarily problematic for individual scientists to be extremely reluctant to question the foundational elements of their research program; it is, in many cases, epistemically beneficial for them to be so reluctant, as it is precisely this confident advocacy on behalf of an approach that drives scientific progress. One attempts to work out the details of a paradigm or a research program as far as one can, until it degenerates far behind its competitors; it is in this regard that Lakatos writes of “the rationality of a certain amount of dogmatism” (Reference Lakatos, Lakatos and Musgrave1970, 175). Progress in science is best ensured not by demanding of individuals that they be open to everything but, rather, by distributing the resources of a community into various lines of research and letting each of these programs doggedly pursue its own course. When this occurs, communities end up being responsive to criticism—those programs that are subject to serious, legitimate criticism degenerate or become overwhelmed with anomalies, to the point that they finally die out. This responsiveness to criticism does not take place at the level of the individual, however: it takes place at the level of the social.

The point of this discussion is not to suggest that either Lakatos's or Kuhn's theory of science is entirely satisfactory; on the contrary, both suffer from well-known difficulties. It is also clear that the notion of individuals as advocates requires further elaboration, as there are many respects in which advocacy can be taken too far. As both Kuhn and Lakatos acknowledge, paradigms or research programs can be defended for too long, and while it is epistemically beneficial for individual scientists to be committed to certain aspects of a paradigm or research program—for example, the hard core of a research program—there are other areas in which such commitment is misplaced. Spelling out the details must await another occasion. There are strong reasons, however, for believing that the conception of individual epistemic agents implicit within the theories of Kuhn and Lakatos is much more plausible than that presupposed by the ideals of Mill and Longino. Not only more plausible—it is also a conception that is much more suitable to a truly social theory of science. To conceive of individuals as advocates rather than unencumbered selves is to view individuals as intrinsically socially located and thus committed to certain aims, evaluative criteria, and experimental techniques as a result of the fact that they are socialized within certain contexts. On this view, it is clear that the development of scientific knowledge, even in the ideal, requires communities in which intellectual labor is dispersed over a wide range of approaches, methodologies, and so forth. On this view, responsiveness to criticism and the avoidance of intellectual stagnation result not primarily from the complete open-mindedness of individuals but, instead, from epistemically well-designed social systems.