In contemporary philosophy of science it has become a truism to claim that scientific knowledge is social knowledge. Yet there is a diversity of views about what is “social” in scientific inquiry and why it is of epistemic interest (Rolin, 2004). One approach to understanding the “social” in science focuses on scientists’ social values, that is, value judgments concerning a desirable social order. A number of philosophers argue that social values are of epistemic interest because they affect, for good and bad, scientists’ assessment of theories and hypotheses (e.g., Anderson, 1995, Anderson, 2004, Kincaid et al., 2007, Lacey, 1999, Longino, 1990, Longino, 1995 and Machamer and Wolters, 2004). Another approach to understanding the “social” in science focuses on social relations among scientists. Some philosophers argue that collaboration among scientists is of epistemic interest because it contributes to the epistemic success of science (e.g., Thagard, 1999, Wray, 2002 and Wray, 2006). Others argue that a division of research effort among scientists is of epistemic interest because it contributes to the epistemic success of science (e.g., Hull, 1988, Kitcher, 1993 and Solomon, 2001). Yet others focus on the epistemic role of trust and testimony in science (e.g., Hardwig, 1991, Kitcher, 1992 and Shapin, 1994). And some philosophers use the notion of distributed cognition to analyze how the social organization of science contributes to its epistemic success (e.g., Giere, 2006). The two approaches to understanding the “social” in science are not exclusive of each other. Some philosophers, most notably Longino, 1990 and Longino, 2002, discuss both social values in science and social relations among scientists.
Gilbert (2000) introduces yet another dimension to this debate. She claims that scientific knowledge is social knowledge in the sense that it includes collective beliefs held by scientific communities. By collective beliefs she means beliefs which cannot be accounted for in a summative way. According to a summative account, a community believes that p if and only if all or most of the members of the community believe that p (Gilbert, 2000, p. 39). For Gilbert such beliefs are not properly speaking collective beliefs because a community believing that p in a summative sense can be reduced to its members believing that p. Collective beliefs in the proper sense of the term cannot be reduced to the members of the community believing that p. As Gilbert explains, “scientific communities do have scientific beliefs of their own” (2000, p. 38).
According to Gilbert, collective beliefs are held by communities as plural subjects. To say that a community as a plural subject believes that p means that the members of the community are jointly committed to believe as a body that p (Gilbert, 2000, pp. 39–41). A joint commitment creates obligations and rights among the community members (Gilbert, 2000, p. 40). A scientist’s participation in a joint commitment of the type in question requires her not to deny that p without qualification (Gilbert, 2000, p. 44). In case she denies that p without qualification, the other members of the community are in the position to call her on it (Gilbert, 2000, p. 40). A plural subject account of collective belief differs from a summative account of beliefs held by communities in an important respect. In a plural subject account of collective belief it is neither a sufficient nor a necessary condition of a community believing that p that all or most of its members believe that p (Gilbert, 2000, p. 39). A community having a collective belief that p involves a consensus but, as Beatty explains, it is a “consensus at a different level”: not agreement concerning p but rather agreement to let p stand as the position of the group (Beatty, 2006, p. 53).
My aim is to explore to what extent scientific knowledge is properly understood as collective knowledge. By knowledge I mean justified true belief or acceptance (see also Wray, 2007). Thus, collective knowledge is justified true belief or acceptance held or arrived at by groups as plural subjects.1 In addressing the question to what extent scientific knowledge is collective knowledge, I focus on the part of the question inquiring to what extent scientific knowledge is held by groups of scientists as plural subjects and leave aside the part of the question inquiring to what extent scientific knowledge is true belief or acceptance. I assume that belief or acceptance has to be justified in some sense to deserve to be called scientific.
In Section 2, I discuss Gilbert (2000) view that scientific knowledge includes collective knowledge held by scientific communities. In Section 3, I discuss Wray (2007) argument for the claim that neither the scientific community as a whole nor the various communities that constitute particular sub-fields are capable of having collective knowledge. Wray (2007) argues contra Gilbert (2000) that merely research teams are capable of having collective knowledge. I argue contra Wray (2007) that collective knowledge is not limited to research teams. As Gilbert (2000) assumes, scientific communities are also capable of having collective knowledge. However, Gilbert’s account of collective knowledge in science is limited because it does not help us answer the question of why scientific communities have an interest in collective knowledge. In Section 4, I introduce a contextualist theory of epistemic justification in order to answer this question. In Section 5, I argue that scientific communities have an interest in collective knowledge because it enables them to establish a context of epistemic justification.