20th World Congress of Philosophy Logo

Theory of Knowledge

Defending Longino's Social Epistemology (1)

K. Brad Wray
University of Calgary

bluered.gif (1041 bytes)

ABSTRACT: Though many agree that we need to account for the role that social factors play in inquiry, developing a viable social epistemology has proved to be a difficult task. According to Longino, it is the processes that make inquiry possible that are aptly described as social, for they require a number of people to sustain them. These processes not only facilitate inquiry, but also ensure that the results of inquiry are more than mere subjective opinions, and thus deserve to be called knowledge. In this paper, I explain Longino’s epistemology and defend it against criticisms recently raised by Kitcher, Schmitt and Solomon. Longino rightly recognizes that not all social factors have the same (adverse) affect on inquiry. She also recommends that we reconceptualize ‘knowledge,’ distinguishing knowledge from opinion by reference to a social standard.

bluered.gif (1041 bytes)

Though it is agreed that epistemologists need to account for the role social factors play in inquiry, developing a viable social epistemology has proved to be a difficult task. According to Longino, it is the processes that make inquiry possible that are social, requiring a number of people to sustain them. These processes, she claims, not only facilitate inquiry, but also ensure that the results of inquiry are more than mere subjective opinions, and thus deserve to be called "knowledge." Here, I want to both explain and defend Longino's epistemology.


Longino defines her account of scientific knowledge relative to positivist and wholist accounts. Though many regard positivism as offering an untenable account of science, because "no comparable sweeping and detailed philosophical view has replaced it," Longino believes that it still needs to be reckoned with (L1990, 21). Wholists are significant because they have been the greatest critics of positivism. After presenting these accounts, and explaining the difficulties that Longino has with them, I will present Longino's own account of scientific knowledge and inquiry.

This discussion focuses on two issues: the relationship between evidence and hypotheses; and, the role of "contextual" values in inquiry. Longino contrasts contextual values with constitutive values. The latter, the "values generated from an understanding of the goals of scientific inquiry," "are the source of the rules determining what constitutes acceptable scientific practice or scientific method" (L1990, 4). That these values influence inquiry is not a problem. But the former, "personal, social, and cultural values," are thought to threaten the integrity of scientific inquiry (L1990, 4-5).

According to positivists, "the fundamental base of inquiry, the source of confirming or disconfirming instances, is a set of ... observation statements that are established independently of any theory" (L1990, 26). (2) Observation statements, expressed in a theory-neutral language, provide a foundation for our theories. Theories are true insofar as they are confirmed by observations. Further, positivists construe the relation between evidence and hypotheses to be syntactic (L1990, 23). Consequently, "what would count as evidence for a hypothesis is determined by the form of the hypothesis sentences and evidence sentences not by their content" (L1990, 48). The criteria for confirmation is similar to "the formal criteria for the validity of deductive arguments" (L1990, 23). By construing hypotheses and evidence to be related syntactically, positivists ensure that "inference to a hypothesis is not mediated by possibly value-laden assumptions" (L1990, 48). Positivists regard such assumptions as a threat to the integrity of scientific inquiry.

They acknowledge that scientific inquiry is not completely value-free, allowing "for a subjective, nonempirical element in scientific inquiry by distinguishing between a context of discovery and a context of justification" (L1990, 64). Though values may play a causal role in the context of discovery, "in the context of justification these generative factors are disregarded, and the hypothesis is considered only in relation to its observable consequences, which determine its acceptability" (L1990, 64-5).

Consider the wholists' account. (3) Wholists reject the positivists' "fundamental assumption of the independence of observation from theory" (L1990, 26). They argue that "confirming or disconfirming observations ... cannot be specified independently of a theory but are themselves given content, at least in part, by theory and described in language whose meaning [is] dependent on the whole of a theory" (L1990, 27).

The consequence of theory-ladenness is incommensurability: two (or more) opposing theories accounting for the same phenomena cannot be compared with each other and against 'the facts' in any way that enables us to determine which is false and which, if any, true. (L1990, 27)

Because competing theories are incommensurable "theory choice in science is no longer a uniquely pure expression of rationality and objectivity but is described as nonrational or irrational, and certainly not evidence determined" (L1990, 27). Wholists believe that a scientist's values may be responsible for determining which theory she accepts.

Longino finds both accounts unacceptable. Consider her criticism of positivism. First, Longino claims that the positivists are mistaken in believing that there is a theory-independent language of observation statements. She claims that "the absolute and unambiguous nature of evidential relations presented in the positivist view cannot accommodate the facts of scientific change" (L1990, 81). Sometimes advocates of competing theories appeal to the same body of data as support for their competing theories. Were evidence as unequivocal as the positivists suggest, this should not happen.

Second, Longino believes that the positivists are mistaken in regarding the relation between evidence and hypotheses as syntactic.

Data ... do not on their own ... indicate that for which they can serve as evidence. Hypotheses, on the other hand, are or consist of statements whose content always exceeds that of the statements describing the observational data. There is, thus, a logical gap between data and hypotheses. (L1990, 58)

This gap allows contextual values to influence decision making (L1990, 52).

Third, Longino objects to the positivists' account of the role of values in inquiry. If the positivists are correct, then good science should generally proceed according to the positivists' prescriptions. But, the "historical work of the wholists' decisively refutes the empiricists' claim that their prescriptions can also function as descriptions of scientific practice" (L1990, 28). Scientists are more affected in their decision making by values than positivists suggest. (4)

Longino is critical of wholism too. First, Longino argues that wholists "create a bond between evidence and hypothesis impossible to break and even destroys ... the concept of evidence as something to which one can appeal in defending a hypothesis" (L1990, 57). If wholists are right, evidence for one theory could not compel someone who accepts a competing theory to change her mind. Data would only appear to support the theory if one already accepts the theory it is intended to support.

Second, Longino believes that wholists have exaggerated the significance of incommensurability. They claim that it is because of radical incommensurabilities that theory choice is not evidence determined. Longino claims that "the incommensurability of theories in the wholist view cannot do justice to the lively and productive debate that can occur among scientists committed to different theories" (L1990, 81). Wholism gives rise to a paradox: "if we regard the meaning of a term occurring in one theory as changed when it occurs in some other theory, then we cannot say that any theories contradict one another" (L1990, 28).

Third, Longino believes that because wholists claim that hypothesis acceptance that is not based on evidence is irrational, they implicitly accept the positivists' conception of evidential relations as syntactic (L1990, 57-8). As suggested earlier, such an account of evidential relations is unacceptable.

Longino develops an alternative account of scientific knowledge, "contextual empiricism." The following two features constitute the core of Longino's account.

First, Longino offers an alternative account of the relation between hypotheses and evidence. Hypotheses and evidence are related by assumptions that scientists bring to their inquiries. According to Longino, "a state of affairs will only be taken to be evidence that something else is the case in light of some background belief or assumption asserting a connection between the two" (L1990, 44). "In the absence of any such beliefs no state of affairs will be taken as evidence of any other" (L1990, 44). Thus, contextual background beliefs bridge the gap between hypotheses and evidence. And,

relativizing evidential import to background assumptions thus involves abandoning the attempt to specify the relation between evidence and hypotheses by means of syntactic criteria and seeing this relation as involving substantive assumptions instead. (L1990, 59)

Second, Longino suggests that we change our understanding of the nature of scientific method. We must "return to the idea of science as practice" and "regard scientific method as something practiced not primarily by individuals but by social groups" (L1990, 66-7). This "shift in perspective" is required because

the application of scientific method, ... of any subset of the collection of means of supporting scientific theory on the basis of evidential data, requires by its very nature the participation of two or more individuals. (L1990, 67)

Longino situates her account between positivism and wholism, avoiding the weaknesses of both.

First, by invoking background assumptions Longino is able to explain how the same data can support competing theories or hypotheses. Advocates of different theories bring to their inquiries different background assumptions, and "in the context of their differing background ... assumptions different aspects of the same state of affairs [become] evidentially significant" (L1990, 47-48). The apparent instability of evidence that leads wholists to claim that competing theories are incommensurable is due to the fact that the states of affairs that function as evidence can be described in different ways, and different descriptions will draw our attention to different aspects. But, Longino insists that hypotheses, background beliefs, and the states of affairs that count as evidence are independently specifiable (L1990, 57).

Second, the background assumptions that facilitate our inferences from evidence to hypotheses make room for the influence of values in inquiry. Because the assumptions that mediate our evidential reasoning are value-laden, an inquirer's values will shape scientific knowledge. Nonetheless, Longino insists that this need not threaten the integrity or objectivity of science, as positivists suggest. Longino construes the demand for objectivity as the demand "to block the influence of subjective preferences at the level of background beliefs" (L1990, 73). When the background assumptions that play the mediating role in evidential reasoning do not reflect merely subjective preferences, a community's methods are as objective as is possible.


I will now examine and address three criticisms that have recently been raised against Longino's account of knowledge and inquiry in an effort to clarify and defend her view.

Philip Kitcher believes that Longino's view collapses into relativism (K1994, n.26, 132). He argues that because she believes that "the only useable notion of truth is one that identifies truth with some type of acceptance" (K1994, 122), her account of knowledge does not provide inquirers with a basis from which they can make principled judgements. Kitcher, though, doesn't believe that social factors play such a prevelant a role in inquiry, and argues that the traditional correspondence theory of truth is more plausible than Longino suggests.

Kitcher's criticism implies that Longino's view does not differ significantly from sociological accounts of science. This is a mistake. Longino distinguishes her view from sociological accounts in two respects. First, unlike the Strong Programmers, Longino does not believe that

science is socially constructed in the sense that the congruence of a hypothesis or theory with the social interests of the members of a scientific community determines its acceptance by that community (rather than a congruence of theory/hypothesis with the world). (L1994, 136)

Longino grants that how we describe things is a matter of convention (L1990, 42). But once we commit ourselves to a way of describing, the right description is not merely a matter of convention. Consequently,

the fact that the boundaries of classificatory categories are conventional and determined by a linguistic community does not show that the boundaries are adopted because of their semantic relation with social values. (L1994, 136) (5)

Second, Longino believes that sociological accounts of science mistakenly make no distinction between knowledge and opinion. As Longino expresses it: "the fate of knowledge as it is treated in social theories of science is to collapse into what is believed or what is accepted" (L1994, 138). Such accounts "are too concerned with finding the criteria that do govern scientific selections ... not the criteria that ought to govern them" (L1994, 137-8).

Despite the fact that Longino accepts the traditional knowledge/opinion dichotomy, her conception of knowledge differs significantly from traditional conceptions. She argues that knowledge is the outcome of interaction between people that is mediated by the appropriate social processes (L1994, 142). Such processes, she suggests, enable us "to transform the subjective into the objective" (L1994, 144). Longino calls the interaction that leads to knowledge "transformative criticism." She suggests that the following four features of "the design and constitution of a community ... facilitate transformative criticism and enable a consensus to qualify as knowledge": public forums for criticism; uptake to criticism; publicly recognized standards; and, equality of intellectual authority (L1994, 144-5). Insofar as our interactions with others satisfy these procedural conditions, the outcome of our inquiries deserve to be called knowledge. But, Longino neither reduces truth to some form of acceptance, nor even identifies truth as the end of inquiry.


Frederick Schmitt argues that Longino's view is incoherent. He attributes a "multiperspectival or consensus theory of rational choice" to Longino. Given such an account, "the rational theory choice is the choice that is accepted from each of various perspectives representing opposing interests" (S1994, 26). Advocates of such an account of rationality, Schmitt suggests, claim that, though interests "inevitably cause theory choice," our aim is to reduce their effects (S1994, 26). Schmitt suggests that such accounts are incoherent; they regard the influence of interests on theory choice as ineliminable, and yet seek to alleviate their effects (S1994, 26).

Indeed, Longino does believe that it is impossible to eliminate the effects that social factors have on decision-making. But, she doesn't claim that we should seek to eliminate the effects of social factors on decision-making. Some social factors, like background assumptions, play a significant role in scientific reasoning, and it is unreasonable to expect to eliminate the effects of such factors. Longino claims that we should mobilize the right sorts of social factors — those that permit transformative criticism — in order to ensure that our inquiries result in knowledge, rather than mere opinion. It is only the effects of subjective preferences that must be eliminated. Because Longino distinguishes between (1) social factors that permit transformative criticism and (2) merely subjective preferences, she is not guilty of the incoherence that Schmitt identifies.

Schmitt also misunderstands what role consensus plays in Longino's account. (6) Longino doesn't believe that the end of inquiry — truth, knowledge, or rational choice — is determined by consensus. She believes that there must be a consensus about background assumptions in order for inquiry to be possible. (7) She claims that

observational data consist in observation reports that are ordered and organized. This ordering rests on a consensus as to the centrality of certain categories, the boundaries of concepts and classes, the ontological and organizational commitments of a model or theory, and so on. (L1994, 140)

Consensus plays a crucial role in organizing our observation reports so that they can function as data.

A similar consensus is required for us to reason effectively. Reasoning involves "bringing the appropriate considerations to bear on a judgment" (L1994, 141). And, "what counts as an appropriate consideration, as a reason, is determined and stabilized through discursive interaction" (L1994, 141-2). Essential to the process of reasoning are the assumptions common to those who are part of one's social context, one's community. As Longino explains, "every assumption upon which it is permissible to rely is a function of consensus among the scientific community" (L1994, 142). It is because there is a consensus about the assumptions that one draws on in one's reasoning that one's actions count as reasoning at all.


Miriam Solomon argues that Longino is mistaken about the role that the community plays in scientific inquiry. Solomon claims that though Longino rightly "regards some social processes as constitutive of scientific objectivity," she "envisages these social processes as practices of criticism that help individual scientists to reason better" (S1994, 219). Solomon argues that Longino's account is too individualistic. Solomon "argues for a more social epistemology," one that recognizes that the community is the locus of scientific rationality (S1994, 219). As Solomon explains, "social groups can work to attain and even recognize epistemic goals without individual rationality or individual cognizance of the overall epistemic situation" (8) (S1994, 219)

Longino argues that in our efforts to account for the influence of social factors on inquiry, "individuals are not to be replaced by a transcendent social entity" (L1994, 143). She claims that

to think or see in these terms is still to see the knowledge-productive feature of a knower as internal to the knower, or as a matter of the relation between knower and known, rather than a matter of relatedness of the knower to other knowers. (L1994, 146)

It is the processes that mediate our interactions with each other that are aptly described as social, not the knowing agent.

Further, Longino believes that if we construe the community to be a knowing agent we are at risk of overlooking the significance of the role that individuals play in inquiry.

Without individuals there could be no knowledge: it is through their sensory system that the natural world enters cognition; it is their proposals that are subject to critical scrutiny by other individuals, their imaginations which generate novelty. (L1994, 143)

Though Longino rejects the notion of a "transcendent social entity," there is a sense in which her account of inquiry is not aptly described as individualistic. Epistemologists have traditionally construed knowledge as involving a relationship between the knowing agent and the object of knowledge. Such epistemologies are described as individualistic because they focus on individuals and their relationship to the world. Longino recommends that epistemologists shift their attention from the relationship between knower and known to the processes that mediate our interactions with others. Knowledge, she claims, is the outcome of the appropriate sorts of social interactions.

In summary, I have tried to show how robust Longino's epistemology is. Longino offers a viable alternative to positivism and wholism, one that recognizes just how thoroughly values influence inquiry. I have argued that her critics have misunderstood her view and have thus failed to raise insurmountable challenges for her. Though Longino acknowledges the significant influence that social factors have on inquiry, she also recognizes that not all social factors have the same affect. A key component of her account of inquiry is the way she reconceptualizes "knowledge." Though Longino insists on the traditional knowledge/opinion distinction, she proposes that we distinguish knowledge from opinion by reference to a social standard.

bluered.gif (1041 bytes)


(1) I gratefully acknowledge the very useful feedback I received from numerous people on earlier drafts of this paper. Lori Nash, Kathleen Okruhlik, and John Nicholas each critically read numerous drafts. And, Alison Wylie, Bruce Freed, Cheryl Misak, and Marty Kreiswirth also provided useful feedback.

(2) Longino identifies Hempel and Carnap as typical positivists.

(3) Longino regards Hanson, Kuhn, and Feyerabend as wholists.

(4) In "A Locus of Value in Science," Kathleen Okruhlik (1985) argues, similarly, that "even if we grant the methodological rationalists their claim that theory choice is entirely subject to objective evaluation, nonetheless we can simultaneously claim that the very content of scientific theories is to a significant extent socially determined" (Okruhlik 1985, 1). For example, Okruhlik suggests that social factors play a significant role in determining what theories are generated, and thus "how our options came to be determined in the particular ways that they are" (Okruhlik 1985, 13).

(5) For a recent restatement of the Strong Programme's programme see Barnes', Bloor's and Henry's (1996) Scientific Knowledge.

(6) Kitcher makes a similar mistake when he claims that Longino "identifies truth with consensus belief in societies that follow certain types of procedures" (Kitcher 1994, n.26, 132).

(7) I argue a similar point in my (forthcoming) "The Role of Solidarity in a Pragmatic Epistemology."

(8) I have criticized Solomon's account of rationality elsewhere. See my (1997) "Rational Communities." I argue that though social groups can realize epistemic goals without individual rationality, when they do, they are not aptly described as acting rationally.


Barnes, B., D. Bloor, and J. Henry. 1996. Scientific Knowledge. Chicago: The University of Chicago Press.

Kitcher, P. 1994. Contrasting Conceptions of Social Epistemology, in F. Schmitt's (ed.) Socializing Epistemology:The Social Dimensions of Knowledge. Lanham: Rowman and Littlefield Publishers, Inc., pp. 111-134.

Longino, H. 1990. Science as Social Knowledge. Princeton: Princeton University Press.

———, 1993. Subjects, Power, and Knowledge: Description and Prescription in Feminist Philosophies of Science, in L. Alcoff's and E Potter's (ed.) Feminist Epistemologies, New York: Routledge, pp. 101-120.

———, 1994. The Fate of knowledge in Social Theories of Science, in F. Schmitt's (ed.) Socializing Epistemology:The Social Dimensions of Knowledge. Pp. 135-157.

Okruhlik, K. 1985. A Locus of Value in Science. London ON: unpublished manuscript.

Schmitt, F. 1994. Socializing Epistemology: An Introduction Through Two Sample Issues, in Socializing Epistemology: The Social Dimensions of Knowledge. Pp. 257-287.

Solomon, M. 1994. A More Social Epistemology, in F. Schmitt's (ed.) Socializing Epistemology: The Social Dimensions of Knowledge. Pp. 217-233.

Wray, K. B. (forthcoming) The Role of Solidarity in a Pragmatic Epistemology, Philosophia: Philosophical Quarterly of Israel. Vol. 26: 3-4.

———, 1997. Rational Communities, Perspectives on Science:Historical, Philosophical, Social. Vol. 5, No. 2, pp. 232-254.

bluered.gif (1041 bytes)


Back to the Top

20th World Congress of Philosophy Logo

Paideia logo design by Janet L. Olson.
All Rights Reserved


Back to the WCP Homepage