In the beginning, God and Lilith and Adam sat in the garden negotiating the construction of nature. Lilith, who was farsighted, quick and strong of mind, negotiated well and the world began to be pleasing to women. But God and Adam were not pleased; they muttered together; and they banished Lilith from the garden and they made a pact, lt was agreed that Adam could decide what was what if God could be constructed as almighty; they both vowed to claim that nature was a given and that Adam could know it and name it. The next day, Adam and almighty God constructed Eve and told her it was all over before she got there and women just were second rate by nature. Now Eve and her daughters doubted this in their hearts, but many years passed before the daughters of Eve—moved by the power and desire of Lilith, a living spirit in the root of their minds—reopened negotiations and reshaped the world.
Epistemological Individualism and the Private Language Argument
Modern philosophy has been deeply committed to epistemological individualism, the assumption that the individual is the source of and principle agent in the production of knowledge. Thus, Descartes in-vites us in the Meditations to conceive of an isolated individual who wants to know whether his sensory ideas allow him to know the world beyond his own mind. The isolation of the individual mind, alone with its sensory ideas, is the fundamental situation of the epistemic agent in the Cartesian project, and the fundamental epistemological project is to show that the isolated, individual mind can be sure that he has knowledge of the external world.
And most empiricists from Locke through Russell to the present have also taken it for granted that knowledge begins with the sensory ideas, impressions, data, or other mental content of a single individual.
Here the project is to show how, using only an economical n u m b e r of innate mental abilities, an individual can generate the structure of knowledge with which we are all familiar. Thus, Locke argues in Book II of the Essay Concerning Human Understanding that the individual has simple sensory impressions, forms simple ideas based upon them, and manipulates these ideas—compounding them, abstracting them, and so on—to generate complex ideas. Not until he sets out his phi-losophy of language in Book III does the society or community of speakers of English appear. But when he turns, in Book IV, to explain the various ways an individual connects his ideas in order to produce knowledge, the community disappears again.
Even Quine, who rejects “ o l d epistemology,” assumes that the in-dividual is the proper object of attention for the new naturalized ep-istemology:
The old epistemology aspired to contain, in a sense, natural science; it would construct it somehow from sense data. Epistemology in its new setting, conversely, is contained in natural science, as a chapter of psy-chology.... We are studying how the human subject . . . posits bodies and projects his physics from his data, and we appreciate that our po-sition in the world is just like his. Our very epistemological enterprise, and the psychology wherein it is a component chapter, and the whole of natural science wherein psychology is a component—all this is our own construction or projection from stimulations like those we are met-ing out to our epistemological subject.1
The psychology Quine has in mind here is not social psychology but the neuropsychological science of the individual brain.
These epistemologies begin as if solipsism characterized the orgi¬
nary h u m a n epistemic scene. And they are often criticized for their failure to get out of the solipsistic moment. But they do not intend to be solipsistic epistemologies; instead, as Hilary Putnam has said of Carnap, he understood himself to be arguing for methodological so-lipsism, not real solipsism: the methodological solipsist is an individ-ual who “ h o l d s that everything he can conceive of is identical (in the ultimate logical analyses of his language) with one or another complex of his own experiences. What makes him a methodological solipsist as opposed to a real solipsist is t h a t . . . everybody is a (methodological) solipsist.”2 Thus, methodological solipsism is distinguished from real solipsism in that the real solipsist believes himself or herself to be the only person (or mind) in the universe; methodological solipsism, how¬
ever, suggests that when each individual learns language and produces knowledge, each does so, for all intents and purposes, alone, in the same way that he or she would if he or she were the only person in the universe. Epistemological individualism is marked by this as-sumption that, although other people exist (so real solipsism does not obtain), knowledge is produced by each person alone from his or her own experiences; epistemologically, each is a methodological solip¬
sist.
Individualist epistemologies set u p the agent of knowledge as a methodological solipsist. And inasmuch as the methodological so¬
lipsist learns language and produces knowledge privately, these ep-istemologies presume that private languages are possible. This pre-sumption is found not in explicit claims that there exists only one person or one mind in the world but in analyses according to which language and the knowledge embodied in it are produced as if there were. The new-born field linguist brings order out of his or h e r sense data, sensations, or experiences. Experiences that we would describe as “ m o t h e r pointing” or “saying ‘ r e d ’ ” are just m o r e sense data, m o r e experiences, more sensory input to b e ordered and conceptualized.
From such private experiences the epistemologically independent in-dividual is supposed to produce all the important distinctions and concepts we take for granted every day. The private language turns out to be our language after all.
It is this feature of a private language—that it is essentially our language, containing the fundamental distinctions of the language we speak now—that gave rise to Wittgenstein’s attack on the very possi-bility of a private language. Thus, Wittgenstein pointed out that the individual who speaks our language knows, inter alia, the difference between true and false statements and so can make a distinction be-tween truth and falsehood, truth and illusion, or truth and any n u m b e r of other “infelicities,” but Wittgenstein argued that the speaker of a private language cannot make such distinctions.3 He is unable to do so because making the distinction between truth and falsehood re-quires an ability to distinguish true statements from those that seem true (and false statements from those that seem false). Against the possibility of any private language, Wittgenstin argued that there is no way for the isolated individual—who doesn’t yet have any concepts including especially concepts of “ s a m e ” and “different,” the concept of pointing to something and naming it, or the concept of r e f e r e n c e -to make this distinction between statements that are true and those that seem to him true.4 Appeals to his memory are no help because he must be able to distinguish veridical memories from those that merely seem veridical in order to use them to correct his statements.
Therefore, the isolated individual cannot produce language—much less the knowledge it emobodies; language must be public, and this m e a n s that two or m o r e people are necessary for concepts like truth and reference to work. Alone, one person cannot make the distinction between how things are and how they seem, but two or m o r e can make it—though their agreement doesn’t guarantee truth or successful reference; rather, Wittgenstein points out, one checks his belief that a and b are the same against the beliefs of other people. The possibility of “ c h e c k i n g ” how it seems to me by asking another person allows, because it entails, the possibility of correcting my belief; and this possibility is just the possibility that I might be correct or mistaken, that my belief might be true or false.
In defense of private languages, philosophers have had recourse to innate ideas (no longer implanted by God in the mind but by evolu-tionary forces in the brain), to the self-identifying nature of private experiences, or to innate “ a b i l i t i e s ” to make certain distinctions.
These “ b l a c k box” defenses insist that early on in the life of language someone must “ j u s t know" when two things are the same and must
“just r e m e m b e r ” them and so have made the distinction between true and false statements about those things. But these arguments reveal just how deep-seated epistemological individualism is. Philosophers like Carnap did not intend methodological solipsism to collapse into real solipsism;5 instead, Carnap assumed there to be other minds and a natural world available to science, but he wanted knowledge to be p r o d u c e d (in principle) by each man alone. The originary alternative—
that a community of individuals together decided (and continues to decide) when two things are similar enough to treat in the same w a y -is not taken up. But th-is alternative -is made requ-isite by Wittgenstein’s attack on the possibility of a private language. Any general concept allows or enables users to classify together or pick out two instances of that concept as similar; general concepts have, therefore, some-times been thought of as rules or as rule-governed. But Wittgenstein argued that rules do not wear on their sleeves directions for following them or even criteria for what counts as following them. And from the fact that classifications change over time with changing h u m a n needs and interests, we can see that people sometimes decide, for good reasons (or for bad ones), to change what counts as “ s i m i l a r to X.” We might say that they decide to change the rules. Wittgenstein’s argument against the possibility of a private language can be read as showing why it is meaningless for a person who is in principle alone to try to make such decisions. These decisions, then, must be made by two or more people. And these are not just decisions about how to use words but about which things are similar and so belong in the
same class; they are, then, decisions about classificatory beliefs. The impossibility of a private language means that private beliefs are not possible in the way that empiricists from Locke to Carnap thought.
And the proper inference to draw from this impossibility is that two or m o r e epistemic agents are required for the possibility of language and, hence, for the possibility of belief in general and, because knowl-edge is a species of belief, for the possibility of knowlknowl-edge in particular.
If Wittgenstein is right and the individual is not linguistically prior to the community, then the individual cannot be epistemically prior either. And it follows that the epistemic community cannot be com-prised of a set of epistemically independent individuals; we must, therefore, begin to view the community as comprised of epistemically interdependent individuals. Moreover, any adequate epistemology must analyze knowledge first in terms of the community and only then attend to the individual. Idealized models of epistemic agency proceeding as though the individual were, if not the source, then certainly the principle agent of knowledge, are at worst mistaken and at best put the epistemological cart before the horse. We will take it as an axiom, therefore, that the epistemological community is the primary agent for the production of knowledge and that any adequate epistemology must account for knowledge in social terms. Whatever else we may wish to say about knowledge, we must recognize that it is a social affair.6 This in turn will allow us to see what cannot be seen by individualist epistemologies: the c o m m u n a l nature of knowledge production and the ways in which the politics of gender, class, and other axes of oppression are negotiated in the production of knowl-edge.7
Epistemological Decisions
Epistemic decisions provide one important site for the intersection of gender politics with the productions of knowledge that may appear impervious to such influences. In this section, we will see how this is so even on an empiricist understanding of knowledge inasmuch as philosophers working in empiricist traditions have developed models of knowledge to which we can turn for an initial understanding of epistemic decisions. Mary Hesse’s Network Model of scientific theo-ries is empiricist in the broad sense that it insists upon empirical adequacy as the primary aim of any scientific theory of the natural world. Although her model is offered as an account of scientific knowl-edge and has been taken by many philosophers to capture the pro-duction of knowledge by an individual, we can use it as a first ap-proximation of knowledge produced by an epistemic community. The
Network Model is particularly useful for making it clear that episte¬
mological decisions are necessary even when our first c o m m i t m e n t is to empirical adequacy.
Hesse reminds us that physical situations have indefinitely many aspects and at any given time we notice only some of these; it follows that every time we notice things in the world or observe things, we drop out information that could be taken up at other times or by other people. The information that is dropped includes, of course, all the ways in which the p h e n o m e n a we observe are similar to other phe
nomena. Hesse understands scientific laws to classify p h e n o m e n a on the basis of resemblances among them. Thus, when the scientist es
tablishes a law, she picks out some of the respects in which phenom
ena resemble one another and ignores their differences and the dif
ferent ways in which they resemble yet other phenomena. Any scientist is, then, constantly faced with decisions as to whether two things are similar in some respects and different in others, the question becomes,
“Which respects are m o r e important, the similar ones or the dissimilar ones?” When the data are all in—here observations of the repects in which p h e n o m e n a do and do not resemble on another—decisions must be made about which data are significant. This is a fundamental case of “ i n t e r p r e t i n g the data.” Data alone, observations alone, do not determine a law or generalization; for example, we observe that whales swim in the water and so are like fish, but we also observe that they are viviparous like mammals. Are they fish o r mammals?
Because similarity is not transitive, a decision must be made on grounds other than observed similarity. That is, b may resemble a and b may resemble c, b u t a and c do not therefore resemble one another;
how, then, should we classify b? As an α or as a c? Because any decision here is underdetermined by the data, it has to be determined on other grounds.
One criterion at work in such a case is logical coherence throughout the system; however, we cannot claim that this criterion alone is suf
ficient to account for theory production. Scientists do not always de
cide between conflicting observations on the grounds that one gen
eralization provides c o h e r e n c e with the greatest n u m b e r of other generalizations. The problematic generalization may instead be the occasion to decide that most of the generalizations in the theory are wrong.
At this point, the mainstream philosphers who adopt a network model argue that scientists either do or should have recourse to cog
nitive virtues. Scentists hold or should hold certain assumptions about what constitutes good systems of laws or “ g o o d theories.” Just so, Quine has argued, the assumptions that good theories are “ c o n s e r v ¬
ative” or are “ s i m p l e ” guide the scientist to make the decision that conserves most of what has been held true in the past or the one that makes the system simpler.8 Hesse refers to the virtues as “ c o h e r e n c e conditions” and argues that they also include assumptions such as the goodness of symmetry and of certain analogies, models, and so on.9
However, feminists, as critical science scholars, argue that we need to look and see what assumptions scientists actually hold to when they decide between conflicting generalizations. The feminist working hy-pothesis is that the assumptions guiding classificatory decisions may be androcentric or sexist.
The assumption of some cognitive virtue(s) can determine which system of beliefs is desireable, but so can the assumption of some other principle—for instance, that male behavior is the n o r m , that male behavior is crucial to evolution, that hierarchies are functional, that hierarchical models are better than nonhierarchical ones, and so on. The suggestion here is that feminist studies of knowledge pro-duction in the sciences and in other areas of life should look carefully at the constraints affecting the choices people make between conflict-ing generalizations. On a network model, each belief in the system is—at any given time, though not at all times—corrigible, so there is nothing theoretically to prevent us from discovering that even the most innocent choice is constrained ultimately by an androcentric or sexist assumption.
The flexibility of any system of beliefs means that choices among beliefs can be made that allow at once some degree of empirical adequacy, of coherence, of fruitfulness, simplicity, and faithfulness to preferred analogies or models and the maintenance of androcentric or sexist assumptions. Thus, the model makes it clear that even good scientific theories, by all the traditional criteria, can be androcentric or sexist in the sense that a sexist or androcentric assumption con-strains the distribution of truth values throughout the system.
We can see from this model that although at any one time we must hold most of our beliefs beyond question, there are still times when we must make epistemic decisions. And although the u n d e r d e t e r m i ¬ nation of scientific and other beliefs by evidence does not entail that social assumptions influence the agents’ choice of beliefs, nevertheless we can no longer assume that all decisions are based solely on tech-nical grounds and are politically innocent. Instead, we must look at each case to see what constrains the choice.
Micronegotiations
Sociologists who have looked at the production of scientific knowl-edge to see how scientists decide which beliefs to adopt have found
t h a t s u c h d e c i s i o n s a r e u s u a l l y n e g o t i a t e d . B a s e d u p o n e x t e n d e d p a r ¬ t i c i p a n t o b s e r v a t i o n of s c i e n t i s t s ’ daily l a b o r a t o r y w o r k , K a r e n K n o r r C e t i n a n o t e s t h a t k n o w l e d g e is p r o d u c e d , n o t by t h e l o n e s c i e n t i s t b u t t h r o u g h s o c i a l i n t e r a c t i o n :
All laboratory studies of which substantial results are available dem-onstrate the interactive basis of scientific work, whether they address the p h e n o m e n o n explicitly or not.10
A n d t h e s e i n t e r a c t i o n s u s u a l l y t a k e t h e f o r m of n e g o t i a t i o n : Studies of scientific “ r e a s o n i n g ” in the laboratory, during controversies, and generally on occasions when scientists communicate with each-other, tend to document the negotiated or socially accomplished char-acter of technical outcomes. Whether it is the nature of the things one
“sees” in scientific observation, the p r o p e r conduct of an experiment, or the adequacy of a theoretical interpretation, scientific agreement ap-pears to be open to contestation and modification, a process often re-ferred to as “ n e g o t i a t i o n . ” Through contestation and modification, the meaning of scientific observations as well as of theoretical interpreta¬
tions tends to get selectively constructed and reconstructed in scientific
tions tends to get selectively constructed and reconstructed in scientific