A VIEW FROM THE EXTREMITIES OF THE ENLIGHTENMENT

Since World War II, Northern agencies have tried to modernize the socalled underdeveloped societies of the South so that their standards of living would catch up to those in the North. Yet there is now general agreement that standards of living have deteriorated during the development decades for the majority of those living in the underdeveloped societies-namely, those already most economically and politically vulnerable. Reevaluations of modem science and its philosophy figure in these assessments because development was conceptualized as transferring to the South sciences, technologies, and their philosophies that were presumed to be responsible for the industrial development of Europe and North America in the nineteenth and early twentieth centuries. Modem science is also at issue because of the terrifying escalation of environmental destruction in the South, as well as globally. Northern philosophies of nature seem to be implicated in that debacle.

The gender, environment, and sustainable development (GED) analyses enter these issues about the global development and environment crisis, as it is called, from the standpoint of the majority of women's lives in the South. These accounts have their origins in the early "women in development" attempts to get governmental and nongovernmental development agencies to assess the impact of development policies on women and, usually, to "add women" as recipients of Third World development benefits (as they were then understood). In the three decades since, these feminist analyses have come to challenge most of the terms in which the original issues were posed, as have other critical development analyses with which this branch of feminist thinking has interacted.

In seeking to determine just how development policies should be changed, this literature also offers resources to philosophy that are otherwise undervalued or unavailable. It emerges from locations in global politics that enable the identification of otherwise hard-to-detect assumptions in the philosophies of science that have guided development thinking and that mark these philosophies as considerably less than universally valid. Moreover, these analyses are responding to urgent practical concerns to develop more reliable patterns of knowledge, and processes of legitimating it, for projects important to the world's economically and politically most vulnerable citizens in non-Western cultures-and, indeed, to many of the rest of us. Thus they must forfeit the luxurious "handmaid" role favored in mainstream philosophy of science and much of northern science studies: these Owls of Minerva must work a second shift in daylight. Such engaged philosophy sharpens the ability to detect how philosophic assumptions function in daily life. Finally, such concerns have led the GED movement to an expanded set of coalitions with other progressive science groups. These coalitions contribute innovative critical perspectives to the project of figuring out how to extract ourselves from the now problematic elements of Enlightenment scientific rationality.1

The Enlightenment philosophies defined the growth of scientific knowledge and the social progress this was supposed to bring in ways that devalued women, nature, and "backward cultures." The new philosophies of knowledge and power emerging from the gender, environment, and sustainable development discussions and the analyses on which they draw represent the return of the Enlightenment's others-the return of women, nature, and "backward cultures" from positions of more than instrumental value (at best) in modernity's thinking. Knowledge is power, as the familiar saying goes; and it is from the extremities of knowledge-power networks that we can best perceive the limitations of how knowledge and power create and nourish each other at the centers.

The project of this paper is, first, to show how the GED debates link criticisms of androcentric, economistic, and nature-blind aspects of development thinking to challenge the epistemology and philosophy of science of "the Enlightenment dream."2 Sections 3 and 4 show how GED analyses can be supported with and, in turn, support arguments produced in post-Kuhnian and postcolonial science studies. Thus there are three influential post-World War II schools of science studies, in addition to northern feminism, that arrive, from somewhat different starting points, at common assessments of Enlightenment philosophies of science. The concluding section summarizes how philosophic positions in this collection of analyses differ from those that were centered in the Enlightenment philosophy. Obviously, this essay can only map in very broad outlines a philosophic terrain that deserves more detailed attention (some of which, of course, it has already received in the literature indicated).

GED CRITICISMS OF ENLIGHTENMENT ASSUMPTIONS

The focus below will be on three issues.3 These are how Enlightenment philosophies appear complicitous with the androcentrism, economism, and devaluing of nature in Third World development policies and practices.4 Is Development Gendered?

Four issues link the GED and other feminist analyses: women's work, gender as an analytic category, the androcentrism of science and technology thinking, and the standpoints of women of color. First, a main theme in early feminist criticisms of development was that women were being left out of development, as literacy and job-training programs were designed for men only, and men were given favored access to income-generating work. Often the only attention women officially received from development planners occurred under the heading of controlling women's reproduction. Moreover, as men were drawn into urban manufacturing, mining, or plantation agriculture, women were left to become a higher proportion of rural populations, with increased responsibility for the care of the young, old, and disabled and with fewer social and environmental resources to sustain life, environment, and community. Development policies bypassed women, the argument went.

Or were women actually left outside such modernization planning? A second round of analyses showed that these very processes of "leaving women out" actually provided necessary new resources for modernizing national economies. Achieving economic growth required increasing women's unpaid domestic labor, enticing or forcing women into the lowest-paid manufacturing and agricultural labor, and appropriating their inherited land rights. Their land rights tended to shift to men when only men were taught to farm in modern, scientific ways. At other times, these rights were directly appropriated, so the land could be used for export production by denying women access to community-owned common agricultural, forest, or grazing areas. Peasants as a class, both women and men, also suffered from some of these forms of appropriation (Mies 1986, Shiva 1989).

Thus the dedevelopment of women and peasants was a necessary condition for development in the South. "Progress for humanity" meant regress for women (and peasants), as Northern feminist historians had put a similar point.5 Did the scientific and technological rationality being transferred from North to South include directions or, at least, permission for such banditry (as it could reasonably be called)? At any rate, the language of development, modernization, and scientific progress was being used to obscure the actual mechanism responsible for much development success. Structural adjustment policies intended to resolve the debt crisis of the 1980s followed the same pattern, further undermining the conditions of women in the developing countries. Here the International Monetary Fund and the World Bank ordered the indebted Southern governments to cut their social services in order to repay development loans to the Northern investing classes. Thereby women's unpaid labor was to be substituted for their formerly paid labor in state-provided educational, health, child care, and other social services so as to maintain the wealth of the most advantaged classes in the North (Sparr 1994).

Some of the feminists concerned with these labor issues also began to identify problems with the economism, the production model, of both neoclassical and neo-Marxian economists' attempts to address women's issues, and with the deteriorating environments that development policies created for women's subsistence and wage labor. These concerns lead to coalitions with relevant other groups.

A second important resource for the GED analyses was the more complex and comprehensive understanding of gender that they used. It was gender relations, not just women (as in the early "women in development" accounts), that were the object of GED concern. Gender was conceptualized not primarily as a property of individuals, but as an analytic category like race and class, through which one could understand the structure of societies and their symbolic systems. In order to understand women's situations and the meanings of the womanly or feminine in development policies and practices, one had to look also at men's situations and the meanings of the manly or masculine.

Gender, like class and race, is fundamentally a relationship. Thus GED accounts, like other feminist ones, argued that the problem with Enlightenment philosophies was not only that women had been excluded from articulating them and overtly maligned in them, but that Enlightenment standards of the human, the good, progress, social welfare, and economic growth, as well as of objectivity, rationality, good method, and what counted as important scientific problems, were all defined in terms of masculine and bourgeois interests and meanings. They were part of historically varying but nonetheless persistent androcentric and class discourses. Among its other benefits, understanding gender as structural and symbolic facilitated the integration of GED analyses with other groups' criticisms of structural and representational aspects of Enlightenment assumptions.6

This brings us to a third issue. Central themes in northern feminist science and technology accounts independently emerged or were transformed in the GED analyses. The GED discussion was created mainly in regional and international agencies rather than in the university and laboratory contexts in which northern feminist science and technology discussions were shaped.7 The androcentric structures and meanings of modem scientific and technological worlds shaped also international, national, and local development agencies and their thinking. The obscuring and often misogynous dualisms identifed by northern feminists, postcolonial studies, and the environmental movement operated through development rhetoric to shape policies that systematically discriminated against the economically and politically most vulnerable populations (Barker 1998). And the questions, the problems, that development addressed were never ones defined by women or from the standpoint of women's lives. Development was gendered, as even one document from the United Nations Commission on Science and Technology for Development put the point by the mid-1990s (Kettel 1995). Feminist concerns with the androcentric standards for objectivity, rationality, evidence, good method, and what counts as science were centered in GED analyses also.

Finally, GED discussions paralleled analyses in the writings of African American and other feminists of color in the metropolitan centers, which made possible additional coalitions. These discourses, both in the North and in the GED accounts, helped to redefine subjects of knowledge as having multiple and often conflicting identities because of their race, class, gender, ethnicity, sexuality, and other histories. They revealed the multiplicity of conflicting knowledge systems that different cultural histories will produce. They insisted on the importance of empowering marginalized racial and ethnic groups as a condition of democratic dialogue and coalition; difference, as well as affinity, must be recognized and respected. In both the South and the North, these writings have produced a powerful critique of positivism and neopositivisms, and have developed illuminating forms of feminist standpoint epistemology. (In the north, for example, Anzaldua 1987; Collins 1991)8 These feminist origins of Enlightenment reevaluations were strengthened and expanded in the GED accounts through their links to criticisms of development economism and ignorance about nature's limits.9

Is Enlightenment Philosophy Economistic?

Does the Enlightenment scientific ethos depend on a bourgeois economic model of human progress? Of course, in many respects, Enlightenment philosophies of science can usefully be read as radically democratic with respect to class, race, and other social markers, considering that it is the effective use of scientific methods (or some other property of scientific processes) that is to ensure the reliabilty of knowledge claims, rather than the social status of the knower, inherited or not. The modern liberal state is supposed to be neutral toward the often diverse conceptions of "the good" of its subgroups and their members, whether such conceptions are biological, religious, ethnic, racial, class, or gender based. Any differences in treatment must be rationally justified. And modern states' information-producing institutions must be similarly value-neutral to enable the state to achieve policies unbiased toward conceptions of "the good." Through rigorous research methods, the goal of value neutrality has been "operationalized" for the sciences.

Of course, hardly anyone thinks it defensible any more to assume that the natural sciences are really neutral to the values and interests of their cultures. Even post-Kuhnian studies-the least overtly politically radical of the science studies movements that have emerged since World War Il-have been busy showing the integrity of modern sciences with their historical eras, to paraphrase Thomas Kuhn himself (1970, 1). That is, scientific projects and their results always bear the fingerprints of the historic eras in which they emerged and that continue to find them valuable. I cannot discuss here the value-, interest-, discourse-, and method-ladenness of modem science in general (See Harding 1992; 1996; Proctor 1991). The issue instead is whether economistic values and interests permeate it.

Development was initially conceptualized as economic growth. Thus human progress was thought of in terms of increased production and consumption. In the first place, this approach failed to perceive women's work inthe household as real work, or, therefore, as activity that contained elements of a history of human progress. This conception was as prevalent in the Marxian as in the liberal analyses. Thus the need for childcare and household labor was perceived by development thinkers as a drain on maximum economic growth by peasant and working classes in the South, and as an opportunity to recruit middle- and upper-class women North and South into the "consumptionwork" that was required to keep production profitable. Poor women were to be drawn into productive agricultural or manufacturing labor and left to get childcare and domestic work done as best they could, while women in the economically advantaged classes were to devote increased time and energy to childcare and domestic tasks, regardless of the purportedly labor-saving devices and services they could use, so as to consume at higher and higher levels. Thus to feminists, conceptualizing development and human progress only in economistic production terms left women and the life of the household intensely vulnerable to exploitation.

In the second place, modernization theory routinely conceptualized population growth in developing countries as a major obstacle to raising standards of living. Population growth causes poverty, this theory insisted. From this perspective, women's bodies were a major obstacle to social progress, and coercive population control policies appeared justifiable. Finally, in the 1990s, even the United Nations Population Conference officially recognized what feminists and progressive economists had been arguing for years: it is poverty that causes population growth, not the reverse. Only many children can provide poor households with the economic and social supports that the state and the economy provide to middle-class households. Conventional Western scientific wisdom had the causal direction backwards.

A third problem with the economic growth conception of development was that nature itself limits economic growth, as feminists, environmentalists, and critics of neoclassical and Marxian economics argued. The world does not have enough resources to support today's global population even at the consumption levels of moderately well off Third World middle classes. And achieving that standard of living for today's most politically and economically vulnerable populations would require a lowering of consumption levels among the more advantaged half of the world's population that is virtually unimaginable. What political process could bring this about?

Finally, conceptualizing development in terms of greater economic productivity and consumption ignores and devalues all other "goods" that women and their cultures prioritize, such as ethical, political, aesthetic, and spiritual values.

Such considerations lead to the suspicion that rational man, who seeks information always in order to maximize his own benefits, ensures the destruction of the very conditions necessary for his survival when those benefits are conceptualized solely as economic. (This is a point about the values and standards of development and other modem institutions concerned with human progress, not those of individual scientists or institutional actors, of course.) Neither nature nor social life can be sustained when such a rationality is the dominant institutional rationality of states and transnational corporations (TNCs) and is held accountable to no other social values.

The question arises as to what extent this self-destructive rationality is an inherent feature of Enlightenment rationality. Modern sciences emerged as part of European postmedieval economic, political, and social formations. One issue is that the modern scientific ideal of value neutrality and the autonomy of knowledge seeking from social accountability makes modern sciences and technologies a "fast gun for hire." Ethics committees in scientific work sites are better than nothing, but they have no power over TNCs, which at present appear accountable to no civic groups at all. Must the sciences internalize democratic ethical and political principles in order to avoid the "fast gun" status? Another issue is whether the increased access to nature's resources that is one of modem sciences' central goals encodes or legitimates using up and degrading environments in order to benefit the groups that modern sciences serve. Nature limits economic growth and any form of human progress that requires such growth. But where is this recognized in Enlightenment philosophies of science?

An Enlightenment Philsophy of Environmental Destruction?

It would take unimaginable sacrifices by middle and upperclasses in the North and South to bring underdeveloped populations up even to standards of living of lower middle classes in these societies. What governance practices could bring that about? Furthermore, the growth models of development consistently sacrifice sustainable environments to short-term consumption goals. Natural resources are disappearing not only through consumption, but also through the effects of military activities, agricultural, urban, and other kinds of toxic pollution of air, water, and other nutrients that human and nonhuman life require.

Women suffer in distinctive ways from the limits that nature places on economic growth, and their disadvantage is passed on to children and others dependent on their energies and resources. They are frequently last in line for economic resources in their households, and disproportionately among the last within their societies. To them is assigned responsibility for doing or managing daily sustenance and the health and welfare of dependents, household, kin, the elderly, the sick, as well as their communities and environments. Moreover, manufacturing and rural wage labor expose them as well as men to toxic dangers in addition to the toxic threats endemic in poor people's household life, such as vermin, gasses from open hearths, and the like. Life- and healththreatening conditions in mining, construction, manufacturing, and agriculture make for nasty, short and brutish lives for the men as well as the women who constitute the politically and economically most vulnerable classes.

Wherever labor and interactions with nature are sex-segregated, environments are usefully conceptualized as gendered. The point here is not that environmental preoccupations with men's issues should be replaced by a preoccupation with women's issues, but that addressing men's problems does not automatically address women's issues in this case or any other.

Critics argue that the Enlightenment entrenches a faulty philosophy of nature. Nature is not a cornucopia, available to satisfy limitless desires, as in the infant's dream of mother. Moreover, sciences and philosophies of nature and of science, like all other human creations, are importantly in nature, not autonomous from it. Sciences, their philosophies, and their relations with the societies that use them should all be explained together. Yet modern philosophies' attempted isolation and immunization of natural sciences from social explanation, and their devaluation of local knowledge, have worked against such comprehensive understandings. Even the language in scientific philosophies about nature-whose principles exist outside of all human cultures and whose unique order can be identified and explained only by a universally valid science-though it has many appealing features, nevertheless obscures what happens in human environments. Yet such environments are by definition the only parts of nature with which humans interact, whether this nature is located en route to Mars, out past Jupiter, in the factory, or in the kitchen. We need philosophies of environments and of human interactions with them to replace Enlightenment philosophies of nature and science. In drawing together resources from feminism, political economy, and environmental studies, GED theory shows the importance of intellectual and political coalitions between analyses that often have been at odds with each other. Moreover, Enlightenment philosophies are our worldview-modernity's world view-and their assumptions permeate institutions and practices far removed from the studies, libraries, classrooms, and conferences where we are used to doing philosophy. Transforming Enlightenment philosophies of science requires insights from all the centers and peripheries where such beliefs have come to structure social relations and their meanings.

For those of us working in the North to refocus Enlightenment philosophies in directions more useful for guiding the production of the knowledge we need for democratic politics, the resources of two additional science studies movements will also be important. The insights of both the postcolonial and the post-Kuhnian science and technology studies support and extend the GED synthesis of Enlightenment critiques in directions valuable to northern postEnlightenment projects.10

ANOTHER PERSPECTIVE ON MODERN SCIENCE'S SUCCESS STORY: POSTCOLONIAL SCIENCE AND TECHNOLOGY STUDIES

Postcolonial science and technology studies criticize Enlightenment philosophies and their effects in development policies from the perspective of anti-Eurocentric histories of modem sciences, and also the related studies of non-Western cultures' own science and technology traditions. In these accounts, development policies are understood as a continuation of the European expansion that began in 1492. European expansion, moreover, has played a far greater role than previously acknowledged in the growth of modern science in Europe and its achievement of a unique epistemological status as capable of providing the only true account of nature's order. That is, the plausibility of the epistemology of modem sciences has depended on the success of European expansion, not just the self-regulating powers of rationality. Modem sciences need the power of national or international state institutions to legitimate and conduct scientific work. Expansionist state power makes it possible to forage in other cultures' knowledge traditions, to test hypotheses in non-European environments around the globe, and to destroy, intentionally or unintentionally, those other traditions that could have created competition for modern scientific claims and practices. This school of thought can trace roots back to the 1940s, but it began to flourish with conferences and publications in English only in the 1980s." Two of its focuses are especially relevant here: comparative ethnosciences and "science-andempires."

A new kind of comparative ethnoscience emerged in anthropology and history from the older, Eurocentric colonial frameworks. The latter had represented other cultures' knowledge traditions as the products of "savage minds," superstitions, magic, or mere speculation, inextricably mired in religious and other cultural beliefs, or as mere technological knowhow. Other cultures had local knowledge systems, but only modem science produced claims that were universally valid, according to the Eurocentric view.

The anti-Eurocentric comparative ethnoscience movement began to reevaluate the sophistication of other cultures' scientific and technological achievements and the contributions these had made to the development of modem sciences in Europe that have gone unmentioned in the conventional histories of science. It also began to use the tools of ethnography and social history to reexamine modem Western sciences as local knowledge systems, integrated into their particular cultures. For example, historian Joseph Needham shows the distinctively Christian meanings of the notion of laws of nature as it directed scientific method until the twentieth century (Needham 1969). To take another case, European expansionist thinking and its representations of the Edenic Americas, the declining cultures of Asia, the destiny of Christian Europe, and so forth, pointed toward the need for knowledge about particular parts of nature's order that greatly advanced the development in Europe of oceanography, climatology, geology, cartography, diverse engineering projects, tropical medicine, pharmacology, agricultural sciences, evolutionary biology, and many other modern sciences (see, for example, Brockway 1979, Goonatilake 1984, Kochhar 1992-93, McClellan 1992). Culturally local discourses have positive effects on the growth of science, not just the negative effects on which conventional philosophies focus.

In a second and related project, the "science-and-empires" approach, the postcolonial histories of science and technology emerged as part of the new anti-Eurocentric global histories. Here they look at how European and nonEuropean cultures have been interacting for perhaps as long as a millennium, and certainly actively in the five centuries since 1492. In such encounters, cultures exchanged beads, manufactured goods, cattle, women, and scientific and technological ideas. Thus the Voyages of Discovery and the subsequent colonial era, on the one hand, and the development of modern sciences in Europe, on the other, were conditions for each other's success (see, for example, Brockway 1979, Goonatilake 1984, McClellan 1992, Petitjean et al.1992, Sardar 1988). They continue to nourish each other through development policies, described in this literature as the continuation of colonialism by other means. That is, development policies and their scientific and technological questions primarily continue to advance European expansion and not the societies that are the policies' overtly intended beneficiaries.

What does the postcolonial literature mean by calling both modern sciences and the knowledge traditions of other cultures "local knowledge systems"? This notion has moved to center stage in all the other science movements discussed here (see Harding 1998a, 1998b; Watson-Verran and Turnbull 1995). One way to conceptualize the point is that cultures have distinctive locations in heterogeneous nature and distinctive interests in those surroundings. People living in deserts or beside oceans will tend to produce different patterns of knowledge (and ignorance). Their hypotheses (usefully) always extend considerably beyond the available evidence, which is one reason why these patterns of knowledge cannot fit together like the pieces of a jigsaw puzzle. Even in the same environment-along the Atlantic, say-different cultures will have distinctive interests in the ocean and produce different patterns of knowledge and ignorance depending on whether they are interested in fishing it, desalinizing drinking water from it, dumping garbage in it, mining minerals under its floor, or using it as a trade route or military highway.

Two more features provide local resources for cultures' knowledge projects. Different cultures have access to different discursive resources: Christian models of nature's order; organicist, mechanistic, or biblical models of nature; or environmental notions of lifeboat Earth or spaceship Earth. Each such discourse directs scientific attention to different aspects of nature's regularities and orders them in different causal configurations. Finally, there is no one scientific method or, more generally, one way of organizing the production of knowledge that can claim credit for the various kinds of knowledge of nature's order that different cultures have produced-or even that modem sciences have produced. Fruitful inquiry methods are as varied as human styles of thought and social organization.

Of course, not all local knowledge systems are equally powerful for all projects. Modem biomedicine is valuable for many purposes; but acupuncture, chiropractic, and vitamin and exercise therapies may be more valuable for some health purposes that modem biomedicine has neglected or misunderstood. Modem philosophy of science's claims to unique and universal validity obstruct our ability to think our way through such issues. The universality claims are scientifically and politically dysfunctional.

Thus these postcolonial studies replace the Enlightenment histories and epistemologies of modem sciences with more objective ones, stimulated initially by questions from the standpoint of those who have benefitted least from the development of modem sciences in Europe. Such questions propose that the sciences that were best for the West are not necessarily best for the rest of the world, and that the "Western" sciences we have today are also not good for the West. Those sciences produce systematic ignorance that endangers the human species, destroys nature and other valuable knowledge traditions, and produces antidemocratic social relations. Moreover, this literature is full of specific reevaluations of Enlightenment philosophies of science. These have focused especially on the philosophies' unearned presumptions to model unique ideals of objectivity, rationality, and good method, and on their ontologies that conceptualize nature as isolated bits of dead matter in motion. The reevaluations have criticized the idealized claims of scientific experts who are often ignorant of local conditions and alternative knowledge systems, and the rejection of internal moral and political constraints on the accountability of scientific rationality. The postcolonial studies perceive the epistemology of modem science in effect as predatory, as legitimating sciences' foraging in, while also destroying, all other knowledge systems. The unique status of modem sciences is primarily a result of that epistemology and the successes of European expansion.

Issues about women and gender have been largely absent from the grand narratives of these postcolonial accounts, except as GED writers have occasionally participated in them-for example, Vandana Shiva ( 1989). Yet their questions about Enlightenment philosophies converge with those in the GED discussions.

AFTER "AUTONOMOUS SCIENCE": POST-KUHNIAN HISTORY PHILOSOPHY AND SOCIAL STUDIES OF SCIENCE

It is useful to save for last a review of the challenges to Enlightenment thought produced by post-Kuhnian Northern philosophies and social studies of science and technology. Many would regard this school as the most conservative of the movements of interest here, and yet its analyses have often followed much the same lines as the most radical tendencies in the others, even though these Northern accounts emerge from the older historical and geographical map. One could start by pointing out that all the other literatures set out to show how modem sciences and technologies are integrated with different aspects of their historical eras-with gender structures and meanings, with local and global economic projects, with historically distinct environmental attitudes and conditions, with European expansionist and colonial projects, and with attitudes toward and conditions of women's lives under development policies. On this point alone, the post-Kuhnian science and technology studies join these other movements in challenging the conventional idea that the philosophically relevant histories of modem sciences are only intellectual ones.

Let us consider briefly two focuses of this kind of Northern Enlightenment critique that offer opportunities to expand and strengthen coalitions between the post-Kuhnian, postcolonial, and feminist projects. First, did a European Scientific Revolution appear, as part of the "European Miracle," out of the Dark Ages? All five terms of such a claim-European, scientific, revolution, miracle, and Dark Ages-now appear problematic. Historical studies of medieval Europe and its sophisticated scientific and technological activities show that thinking of the period as the Dark Ages tells more about the speaker than about medieval Europe. There was no European miracle or scientific revolution, but only a slow process that began as far back as the eleventh century. In that process, the components of what were eventually dubbed modem science came together as part of equally gradual political, economic, and social changes. Moreover, given the immense presence in medieval and early modern Europe of Islamic culture, the borrowings in early modem sciences of Egyptian mystical thought and other non-European elements, and the political fragmentation, until at least the late Renaissance, of what we now call Europe, it is misleading to speak of modem sciences as European. Nor do early modern sciences look much like what counts as science today, with their empirical claims shaped by alchemy, sun worship, astrology, and other mystical beliefs and investigatory practices-not to mention overtly Christian metaphysics and epistemology, as Needham and others point out (1954, 1969; Blaut 1993; Yates 1969).

This kind of reconceptualizing of the origins of modem science links postKuhnian accounts with postcolonial studies and with the interests in local knowledge systems found throughout the GED literatures. Through such analyses,-even of late Twentieth Century physics-Northern science studies have shown that no element of science is immune from cultural shaping, and every element of science has epistemological consequences (Forman 1987; Pickering 1992).

For a second focus, consider recent arguments against the unity of science thesis the form in which Enlightenment assumptions about the universality of modern scientific claims coalesced in the late nineteenth and early twentieth centuries and remain powerful today (see Galison and Stump 1996; Harding 1998a). The unity of science thesis held that there was one world, one truth about it, and one and only one science capable of capturing that truth. Obscured in the scientists' and philosophers' formulations of the argument was a fourth assumption: that there was one "class" or group of humans capable of articulating that science and thus recognizing that truth-scientific experts.12 Leaving aside questions about the usefulness of the "one world" hypothesis (but see Dupre 1993), the other three assumptions have been firmly undermined by post-Kuhnian science studies. This is not to say that these positions are uncontroversial, but rather that many historians, sociologists, ethnographers and philosophers of science agree that the unity assumptions have outlived their usefulness, at least in their stronger initial formulations and in the ways they are articulated today. The "one truth" hypothesis requires various forms of reductionism that are unrealistic; even as an ideal, this claim blocks knowledge. The "one science" hypothesis is meaningless even in modern sciences, in light of the valuable proliferation there of specialized research fields, their distinct methods, and distinct representational resources. Modem sciences are themselves many local knowledge systems linked in pragmatic but not necessarily perfectly coherent ways, from this perspective.

The challenge then becomes not how to reduce this multiplicity to one unity through some kind of fancy discursive footwork, but rather to understand how such local knowledges do travel from one culture to another, and how elements from different contexts of production are linked and reconstructed in ways that help to produce new knowledge. How can practitioners from fields with conflicting definitions of key terms come to work together effectively?'3 What is gained and what lost when knowledge is detached from one cultural system and inserted into another at a distant time or place, as has occurred recently, for example, in the case of acupuncture's move into modem biomedicine? How is information that must be passed on over generations standardized in cultures of only oral literacy (Watson-Verran and Turnbull 1995)?

The universality ideal also appears scientifically and epistemologically dysfunctional, for many reasons. It legitimates appeal to monolithic science to support otherwise inadequately supported individual scientific claims, and it legitimates resistance to valuable criticisms that are inconsistent with prevailing views. lt inappropriately devalues cognitive diversity. It obscures the inevitable limitations of any one science. It also promotes systematic ignorance in the social sciences and other fields that model themselves on philosophies of the natural sciences. Moreover, the universality ideal is politically costly in three ways. It devalues and destroys knowledge traditions that are crucial to the survival of other cultures. It elevates a model of the admirably human that is defined in terms of its opposition to and distance from the womanly, non-European, and economically vulnerable. And it elevates cognitive authoritarianism and problematic "religious" ideals to the status of the highest human ideals; it does this through its monovocality and xenophobia; its hierarchical social structures with their elite group of experts who have the status of "chosen people," and its formal and informal ways of protecting from public scrutiny the complete processes of sorting belief.14

Finally, this school of science studies, too, shows that the standards of modem science have historically changed over time and are always rhetorically constituted and deployed. What counts as good method, as a proof in mathematics, as objectivity and rationality, not to mention as "real" or material, varies from era to era and often from field to field within the sciences. Epistemological and ontological standards of the sciences, too, have an integrity with their historical era (Proctor 1991; Schuster and Yeo 1986; Shapin 1994; Shapin and Schaffer 1985).

Thus post-Kuhnian histories and philosophies of modem science are thematically linked to the others in their critiques of Enlightement philosophies of science, and provide valuable possibilities for the coalition work necessary to stabilize post-Enlightenment philosophies of science. What will such philosophies look like?

POST-ENLIGHTENMENT PHILOSOPHIES OF SCIENCES: NEW QUESTIONS We are now in a position to summarize some of the new philosophical themes emerging from GED analyses that appear also in postcolonial and post-Kuhnian studies.

After the Universality ideal: Local Knowledge Systems

The Enlightenment philosophies were preoccupied with eliminating the local in scientific processes so as to obtain transcultural, universally valid knowledge claims. Elements of many bodies of systematic scientific knowledge (and of systematic scientific ignorance) find homes in later knowledge configurations, though the models of the universe in which they lodge, and the relations they claim between observed phenomena can differ vastly in successive systems. The new philosophies ask how and what kind of local knowledge "travels," and-because cultures are tool boxes as well as prison houses for the growth of knowledge-what is lost and what is gained when it does. They are interested in globalizing processes, but not in achieving universality.

Smart Knowers and Imperfect Knowledge Systems

Such conditions of actual and ideal knowledge production require a model (or rather many such) of knowers and knowledge systems that is different from the familiar one of dumb knowers who must struggle to learn the one correct knowledge system. These philosophies explore other possible models. One, drawn from the realities of both everyday life and contemporary scientific practice, posits smart knowers and imperfect knowledge systems. Here, knowers-scientists or citizens-always have only imperfect knowledge systems with which they must make daily decisions as well as life-ordeath ones, though most knowers these days always have more than one such system available.

Consider, for example, the daily health maintenance practices of middle classes in the metropoles. Here individuals commonly use several conflicting knowledge systems. In the United States we use revised vitamin, acupuncture, chiropractic, dietary, exercise, and meditation therapies, not to mention Grandma's home remedies-which modem biomedicine, until the last few years, claimed were of little or no value. But we also use modem biomedicine. Another version of this situation occurs for people in non-Western cultures who have access both to modern biomedicine and other modem Northern sciences and to indigenous health and other knowledge systems (see, for example Bass 1990). All of us have to be very clever about which knowledge system we use and when; this can be a life-or-death matter. The point is that cognitive diversity is an important scientific value.15

Resisting Relativism and Idealism

In almost all the writings in these diverse literatures, the rejection of Enlightenment absolutism and excessive materialism is specifically not permitted to be a reason to adopt cognitive relativism or idealism. Different cultures actually do have different scientific and epistemological standards, and this is merely an obvious historical or sociological claim-it is historically or sociologically relative. But not all knowledge systems are equally powerful at grasping the diverse aspects of nature's heterogeneous order. Many people lead short and unhappy lives by following the recommendations of local knowledge systems-including those of our own "local knowledge system," modem science.

Consider our own system's toleration, until recently, of smoking and inadequate vitamins, not to mention industrially-caused environmental toxins. Some sciences are better if you want to get to the moon; others if you want to maintain sustainable environments. So instead of pursuing issues about the virtues of absolutism or realism, these accounts are interested in how to articulate such practical standards of belief sorting. And rejecting the idea that our glassy minds can perfectly mirror a reality that is "out there" in the nonsocial universe similarly never forces these accounts to an idealism, in which only human ideas and the social are real. Instead, the dichotomy between material and ideal, like absolute versus relative cognitive judgments, has itself become the object of historical scrutiny: what social and intellectual conditions made such problematic conceptual frameworks look reasonable?

Nature: Social and Emergent

For the Enlightenment philosophies, nature was, in principle, readable by science, and the questions were about how to accomplish such perfect readings. In the literatures examined here, there is no possibility of one perfect scientific reading of nature's order. In part, this is because of the sometimes fruitful, sometimes knowledge-obstructing ways that humans work nature into an object of observation through cultural interests, discourses, and ways of organizing the production of knowledge. This nature that scientists observe is, in such respects, always also social. (Not "nothing but" social; only always also social.) Here the questions shift to how what counts as nature and what counts as the social are coconstituted in societies that favor such a dichotomy, and more generally, how human material and symbolic practices change nature's order. How should sciences and their philosophies, sciences' order and philosophies' order be conceptualized as also "in nature," requiring explanation alongside the accounts they provide of sciences' and nature's order? What resources from philosophies of social sciences could be useful in such explanations?

Another cause of nature's unreadability is its emergent character. Nature is constantly producing new phenomena, such as ozone holes, and new categories of diseases that are unpredictable. How should this aspect of nature be characterized in the new philosophies?

Internalized Democratic Ethics and Politics in Sciences

With a multiplicity of culturally local sciences, presumably a multiplicity of culturally local ethics and political philosophies will be integrated into these sciences. Enlightenment notions of rationality locate ethics and politics outside the borders of rationality. In recent years, feminist theorists, among others, have developed post-Enlightenment theories of rationality, in which moral decisions and the emotions also have their rationality. The postcolonial and GED analyses clearly are exploring appropriate ways to internalize democratic ethical and political ideals within scientific rationality, because such a rationality bereft of such accountability is incapable of adequate self-regulation. What "democratic" will mean, how such a standard will function in particular contexts-these are difficult but important questions the GED and postcolonial accounts address. One possible democratic ethos to internalize in science-one that emerges from several of the kinds of accounts here-would be that those who bear the consequences of scientific and technological decisions should have a proportionate share in making them. In different historical and cultural contexts, different kinds of institutions and practices would best facilitate such an ethos: small, face-to-face communities and large bureaucratic ones need different kinds of democratic institutions and practices.

Whose Philosophies of Science?

And what of philosophies of science, such as those with common themes that are emerging from so many diverse post-World War II science and technology studies? We have noted that they, too, can usefully be conceptualized as "part of nature" and subjected to analyses alongside the other social and natural elements of our surroundings. We can also see that they must remain permanently unfinished and continually regenerated through collective processes achieved through fruitful coalitions and respectful dialogues.