Boundary Objects and Beyond: Working with Leigh Star

Infrastructures Series

edited by Geoffrey C. Bowker and Paul N. Edwards

Paul N. Edwards, A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming

Lawrence M. Busch, Standards: Recipes for Reality

Lisa Gitelman, ed., Raw Data Is an Oxymoron

Finn Brunton, Spam: A Shadow History of the Internet

Nil Disco and Eda Kranakis, eds., Cosmopolitan Commons: Sharing Resources and Risks across Borders

Casper Bruun Jensen and Brit Ross Winthereik, Monitoring Movements in Development Aid: Recursive Partnerships and Infrastructures

James Leach and Lee Wilson, eds., Subversion, Conversion, Development: Cross-Cultural Knowledge Exchange and the Politics of Design

Olga Kuchinskaya, The Politics of Invisibility: Public Knowledge about Radiation Health Effects after Chernobyl

Ashley Carse, Beyond the Big Ditch: Politics, Ecology, and Infrastructure at the Panama Canal

Alexander Klose, translated by Charles Marcrum II, The Container Principle: How a Box Changes the Way We Think

Eric T. Meyer and Ralph Schroeder, Knowledge Machines: Digital Transformations of the Sciences and Humanities

Geoffrey C. Bowker, Stefan Timmermans, Adele E. Clarke, and Ellen Balka, eds., Boundary Objects and Beyond: Working with Leigh Star

Boundary Objects and Beyond

Working with Leigh Star

Edited by Geoffrey C. Bowker, Stefan Timmermans, Adele E. Clarke, and Ellen Balka

The MIT Press

Cambridge, Massachusetts

London, England

© 2015 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher.

This book was set in ITC Stone Serif Std by Toppan Best-set Premedia Limited. Printed and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data is available.

ISBN: 978-0-262-02974-2 (hc); 978-0-262-52808-5 (pb)

eISBN: 978-0-262-33100-5

10 9 8 7 6 5 4 3 2 1

Based on the Conference: A Celebration of Susan Leigh Star: Her Work and Intellectual Legacy, University of California, San Francisco (UCSF), September 2011, http://www.sscnet.ucla.edu/soc/LeighStar/. Organized by Stefan Timmermans (UCLA), Geoffrey C. Bowker (University of Pittsburgh) and Adele E. Clarke (UCSF). Sponsored by Science, Technology, and Society Program, Division of Social and Economic Sciences, National Science Foundation; Department of Sociology, UCLA; Department of Social and Behavioral Sciences, UCSF.

d_r0

Table of Contents

Series page

Title page

Copyright page

Prologue

Introduction: Working with Leigh Star

I Ecologies of Knowledge

1 Revisiting Ecologies of Knowledge: Work and Politics in Science and Technology

2 Ecological Thinking, Material Spirituality, and the Poetics of Infrastructure

3 Don’t Go All the Way: Revisiting Misplaced Concretism

4 Anticipation Work: Abduction, Simplification, Hope

5 Living Grounded Theory: Cognitive and Emotional Forms of Pragmatism

6 Misplaced Concretism and Concrete Situations: Feminism, Method, and Information Technology

II Boundary Objects

7 Institutional Ecology, Translations, and Boundary Objects: Amateurs and Professionals in Berkeley’s Museum of Vertebrate Zoology, 1907–1939

8 Sharing Spaces, Crossing Boundaries

9 So Boundary as Not to Be an Object at All

10 The Concept of Boundary Objects and the Reshaping of Research in Management and Organization Studies

11 Leigh Star and the Appearance of The Structure of Ill-Structured Solutions

12 The Structure of Ill-Structured Solutions: Boundary Objects and Heterogeneous Distributed Problem Solving

III Marginalities and Suffering

13 Power, Technology, and the Phenomenology of Conventions: On Being Allergic to Onions

14 Anatomy Is Frozen Physiology, Or How I Learned to See the Process That Is Everywhere

15 Categorizing Life and Death: The Denial of Civilians in U.S. Robot Wars

16 Infrastructures for Remembering

17 Triangulation from the Margins

18 Reflections on the Visibility and Invisibility of Work

19 Layers of Silence, Arenas of Voice: The Ecology of Visible and Invisible Work

IV Infrastructure

20 Steps toward an Ecology of Infrastructure: Design and Access for Large Information Spaces

21 Mapping the Body across Diverse Information Systems: Shadow Bodies and How They Make Us Human

22 Thundering Silence: On Death, Fear, Science

23 Those Who Are Not Served? Exploring Exclusions and Silences in Transport Infrastructures

24 The Ethnography of Infrastructure

Envoi: When Shadows Become Complex: Weaving the Ŋanmarra

Afterword: On the Distributedness of Leigh

Obituaries, Memorial Events, and Selected Publications of Susan Leigh Star

Contributors

Index

List of Illustrations

Figure 3.1 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.2 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.3 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.4 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.5 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.6 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 3.7 Still from Misplaced Concretism/Film for Leigh StarSource: Wakeford and Keal 2010.

Figure 4.1 Abductive analysis Source: Based loosely on figure 27.2 in Strübing 2007.

Figure 6.1 Diagram and description of a turkey baster

Figure 7.1 Translation in actor-network theory

Figure 7.2 Translations and boundary objects

Figure 10.1 Walker Percy’s triadic semiotic event

Figure 10.2 Percy’s two-person semiotic model

Figure 10.3 Boland’s image for a thought experiment

Figure 10.4 Boland’s version of relationships between standards and residual categories

Figure 12.1 Boundary object: repositories

Figure 12.2 Boundary object: platonic object

Figure 12.3 Boundary object: terrain with coincident boundaries

Figure 12.4 Boundary object: forms/labels

Figure 13.1 Three power continuums

Figure 15.1 Photo 2009.12.18 House Destroyed, Ob50/Noor Behram (Bureau of Investigative Journalism 2011)

Figure 15.2 Photo 2009.08.21 Syed Wali Shah, aged 7 Ob32/Noor Behram (Bureau of Investigative Journalism 2011)

Figure 15.3 Photo Shoes (photo: Noor Behram)

Figure 18.1 Workers at a Foxconn factory in southern Guandong Province. Foxconn employs more than 800,000 workers in China. Source: New York Times, June 2, 2010

Figure 18.2 Workers on strike, Honda Lock factory, Zhongshan, China, June 2010 Source: New York Times, June 16, 2010

Prologue

Geoffrey C. Bowker

I write this with a heavy heart and tremulous hands; with joy and pleasure and love . . . sharing as ever Leigh’s delight in contradiction.

I was graced to weave my life together with hers. She had a soft voice, an impish smile, and a lovely laugh. In pain—and pain was ever present, waxing and waning—her voice would drop and in her eyes I could sense her suffering. In strength—and always was there strength—the beauty of her light shone through. Together we drew down the moon, conjured classifications and discovered poetry abounding.

This volume brings together friends and colleagues who were touched by and who touched her. Not a self-contained set, but offering a flavor of the way she was in the world: intimately, disconcertingly personal; changing lives; generating ideas; thinking the unimaginable with fierce will and phrasing it with delicate touch.

When we made our final move, she had been going through a long and difficult stretch. I am so happy that her last months were full of optimism and renewal. Always a good gauge of her spirit was whether she was writing poetry (or rather, whether she was writing what is formally called poetry): she was, and it was wild and wonderful and copious.

Her next project, which we had played with for years, was to be a work on the poetics of infrastructure. She talked of the orphans of infrastructure. People sequestered away from the ways and wires of the world on an old leper colony island in Hawaii; the Dionn quintuplets whose every evacuation was recorded and whose every moment of idle play was annotated. People who did not fit the units of social programs and software and led strangely beautiful lives. She talked of infrastructure as a thing of beauty . . .

This volume is not a closure, but a living work through which many who do not know her can come to be touched by her magic and those who did can once again feel the beauty of her spirit.

If we collectively have succeeded, then, to quote Leigh, we will speak Her name clearly into your eyes.

Introduction: Working with Leigh Star

Stefan Timmermans

Once the process gets going it keeps on going, given constant interactions with other people and all kinds of humans and nonhumans in the world. I don’t know enough about death to know whether or in exactly what forms it might keep going afterwards, except that the ongoing actions we leave embedded in the world constitute one such action.

(Star 1995a, 20)

Susan Leigh Star (Leigh) became one of the most influential science studies intellectuals of the last decades. Along with other scholars, her work shifted the research agenda from bibliometric and Mertonian preoccupations with the social organization of science (i.e., its reward systems, organization of scientific communities, invisible colleges, dominant values, etc.) to a focus on the production of scientific facts. Summarizing her generation’s approach, Star wrote, Among our common interests and beliefs was the necessity of ‘opening up the black box’ in order to demystify science and technology, that is, to analyze the process of production as well as the product. The methodological directives here for us were familiar: understand the language and meanings of your respondents, link them with institutional patterns and commitments, and, as Everett Hughes said, remember that ‘it could have been otherwise’ (Star 1988, 198). This awareness that it could have been otherwise turned the assumption that science reflected nature or facts into a topic of sociological inquiry.

Drawing from pragmatism and the Chicago school’s sociology of work, Star’s scholarship highlighted the messy practices of discovering science. Where others conducted ethnographies of the work of laboratory scientists’ manipulating tools and data to construct epistemic insights, Star quickly became drawn to the work behind the work: the countless, taken-for-granted and often dismissed practices of assistants, technicians, and students that made scientific breakthroughs possible. One of her goals was to restore agency to scientific work by examining who did the actual work and who received credit for that work. Digging ever deeper into the conditions that make science possible, Star foregrounded the infrastructure of classifications, technologies, paperwork, and regulations that constrained and constituted scientific work. She aimed for a radical symmetry in her work: not only evaluating the winners and losers of science evenhandedly but examining the lives of those caught up against their will in the new globalizing technoscientific regimes. Within the post-Merton sociology of science, her work thus opened up the field of science studies to what others took for granted.

Star was greatly inspired by Everett Hughes’s approach to the sociology of work, as passed on by her dissertation advisor Anselm Strauss and intellectual muse Howard Becker—themselves Hughes students. Looking at science as work, a job to be done one task at the time by someone in a particular organizational form demystified the special character of science and offered a vantage point of the practice of doing (rather than talking about) science. For Star, scientists were neither heroes nor villains, but people doing things together—the things being an activity they called science. Looking at science as work opened up the laboratory to bread-and-butter sociological questions such as how do people come to believe what they believe about nature? How do scientists find a common language across disciplinary boundaries? How does science contribute to the stratification of society? The emphasis on work drew Star to examine the activities that are ignored, invisible, and deleted. Her methodological mantra was to restore the work: following who did what, when, where, how, and with what kind of consequences.

From the Chicago school of sociology, Star was stimulated by the strong focus on conceptualizing, developing sensitizing notions (Blumer 1954), in the grounded theory tradition (Glaser and Strauss 1967). An accomplished poet and wordsmith in her own right, Star was drawn to the intellectual community of grounded theorists and pragmatists and the creativity (stretch your imagination [Star 1991b]) of this theory/methods package (Clarke and Star 2008). Grounded theory did not only stimulate her intellectually but also drew attention to the joys and, occasionally, frustrations of doing scientific research, which considering that her topic was scientific work inevitably drew her to reflect on the commonalities between her own work and the work of scientists (Star 2007). She taught workshops and courses on qualitative methods, showing how well-developed concepts can illuminate hidden dimensions in a research project. Indeed, for Star, pragmatism infused grounded theory with an epistemological compass: How does one create a life-work and remain open; open to the data, open to being wrong, to redoing one’s own work, actively to seek out new views and mistakes? For me, that has come through the privilege of teaching grounded theory, and of collaborating with people who like to work this way. That is, to embrace a continuous, embedded, imbricated, multiple, constantly compared way of making sense of myself (Star 2007, 90).

With a strong pragmatism at her core, Star was further influenced by feminism, race theories, ecological thinking, symbolic interactionism, actor-network theory, ethnomethodology, linguistics, activity theory, metaphysics, theology, cognitive science, phenomenological psychology, distributed artificial intelligence, and anyone who produced exciting intellectual insights.

Star situated science and technologies within ecologies of knowledge (Rosenberg 1979). As such, she aimed to understand the systemic properties of science by analogy with an ecosystem, and equally important, all the components that constitute the system (Star 1995a, 2). Rather than taking the perspective of the university administrator or award-winning professional scientist, Star examined the interrelationships between different layers of a system within an entire ecosphere. Thus, a study of the design of computer chips almost immediately jumped to an analysis of institutions, organization charts, the material characteristics of chips, regulations, and gurus in the field (Star 1995b). And she conceptualized a study of information systems as composed of people and things, in ecological relationships, with representations and signals and fluid ways of working.

Star highlighted how the ecological knowledge perspective eschewed reductionist analyses: by ecological, I mean refusing social/natural or social/technical dichotomies and inventing systematic and dialectical units of analysis (Star 1988). Rather than dualism, Star, following Dewey, saw phenomena as continuous (Star 1991b, 277). She encouraged the simultaneous analysis of human and material worlds: The advantages of such an analysis are that the increased heterogeneity accounts for more of the phenomena observed; one does not draw an arbitrary line between organism and environment, one can empirically ‘track’ lines of action without stopping at species, mechanical or linguistic boundaries, and especially without invoking a reified conception of society (Star 1995a, 13). Unlike Latour, Star focused on nonhuman agency in order to highlight how social life is recalibrated and restratified. Her goal was not to democratize the human-nonhuman divide but to analyze the powers of the nonhuman in reshaping a human world.

From pragmatist philosophy, Star took to heart the difficult notion that understanding is based on consequences not antecedents. Indeed, her advisor Anselm Strauss controversially noted that gender and race cannot be presumed a priori to be necessarily relevant but need to earn their theoretical relevance (but see Clarke 2008). Yet, schooled as a feminist, Star gravitated to examining the consequences of information technologies for those at the margins, those who are both outsiders and insiders. Studying the outliers and forgotten was theoretically important because the voices of those suffering from abuses of technological power are among the most powerful analytically (Star 1991a, 30).

One of Star’s goals was to present the process of building big science from the viewpoints of the excluded rather than from those of management or great scientists. Every enrollment entails both a failure to enroll and a destruction of the world of the non-enrolled (Star 1991a, 49). She encouraged her students to ask the question: Cui bono? Who benefits? Her point was not to do checklist research in which social scientists check off the usual suspects (Shim 2002) of race, class, and gender. Rather, she asked hard questions about the marginalizing as well as liberating powers of technosciences and their stratifying consequences. Star embraced a theory of intersectionality in which marginalities are not pre-given but the systemic consequences of new sociotechnical arrangements that create new—and fluid—forms of exclusion. Every standpoint, every self, every interest counts but they also come with overhead (Bowker and Star 1999, 309). Sometimes these overhead costs were deliberately sought out as acts of subversion, at other times the cost was the collateral damage of facing new forms of science and technology. Her project was to reveal the subversion, the collateral damage and suffering as well as the valuable accomplishments of new information regimens.

Restoring the intended and unintended fallout of new information systems and data structures was analytically important because such things are systematically deleted from official records. Hence, Star’s interest in the sociology of the invisible and articulation work: As the work of negotiating, resource management, and many other processes are deleted from representations, one group’s interests begin to take precedence over another’s, subverting the fundamental pluralism of human interaction. . . . The more the work is rendered deleted and the more invisibles are invoked as explanations, the more suffering there is (Star 1991b, 278–279).

Articulation work refers to keeping things on track through dispersed bits of local knowledge (Strauss 1988). It is the kind of work done by others that allows an executive or a famous scientist to cross multiple worlds. This is the work of nurses, secretaries, research assistants, copy editors, and repair people.

Among the most celebrated of Leigh’s concepts was the notion of boundary object (Star and Griesemer 1989; Star 1989; Star 2010). This concept captures the possibility of cooperative scientific work in the absence of consensus. Cooperation does not always follow from a preexisting consensus but can be achieved with objects flowing through various communities of practice or social worlds. Star used object loosely in a Meadian sense as something that people act toward and with; it may be a thing but also, for example, a theory. A boundary object is something that has different and quite specific meanings in intersecting communities but also has a common meaning to facilitate cooperation across communities. Boundary objects are objects which are both plastic enough to adapt to local needs and the constraints of several parties employing them, yet robust enough to maintain a common identity across sites (Star and Griesemer 1989, 393). Boundary objects thus satisfy the informational requirements of various communities while still having distinct meanings in each one of them. The boundary part of boundary object referred not necessarily to the periphery of communities but to overlapping areas (as in a Venn diagram) between different communities. Considering that boundary objects emerge through work processes, Star noted that they are themselves in flux. They can easily lose some of their distinct meanings for different communities when they become standardized. Such standardizations, however, may in turn produce more residual categories that defy standardization and stimulate the emergence of new boundary objects (Star 2010).

Star developed this concept in a study of Berkeley’s Museum of Vertebrate Zoology. This museum required an extensive collection of natural specimens in order to build up its reputation, and professional biologists were trying to capture California’s biodiversity before it would disappear. They enrolled help from amateur collectors who sought legitimacy through their collecting efforts. Other interested parties were trappers, farmers, and traders who often caught rare animals. A key boundary object that facilitated autonomy and communication across these social worlds became the animal specimens. An economy of bartering, trading, and buying sprung up in which each community kept autonomy over the collection of animal specimens, but archival and research science was made possible with the stipulation of minimum standards about the annotation of habitat, conditions, time and geographical location of capture and preservation, and the specimens. Thus the specimens had different identities in each social world but gained a common identity around the scientific documentation of California wildlife.

This boundary object was further related to other boundary objects that regulated communication and relationships between communities. These included standardized information forms for the collection of specimens, repositories of specimens without specification of the ultimate goal of the repository, guidelines that did not capture the granularity to execute tasks but were flexibly adaptable to local needs, and maps of California that gave different social worlds a set of common geographical boundaries but could be filled in differently by each group to suit its needs and goals. The focus on boundary objects allowed Star to examine how objects achieve coherence across social worlds while still having distinct identities within specific communities. Boundary objects reflect neither a consensus nor a top-down imposition of requirements on others: Rather, boundary objects act as anchors or bridges, however temporary (Star and Griesemer 1989, 414).

The idea of boundary objects has been extremely influential, giving science, technology, and society (STS, also known as science and technology studies) scholars an entire new approach for studying how communities interact and build science. The perspective of boundary objects offers an engaged counterpoint to the emphasis on translating other people’s priorities into scientific ones (translation and intéressement), instead highlighting forms of collaboration that preserve the uniqueness of each community of practice (Clarke and Star 2008).

In contrast, working in computer and information science departments and with computer scientists, Star was struck by the tendency of some computer scientists to create formalisms of mathematical models of the world they aimed to capture while others embraced a more empirical approach—trying to collect every data nuance. Social science, especially the qualitative social science she practiced, was often lumped dismissively with an empirical approach and contrasted with the presumed elegance and intellectual superiority of models. Based on the frictions that her presence generated in computer science circles and drawn by both the intellectual and ethical implications of the formalism-specificity issue, Star developed a life-long interest in the trade-offs between formalisms and the messiness of data, work, and life.

She understood that formalization and its related processes of simplification, standardization, universalization, rationalization, objectification, and representation were required to allow technologies to work outside the context where they were created. Yet, formalisms and the technologies that incorporated them often fail to work, requiring what Gasser (1986) called workarounds or ad hoc tinkering around formal demands. Rather than seeing this as redemption for the empirical approach, as a pragmatist Star asked under what conditions do questions of formalism arise: How can something be simultaneously concrete and abstract? To her question, we can add: Who puts together the database? What is included and what is left out? It was not because people do not do what they are expected to do, or that databases or information technologies have no effect. Again drawing on pragmatist thinking, she saw universalisms as nothing more than standing agreements across communities, not a priori facts. She, along with colleagues in the computer-supported work community, advocated for joining empirical research with formalism, noting that formalism is inevitably locally and temporally situated, distributed, and contingent, and needs to be made to work (Star 1992).

The theme of standards and classification has been present in Star’s work from its early beginnings (Star 1983), but reaches its pinnacle in her magnum opus Sorting Things Out, coauthored with Geoffrey Bowker (Bowker and Star 1999). In it, Star observes that classification systems and standards tend to blend in with the background infrastructure. Yet, these seemingly boring tools embed deeply social and ethical histories of conflict and compromise in a technical nomenclature that then provides the often-invisible infrastructure for contemporary living.

Classifying is a memory practice to both hold on to certain characteristics and send other elements into oblivion. Star gravitated to an examination of how classification systems cause troubles for individuals, coining the concept of torque to notate situations when a formal classification system is mismatched with an individual’s biographical trajectory, membership, or location. Starting from the observation that one person’s standard is another’s confusion and mess, Bowker and Star articulated the moral dimension of a study of classifications:

The moral questions arise when the categories of the powerful become the taken for granted; when policy decisions are layered into inaccessible technological structures; when one group’s invisibility comes at the expense of another’s suffering. . . . We need to consistently explore what is left dark by our current classifications (other categories) and design classification systems that do not foreclose on rearrangements suggested by new forms of social and natural knowledge. (Bowker and Star 1999, 320–321)

Here we vividly see Star’s overarching commitment to social justice through better understanding of the workings of infrastructure.

Political categories become naturalized through standards and classification systems. Bowker and Star link the International Classification of Disease (ICD) to the birth of the welfare state, and note generally how different kinds of political regimes, including the former apartheid regimen in South Africa, depended on various classification systems. With different information requirements and different infrastructures to gather data, each one of the ICD categories constitutes a frozen policy. It is frozen because the category is reified when introduced in a classification system, and it is policy because it nevertheless feeds relentlessly into further political developments. The ICD data, for example, allowed competing groups of medical specialists, public health officials, and laissez-faire economists to take credit for Britain’s decline in mortality rates in the nineteenth century. Bowker and Star significantly note that in the process of making records and producing knowledge, organizations use classification systems to selectively forget the past.

Star and some colleagues joked that their work reads like a manifesto for the Society for the Study of Boring Things (Lampland and Star 2009, preface). What is a dull classification, an annoying information infrastructure, or a set of easy to ignore standards for some, becomes the analytic object for others. For Star, studying these forms of social life meant restoring the work and the political/ethical/social struggles that went into the creation of the formal. She was not against standards or protocols but recognized both the danger of leaving those tools unexamined and the violence they could inflict on those who did not neatly align with them.

References

Blumer, Herbert. 1954. What Is Wrong with Social Theory? American Sociological Review 18:3–10.

Bowker, Geoffrey, and Susan Leigh Star. 1999. Sorting Things Out: Classification and Its Consequences. Cambridge, MA: MIT Press.

Clarke, Adele. 2008. Sex/Gender and Race/Ethnicity in the Legacy of Anselm Strauss. Studies in Symbolic Interaction 32:159–174.

Clarke, Adele, and Susan Leigh Star. 2008. Social Worlds/Arenas as a Theory-Methods Package. In Handbook of Science and Technology, ed. E. J. Hackett, O. Amsterdamska, M. Lynch, and J. Wacjman, 113–139. Cambridge, MA: MIT Press.

Gasser, Les. 1986. The Integration of Computing and Routine Work. ACM Transactions on Office Information Systems 4:205–225.

Glaser, Barney G., and Anselm L. Strauss. 1967. The Discovery of Grounded Theory. New York: Aldine Publishing Company.

Lampland, Martha, and Susan Leigh Star. 2009. Standards and Their Stories: How Quantifying, Classifying, and Formalizing Practices Shape Everyday Life. Ithaca, NY: Cornell University Press.

Rosenberg, Charles E. 1979. Towards an Ecology of Knowledge: On Discipline, Contexts and History. In The Organization of Knowledge in Modern America, ed. A. Oleson and J. Voss, 440–455. Baltimore: Johns Hopkins University Press.

Shim, Janet K. 2002. Understanding the Routinized Inclusion of Race, Socioeconomic Status and Sex in Epidemiology: The Utility of Concepts from Technoscience Studies. Sociology of Health & Illness 24:129–150.

Star, Susan Leigh. 1983. Simplification in Scientific Research: An Example from Neuroscience Research. Social Studies of Science 13:208–226.

Star, Susan Leigh. 1988. Introduction: The Sociology of Science and Technology. Social Problems 35:197–205.

Star, Susan Leigh. 1989. The Structure of Ill-Structured Solutions: Boundary Objects and Heterogeneous Distributed Problem Solving. In Readings in Distributed Artificial Intelligence, ed. M. Huhns and L. Gasser, 37–53. Menlo Park, CA: Kaufman.

Star, Susan Leigh. 1991a. Power, Technologies, and the Phenomenology of Conventions: On Being Allergic to Onions. In A Sociology of Monsters: Essays on Power, Technology and Domination, ed. J. Law, 26–56. London: Routledge.

Star, Susan Leigh. 1991b. The Sociology of the Invisible: The Primacy of Work in the Writings of Anselm Strauss. In Social Organization and Social Process, ed. D. R. Maines, 265–284. New York: Aldine de Gruyter.

Star, Susan Leigh. 1992. The Trojan Door: Organizations, Work, and the ‘Open Black Box.’ Systems Practice 5:395–410.

Star, Susan Leigh. 1995a. Introduction. In Ecologies of Knowledge, ed. S. L. Star, 1–38. Albany: State University of New York Press.

Star, Susan Leigh. 1995b. The Politics of Formal Representations: Wizards, Gurus, and Organizational Complexity. In Ecologies of Knowledge: Work and Politics in Science and Technology, ed. S. L. Star, 88–118. Albany: State University of New York Press.

Star, Susan Leigh. 2007. Living Grounded Theory: Cognitive and Emotional Forms of Pragmatism. In Handbook of Grounded Theory, ed. A. Bryant and K. Charmaz, 75–93. Los Angeles: Sage Publications.

Star, Susan Leigh. 2010. This Is Not a Boundary Object: Reflections on the Origin of a Concept. Science, Technology & Human Values 35:601–617.

Star, Susan Leigh, and James R. Griesemer. 1989. Institutional Ecology, ‘Translations’ and Boundary Objects: Amateurs and Professionals in Berkeley’s Museum of Vertebrate Zoology, 1907–39. Social Studies of Science 19:387–420. [See also chapter 7, this volume.]

Strauss, Anselm L. 1988. The Articulation of Project Work: An Organizational Process. Sociological Quarterly 29:163–178.

I Ecologies of Knowledge

1

Revisiting Ecologies of Knowledge: Work and Politics in Science and Technology

Susan Leigh Star

Editors’ Note: This chapter originally appeared as the introduction to a volume Leigh Star edited titled Ecologies of Knowledge: Work and Politics in Science and Technology (1995a). It offers an important and still useful overview of both early science and technology studies and her own innovative perspectives on the sociology of knowledge and its production processes, questions at the heart of science and technology studies. The original chapter referred extensively to papers in that volume in ways that would be confusing here. We therefore included citations to clarify where these works appear as [1995].

In 1967 Howard Becker wrote an article that became a clarion call to sociology, his presidential address to the Society for the Study of Social Problems. It was entitled Whose Side Are We On? and reminded sociologists that pretensions to value neutrality were themselves value laden. He argued that we must choose to recognize that all perceptions are located in a hierarchy of credibility. In other words, people consider the source of any statement or perception, and discount those produced by lower-status people.

Whose side are we on in social studies of science and technology? What hierarchies of credibility are we tacitly or explicitly assigning? And what language can we invent to investigate these questions honestly?

We could do well to borrow from Patricia Hill Collins’s (1986) essay on black feminist thought and its contributions to the structure of sociological knowledge. She argues that African American women’s radical explorations of the meaning of self-definition and valuation, the interlocking nature of oppression, and the importance of redefining culture constitute a challenge to sociology’s basic beliefs about itself. The challenge takes the form of asking sociology to learn from the outsider within—the double glasses of insider and outsider, articulating the tension of both being a sociologist and being excluded by its frame of reference.

The papers in Ecologies of Knowledge: Work and Politics in Science and Technology (Star 1995a) are all attempts to frame the question of whose side we are on by examining science as a radically contextual, problematic venture with a very complicated social mandate, if any. Our purpose here is more than polemics; rather than valorizing or denigrating science as a monolith, we are taking an ecological view of work and politics. And we, too, are outsiders within, as Woolgar’s [1995] essay argues—both strangers and intimates in the world of science. Our work challenges the moral order of science and technology making—and in turn places us in a complex, often tense moral position.

Ecologies of knowledge¹ here means trying to understand the systemic properties of science by analogy with an ecosystem, and equally important, all the components that constitute the system. This is not a functional (or functionalist) approach, with a closed-system organic metaphor at its core. As Sal Restivo (1988) notes in his description of science as a social problem, we want to approach science as a set of linked interdependencies inseparable from personal troubles, public issues and social change agendas, not a social structure with one or more dysfunctional parts. Science and technology become monsters when they are exiled from these sorts of questions (Law 1991; Haraway 1992; Star 1991a; Clarke 1995a), just as other symbolic monsters have been borne from the exile of women’s strength from the collective conscience, or in the demonizing of people of color. In Michele Wallace’s words: "The absence of black images in the reflection of the social mirror, which such programmatic texts (from Dick and Jane, to Disney movies, to The Weekly Reader) invariably contract, could and did produce the void and the dread of racial questions" (1990). In our exorcism, we simply want to see scientists and technologists as ordinary, as citizens, neither villains nor heroes.

Each of the authors who contributed to Ecologies of Knowledge: Work and Politics in Science and Technology (Star 1995) calls in different ways for an ecological analysis, including a restoration of the exiled aspects of science. Thus by ecological we mean refusing social/natural or social/technical dichotomies and inventing systematic and dialectical units of analysis. I think this reflects the dissatisfaction with conventional ways of approaching organizational scale and units of analysis, a dissatisfaction brought on no doubt in part by our respondents (scientists and technologists), who themselves are continually plagued by these questions. Croissant and Restivo [1995] examine a large-scale set of relationships, as between science and all other institutions and social arrangements. Kling and Iacono [1995] put the context back into analyses of technological innovation, noting that only such an approach can overcome simpleminded technological determinism or technocracy. Star [1995b] makes a similar point for the class of artifacts called formalisms, or formal representations, and their relationship to organizational complexity. Fujimura [1995] looks at the ecology of the workplace in a Hughesian sense, and thus goes beyond the simple adoption of new technologies as a determining factor in scientific social change. Law and Callon [1995], Woolgar [1995], and Latour [1995] each break the traditional boundaries of what can be included in an analysis of technology and social organization, recommending a broader, more democratic kind of analysis that is both moral and deeply ecological. In this, Law and Callon [1995] call attention to the local expertise of the scientist/engineers who, in their eyes, make no distinction between technae and politaea (Winner 1986). Woolgar [1995] refuses another great divide in so doing—that between individual psyche and collective repertoires of behavior. Clarke [1995b] challenges us to examine the stuff of science—the material substrate—instead of ignoring it in the service of idealist theories. Lynch’s (1995) work extends this point to a revaluation of the idea of place in science, arguing instead for distributed, material topical contextures in which to examine scientific genre. Latour [1995] (writing as Johnson) uses the device of analyzing an ordinary, low-tech system—the door—and its surprising complexity to point to the inseparability of the technical and the social.

Our key questions here are those of general political theory and of feminist and third world liberation movements: Cui bono? Who is doing the dishes? Where is the garbage going? What is the material basis for practice? Who owns the means of knowledge production? The approach begins in a very plain way with respect to science and technology by first taking it off the pedestal (Chubin and Chu 1989)—by treating science as just something that people do together. Some of this means looking at science and technology as the occasion for people to do political work—not necessarily by other means, but fairly directly. Science as a job, science as practice, technology as the means for social movements and political stances, and science itself as a social problem—collectively, these articles take science/technology as the occasion for understanding the political and relational aspects of what we call knowledge. This introduction situates these papers with respect to science and technology studies (STS), and gives a sense of the problems they are addressing in social theory.

Most work in STS has not been seen as general social theory or as contributing in a fundamental way to social science theory. While historians, philosophers, and even computer scientists show a great deal of interest in the new sociology/anthropology of science, most social scientists view it as a kind of luxury, an arcane comer of the discipline offering only specialized insights. One purpose of this book is to demonstrate that social studies of science and technology are addressing a set of questions central to all social science. In this selection of science studies research, readers will see that science and technology are the vehicles for analyzing some very old questions. How do people come to believe what they believe about nature and social order? What are the relationships between work practices and social change? What is the trajectory of social innovations? Who uses them and for what purposes? As people from different worlds meet, how do they find a common language in which to conduct their joint work? How can we study people’s work critically, yet as ethnographers or historians respect the categories and meanings they generate in the process? Finally, what are the boundaries between organism and environment; how fixed are they; how can we know them; and are they meaningful a priori?

Perhaps because learning another scientific language is a prerequisite for doing the kind of social studies of science/technology described herein, scholars in science studies tend to be an omnivorous bunch. We read history and philosophy of science as well as scientific tracts in the substantive fields we study, feminist theory, and Science magazine. We often borrow models from those writings as well as from other areas of social science. We work across national boundaries in informal groups clustered around an analytic topic: the use of metaphor, for example, or the extent of technological determinism or infrastructural change. The field is small (although growing), lively, and filled with debates, cross-fertilization, and often surprising collaborations. For example, sociologist Michael Lynch [1995], who is interested in visual representation, has collaborated with art historian Samuel Edgerton on analyzing the pictures astronomers create (Lynch and Edgerton 1988). Steve Woolgar became the project manager of an industrial computer development firm in order better to understand the process of technology construction. Diana Forsythe (1992, 1993), a cultural anthropologist by training, has worked for many years in collaboration with computer scientists building medical expert systems, both critiquing the notion of expertise and acting as a designer. Because the field is so interdisciplinary, the term science studies often replaces the disciplinary-specific label, such as sociology of science.

Historical Review

In the United States, most sociology of science before the late 1970s was dominated by the work of Robert K. Merton and his associates (see, for example, Merton 1973; Zuckerman 1977, 1989) and by a group of researchers conducting bibliographic citation analysis (from which work came the concept of invisible colleges; see Crane 1972; Mullins 1973a,b). While there had been much criticism of functionalist sociology on a number of fronts, particularly from symbolic interactionism and Marxism, the sociology of science received scant attention. Symbolic interactionists, for example, had studied work and occupations, medicine, deviance, gender, the family, urban life, and education, critiquing functionalism in each of these areas. Yet they had produced only a few scattered monographs and articles on science (for example, Glaser 1964; Strauss and Rainwater 1962; Marcson 1960; Becker and Carper 1956; Bucher 1962) and had undertaken no large research programs in this area.

Meanwhile, in Europe in the early to mid-1970s,² a group of researchers began a series of studies to demonstrate, contra Merton, that science was not disinterested, communistic, and universal. Many of them had also been deeply influenced by Kuhn’s Structure of Scientific Revolutions (1970), a book that had questioned the cumulative nature of science and raised the issue of the incomparability of scientific viewpoints or paradigms. They were concerned to show that science was not neutral, that the outcomes and content of science as well as access to it as a profession were determined by structural commitments, political positions, and other institutional considerations. MacKenzie’s (1981) work on the interrelationships of statistics and eugenics is a good example of these efforts. These researchers were also concerned to demonstrate the constructed nature of science and its view of nature. Thus, they were strongly antipositivist. Some of this work was done at the University of Edinburgh, and the interests model became especially associated with the Edinburgh School (see, for example, Barnes 1977). Other important centers with overlapping approaches were in Paris and Amsterdam.

In 1979, Bruno Latour and Steve Woolgar published Laboratory Life, probably the book from the field that is most well known outside of science studies. The book was an ethnographic study of a scientific laboratory, and its purpose was to document the creation of a scientific fact. Using a variety of techniques from anthropology, semiotics, and ethnomethodology, they traced the birth of a biological fact in the context of lab work. They concentrated on a process they called deletion of modalities, a progressive stripping away of contextual information about production, with the end result being a fact bare of its own biographical information. The book was an immediate success and was one of the factors helping spawn a series of laboratory studies and descriptions of act-making, often ethnomethodological in approach.

The combination of fieldwork and antipositivism was familiar to American symbolic interactionists, who welcomed the chance to apply these techniques to science and to learn from colleagues in Europe. A number of collaborations ensued among researchers in America, England, France, and the Netherlands pursuing these viewpoints. (See Fujimura, Star, and Gerson 1987; Clarke and Gerson 1990; Star 1992; and Clarke 1990, 1991 for reviews of this work.) Among our common interests and beliefs was the necessity of opening up the black box in order to demystify science and technology; that is, to analyze the process of production as well as the product. Methodological directives for those of us working in the interactionist tradition were familiar: Understand the language and meanings of your respondents, link them with institutional patterns and commitments, and, as Everett Hughes once said, remember that it could have been otherwise (Hughes [1959] 1971, 552). Many of our colleagues in Europe held similar views, albeit from very different traditions: Do not accept the current constructed environment as the only possibility; try to understand the processes of inscription, construction, and persuasion entailed in producing any narrative, text, or artifact; try to understand these processes over a long period of time (some of this work is represented in Law 1986; Bijker, Hughes, and Pinch 1987; Callon, Law, and Rip 1986).³

There were and are many other groups throughout the world studying science or technology: policy makers, historians, analysts of the impact of technology (particularly computing and automation), and the number is growing rapidly. Another important development began as programs of social science research gelled into science, technology, and society programs at a number of technical institutions and regular universities. New undergraduate and graduate STS programs began to spring up, both within traditional departments and as interdisciplinary programs. Early STS programs represented an amalgam of interests: ethics and values in science and engineering, studies of social impacts of technology, and history of science and technology. They were often an academic home for science criticism, that is, studies that demonstrated scientific bias (racism, sexism, classism) or danger resulting from scientific and technological research and development (nuclear and toxic wastes, recombinant DNA, technological disasters). Criticism of the sacrosanct institution of science and explication of the constructed nature of nature have remained core problems in science studies, and there is currently lively debate about the role of activism in the field, as Croissant and Restivo [1995] indicate.

Questions of Organizational Scale

Questions of organizational scale have always plagued (or some might say, graced) social science. Is social change individual or aggregate? How can we understand the relationship between social facts and individual experience?

These questions appear in science as it is interlaced with presumptions about the nature of scientific inquiry. If researchers accept that nature is simply out there waiting to be discovered, they may append to that belief the idea that anyone can do it, geniuses better than the rest of us. There is nothing that logically ties these two together, but much of the received mythology about science involves great men (sic), great moments, great labs, and great accidents of Nature revealing herself. This combination of individualism, positivism, and elitism conspires to confuse the question of the appropriate level of organizational scale at which to conduct inquiry. So, the secondary literature on science is littered with psychologism, reified societies that act in mysterious ways on believers, and participant histories that claim exclusive centrality for powerful, rich institutions and people.

Against this trend is a lively debate, partially represented in the pages of this book, about the right unit of analysis for studying science. In escaping from the nasty things mentioned in the previous paragraph, sociologists and anthropologists of science have invented, borrowed, or transformed units of analysis from other parts of the discipline or from science itself: bandwagons, social movements, political economy and large-scale work organization, units of action and activity that cross human/nonhuman boundaries, the taken-for-granted truth about Nature that reflects old and widespread conventions (and superstitions).

How is the little black box of the computer, the test tube, or the door-closer joined with phenomena at larger scales of organization? This is a fundamental question about science and technology, but it appears whenever one questions the nature of local social arrangements as articulated with those at a distance or with considerably more power and purview. All of the articles [in Ecologies of Knowledge: Work and Politics in Science and Technology (Star 1995a)] attempt to answer this question ecologically and propose several modes for doing so.

At the largest scale of organization, questions are raised here about the utility and role of science or technology in maintaining or changing the status quo. This is asked not simply in terms of technological determinism, but in terms of larger scale issues, a central one being: Can there be a revolutionary science/technology in the absence of revolutionary social change in other spheres? To the extent that one believes in the interpenetration of spheres and science as a social institution of its historical time and place, the answer must be no. This puts the question of political commitment squarely at the center of science studies. For one thing, it is difficult to escape examining oneself as a scientist while engaged in studying scientists full time. Truly revolutionary science or technology thus means full-scale revolution. The sociology of science might allow us better to understand what that might mean.

Methodological Issues

Interwoven with questions of scale and politics are questions of method. Scientists are very challenging respondents. For one thing, all scientists share with us concerns about reliability and validity of data, robustness of findings, and the meaning of those findings. I never met a scientist who had not thought about the issues raised in this introduction. As a group of respondents, scientists are particularly difficult and rewarding because they have often thought rigorously about the issues we are investigating, and about which we are ourselves uneasy [Woolgar 1995]. So the work of our respondents blends with our own. The meaning of participant observation in this case can begin radically to change.

There are several kinds of work to be juggled in doing research in STS, each of them methodologically challenging. First, there is the map and language that the scientists themselves use in their work. Second, there is our mapping of the work practices and organization. Third, we create maps of the communications between domains. Fourth, a complex nested map is generated that shows who answers to whom, and why. It is at this level that questions of unit of analysis, or scope, often show up in force. Reconciling the different maps is a nontrivial methodological problem, again common across many domains of social science and political life.

Why I Am Not a Nazi: Realism and Relativism in Science and Technology Studies

One of the curious things about being in the STS field is that one is immediately plunged into philosophical debates about realism and relativism. Briefly, realism is the position that there really is a there out there, and it’s true in some absolute sense. Relativism holds that truths are relative to a place, time, or person (often a historical situation or geographic/cultural location). All researchers in science studies have had the experience of being challenged about the underlying truth of science: What about the scientific method? What about truth? What about the laws of gravity?

I have been involved in science and technology studies for about fifteen years, first as a science critic, then as a historical sociologist and ethnographer. I have given over one hundred talks on various aspects of the sociology of science and technology. In almost every presentation, I have been asked some version of what I now call the there there question: But are you saying it’s all socially constructed? Doesn’t that mean anything could be true? Isn’t there anything out there? Are you saying that scientists are making it all up? Are you saying germs don’t really make you sick, or gravity doesn’t really make things fall down?

It is indicative of the central place of science in mainstream Western belief systems that merely to imply that the acquisition of scientific knowledge is work, not revelation, seems to involve the kind of radical idealism (if not radical autism) alluded to above. But this is not necessarily the case.

To say, as Hughes ([1959] 1971, 552) did about social order, that it could have been otherwise is not to say that it is. And to say that the conditions of nature or science are the result of collective enterprise that includes humans and nonhumans is not to imply that the merest whim on the part of an individual could overturn them. Rather, as social scientists, let’s ask: Under what conditions do such questions about reality routinely get raised?

First, the term socially constructed is a reformist term, inserted into titles in sociology/anthropology of knowledge and science. Its initial purpose was to demonstrate that the reports of science that had been stripped of production history were missing important historical and situated accounts; second, to restore accounts of the actual work and its organization to those reports. Furthermore, if one takes society as the scientific problem, then the image of a society out there structuring an experience that is then entered into the canon of research doesn’t make any sense either.

I call the idea that cumulative collective action is flimsy the mere society argument. It is paralleled by simplistic perceptions about, for example, socialization and gender. The argument goes something like this: So, she’s been socialized as a girl. Well, let’s just let her into the corridors of power and de-socialize her, and then everything will be ok. Such a statement depends on a trivial and reified conception of both socialization and gender. Whatever bundle of actions, past and present, we might think of as socialization here is far more complex and durable than most of us realized in the early days of feminism. Similarly, the notion of institutionalized racism has been crucial in understanding that racism is not simply a matter of people not being nice to each other, nor necessarily to be found in a single set of micro-interactions—rather, it is a web of racist discourse and practices that extends through and informs all human practice—and cannot be simply transcended (hooks 1990). The durable bundle of actions and experience that comprise science has a similar sturdy complexity. This complexity does not defy its ontological status as created, however. The constructivist or relativist schools in science studies (and I will not explicate the subtle differences between them here) have often been accused of flimsiness or mentalism on grounds that deeply confuse epistemology (how do you know what you know) with ontology (how are you what you are).

Scholars in science studies have disagreed about this issue and will continue to do so for some time to come. Yet a thread runs throughout the work of the groups represented in this issue: Let’s replace the either/or dichotomy of constructed versus real with more useful concepts. Concepts such as workplace ecology, irréductions (Latour 1987), sociological imagination, networks and translations, and boundary objects (Star and Griesemer 1989) are important here. Wimsatt’s (1980) concept of robustness has similarly been an important replacement for more restrictive concepts of reliability and validity. He, borrowing from biologist Richard Levins, defines it as the intersection of independent lies, or more sociologically, the durability of collective action despite the fragility of any one instance.

During the 1980s there were scores of articles and books addressing this class of questions in science and technology studies. They have important links as well with earlier work in other parts of sociology and anthropology. For example, the debate in the 1960s about labeling deviance asked whether some things aren’t really sick (or unnatural). The sociology of art has been concerned with the question of whether some things aren’t really just beautiful (in a timeless or transcendental fashion).⁴ The enduring concern with ethnocentrism in anthropology has recently exploded in debates about the place of the anthropologist and whether the knowledge constructed by anthropology is rightly seen as a jointly created fiction. In sociology and anthropology of medicine, the debate occurs as a question about whether one can differentiate physiological disease from illness behavior—aren’t some things really just germs?

But the analytic trick in each of these cases is to raise the concept of really to the status of rigorous, reflexive inquiry and ask: Under what conditions does the question get raised?

One of the difficult things about trying to analyze an institution as central as science is that one challenges the received views of things for audiences and respondents. In giving talks that defend the above position, I have sometimes been called a Nazi, or parallels have been drawn between the social construction of science and Nazi science. It took me a while to figure out what people were talking about in these accusations, since being a Nazi is anathema to me.

If one takes the point of view that fascism requires a kind of situation ethics and requires that one redefine the situation according to opportunism or a kind of distorted view of science and nature, then any attempt to make relative any situations (especially natural or scientific ones) becomes morally threatening. This is so because one antidote to fascist ideology is to affirm an overriding value in human life, a universal value that cannot be distorted by the monstrosities informed by local, parochial ideologies of racism and genocide. Ethicists often base their arguments on this presumption. The worst thing for an ethicist is to hear arguments that plead special circumstances—the name of the game is finding good universals.⁵ Yet this criticism of relativists as Nazis shows another kind of confusion, which again relies on a separation of the social and the natural and a separation between the conditions of production and the product. If the relative ontological status of a phenomenon is inextricably embedded in the conditions of production, then it’s not a question of an analyst legitimating genocide or situation ethics. Rather, the question on a meta-level becomes: How can we make a revolution that will be ontologically and epistemologically pluralist yet morally responsible? Can we be both pluralist and constructivist, hold strong values and leave room for sovereign constructions of viewpoints? These are not new questions, either; both the French and American Revolutions were fueled by them. I would claim that there is stronger evidence for Nazism arising from ignorance of the conditions of production of knowledge than from exploring the relative configurations of these conditions in different times and places; more oppression from the appeal to absolute natural law than from negotiations about findings. While I’m not implying here that science studies is the best weapon against totalitarianism, the fact that this question arises so frequently in so many different contexts is to me indicative of the fact that science has been such an inviolate institution, certainly in academia.

Current Intellectual Development in STS

Taking on science as a social construction grew beyond either interest explanations or laboratory ethnographies by the end of the 1980s. Science and technology studies has over the past several years worked hard at two central intellectual currents, both of which are at the core of an ecological analysis of science (or perhaps, in some sense prior to it). The first is the establishment of science as materially based (see especially Clarke and Fujimura 1992a,b; Haraway 1989; Clarke [1995b]; Lynch [1995]); the second is science as a form of practice (see especially Star 1989; Pickering 1992).

It is remarkable for how long accounts of science (in history, philosophy, and sociology/anthropology of science) neglected to notice that much of the activity we call science consists of people manipulating materials, including specimens, media and cultures, breeding colonies, and display items. This material culture of science is important not just as another form of exoticism, but for the ways in which it is constituent of scientific findings and constraining of the ways we perceive scientific meaning.

Of Humans and Nonhumans

One of the issues that appears in different ways in the papers in Ecologies of Knowledge: Work and Politics in Science and Technology (Star 1995a) is the issue of where to draw the line in analyzing science and technology. Traditional studies usually drew the line at the edge of the black box, whatever it might be: the computer, the laboratory, the closed scientific work group. The argument in Ecologies of Knowledge is that science studies in the past have left out some of the most important actors, the nonhuman ones. Many of the new sociologists of science are engaged in a kind of democratization of this analysis, as the papers here demonstrate. If one adopts an ecological position, then one should include all elements of the ecosphere: bugs, germs, computers, wires, animal colonies, and buildings, as well as scientists, administrators, and clients or consumers (see Clarke and Fujimura 1992b and Latour 1987 for an analysis of this). The advantages of such an analysis are that the increased heterogeneity accounts for more of the phenomena observed; one does not draw an arbitrary line between organism and environment, one can empirically track lines of action without stopping at species, mechanical or linguistic boundaries, and especially without invoking a reified conception of society.

On the other hand, this kind of analysis presents some serious ethical problems—on both sides (Singleton and Michael 1993). For many years feminists, radical ecologists, and pantheists have recommended a kind of analysis that does not exclude anything from the natural world. The exclusion of animals, the biological environment, and other parts of the natural context has been one of the major sources of alienation under patriarchy, late capitalism, or religions that are antinature (Griffin 1978; Merchant 1980; Harding 1991). Restoring the natural world to the research context would be an ethical and political advance. On yet another hand, the papers by Kling and Iacono [1995] and that by Star [1995b] are written from within a research context in which it is not humans who have been privileged at the expense of nonhumans, but vice versa. It is computers and automation that have occupied a privileged position vis-à-vis human beings, often because of the inadequate social analysis held by computer movement advocates. An ethical social problems position in this case most likely involves checking the power of nonhumans and their advocates and seeing that humans understand it contextually, not democratizing the nonhuman position. Thus where Latour [1995] is concerned to restore ecology from