«A Little Dirt Never Hurt Anyone: Knowledge-Making and Contamination in Materials Science Cyrus C.M. Mody In the two decades since the genre emerged, ...»
In many laboratory sciences, issues of cleanliness and purity are everpresent concerns. In materials science, keeping things (instruments, materials, people)
clean structures the knowledge-making process. Using the work of Mary Douglas, I
examine various contaminants, impurities and defects that are relevant to materials
scientists. Importantly, though, deﬁnitions of what constitutes ‘dirt’ are multiple,
overlapping and, often, formally contradictory; this means that impurities are as much positive resources as threatening pollutants. In materials science labs, where many kinds of actors and forms of life intersect, pollution may be used to rein in confusion and ambiguity. This paper traces various manifestations of laboratory dirt, then examines how (un)cleanliness enables certain moves in the materials science game.
Keywords clean, ethnography, laboratory, pollution
A Little Dirt Never Hurt Anyone:
Knowledge-Making and Contamination in Materials Science Cyrus C.M. Mody In the two decades since the genre emerged, studies of the social worlds of laboratories have often commented on practices such as abstraction, extraction, puriﬁcation, ﬁltering, rendering, and so forth – practices that bear the family resemblance that they deal with cleaning, cleansing and decontaminating bits of the world to make them more accessible to scientiﬁc study and to translate them into the workable material context of the lab.1 A few analyses have even taken the theme of puriﬁcation a step further, and have explored the ramiﬁcations of terms like ‘clean’ and ‘sterile’: among these, Pearl Katz has looked at the mechanisms of contamination and pollution avoidance in the operating room; Stacia Zabusky has described cleanliness concerns among engineers at the European Space Agency; and Barbara Rawlings and Stefan Hirschauer have both analysed the social organization and symbolic signiﬁcance of sterility in, again, surgical settings.2 In one particularly insightful article, Kathleen Jordan and Michael Lynch discuss contamination as they examine the various avatars of the plasmid prep in biology laboratories.3 Jordan and Lynch’s work is of interest because it uncovers the idiosyncrasies of local interpretations of the plasmid prep along with the forces and entities which can contaminate it and make it succeed or fail; this sort of ﬂexibility in Social Studies of Science 31/1(February 2001) 7–36 © SSS and SAGE Publications (London, Thousand Oaks CA, New Delhi) [0306-3127(200102)31:1;7–36;017579] 8 Social Studies of Science 31/1 discerning contaminants and ways of getting rid of them is a recurrent theme in understanding dirt and cleanliness in laboratories.
What these and a host of other studies show is that dirt and pollution are driving forces in the social life of a wide range of sciences; nevertheless, a detailed examination of the rôle of dirt in laboratory settings, containing an outline of the various kinds of pollution, the ways in which they are dealt with and/or used as resources, and their intertwining with the social organization of laboratory participants, is still lacking. One particularly important lacuna is what might be called the ‘knowledge element’.
Those analyses most focused on purity and cleanliness, such as those by Hirschauer, Katz, Rawlings and Zabusky, concern the pragmatic rôle of contamination (and the social forms attendant upon it) in the achievement of desired outcomes (in surgery or engineering), rather than showing how these practical considerations, along with the dirt embedded in them, go into the production of knowledge.
This is all the more surprising given that one of the foremost contributors to the ﬁeld of science studies, Mary Douglas, is also the foremost contributor to traditional anthropology’s understanding of pollution and its rôle in the creation of socially held knowledge about the world. Many science studies scholars, such as David Bloor, Barry Barnes, Steven Shapin, Michiel Schwarz and Michael Thompson, have capitalized on her later works – Natural Symbols (1970), Implicit Meanings (1975), How Institutions Think (1986) and studies of Risk4 – to unpack the cosmologies and hierarchies of social groups that produce scientiﬁc knowledge.5 Her early explication of dirt, Purity and Danger (1966),6 however, has largely been neglected in science studies, despite the implications it might have for analysing the production of knowledge and culture in the laboratory.
This paper does not aspire to a totalizing, cross-disciplinary analysis of cleanliness and dirt that would do for laboratory studies what Purity and Danger did for anthropology and comparative religion. Indeed, I will try not to take a synoptic view of practices like purifying, etching, polishing, annealing, milling, coating, and so forth. There is no one thing uniting these and other disparate laboratory phenomena: rather, they are overlapping strands making up an analytically interesting thread around themes of purity, cleanliness and contamination. I leave the exact boundaries of this thread to emerge throughout my paper, but its endpoints are generally the practices and techniques of ridding, cleansing and purifying parts of the laboratory world, as well as the entities and phenomena that are thereby cleansed or avoided. I extend my analysis, however, to follow the variabilities of lab practice, where what is labelled as polluting one minute is turned into a necessary tool the next, and where the ‘clean’ and ‘dirty’ parts of the world are often overlapping and intermingled.
This study therefore moves somewhat beyond Douglas’ framework (at least in Purity and Danger), where ‘the central idea is that accusations of causing dirt and deﬁlement are weapons against disorderly behaviour’.7 Dirt as a means to control behaviour is a critical fact in many laboratory sciences; but I hope to show that pollution and contamination are often Mody: A Little Dirt Never Hurt Anyone 9 sought out by scientists as ways to expand the boundaries of their knowledge, as solutions to problems and anomalies, and as guides in their local logic of practice. Nevertheless, the threatening aspects of pollution are rarely far from the surface, in ways that make Douglas’ ideas essential starting points in talking about laboratory dirt.
What is Materials Science?
The ﬁrst such point is that pollution emerges within communities. Douglas explains that virtually all social groups have ideas about dirt and deﬁlement, but that the meanings attributed to contamination can only be understood in relation to local contexts, meanings and practices. Thus it seems clear that practically every laboratory and ﬁeld science has vocabularies of pollution and puriﬁcation, contamination and cleansing.8 While we might be tempted to say something about the rôle of dirt across all these vocabularies, the empirical data needed for such a comparison have not yet been collected. Moreover, even within a particular scientiﬁc setting, ideas about dirt can be highly variable. For the moment it is necessary to take a ﬁne-grained look at a few interrelated and interwoven laboratory contexts and to examine the speciﬁcities of how dirt is (often literally) embedded in those settings.
The particular setting for this dirt story, therefore, is a small group of graduate students working under one professor in the Materials Science and Engineering Department at Cornell University. I carried out my observations continuously over the summers of 1998 and 1999, and intermittently between-times (summers happened to be the periods of most active research for my informants as well). Four graduate students were members of the group during the time of my ethnography: I will refer to them as Henry, Rudy, Angus and Paul.9 They constituted my pool of key informants, but I also observed and talked with their group leader (Professor Sitman), several of their undergraduate laboratory assistants, and the many other technicians, professors, graduate students and laboratory personnel with whom they interacted. In addition, as I became more and more interested in how my informants dealt with cleanliness, I decided to observe the hazardous waste inspectors of the Cornell Environmental Health and Safety Ofﬁce, one of the campus entities charged with regulating certain important kinds of pollution.
There are a number of reasons why materials science is an exemplary site for looking at issues of cleanliness and dirt in the lab. ‘Materials science’ is an umbrella term for an extraordinarily diverse set of practices and occupational and disciplinary afﬁliations. It includes, but is certainly not limited to, the study of metals, polymers, ceramics, glasses and composites, with respect to their electrical, mechanical, thermal, optical, and other properties. We know from Mary Douglas that purity and danger become most relevant at the boundaries between cultures and systems of classiﬁcation. Laboratory settings where different kinds of knowledge interact are places where dirt and deﬁlement are likely to matter.
10 Social Studies of Science 31/1 The institutional arrangements for materials science at Cornell highlight this heterogeneity. My informants belong to the Materials Science and Engineering Department, a ﬁeld within Cornell’s School of Engineering.
They also have a close relationship with the Cornell Center for Materials Research [CCMR], however, which is supervised largely by physics faculty located within the College of Arts and Sciences; moreover, CCMR has facilities in physics, engineering, geology and chemistry buildings, all of which were utilized by my informants during my ethnography. Cornell’s interest in materials science has been intense since 1959, when it was awarded funds for an ARPA Interdisciplinary Laboratory (IDL) in that area.10 Many who receive the appellation ‘materials scientist’ might alternately refer to themselves as ‘scientists’, ‘engineers’, or as practitioners of more academically traditional disciplines (chemists, physicists, applied physicists, geologists). Alternatively, they may present themselves in terms of the material or application with which they work (ceramists, metallurgists, ‘device’ people, ‘plastics’ people). One important aspect of materials science is its relation and orientation to other engineering sciences – materials science is about both making and understanding, and many of its products (both intellectual and material) are embedded in numerous highand low-tech artefacts.11 From this perspective, we can see materials science as encompassing (1) the characterization of the properties of materials; (2) the investigation of how those properties come to be; (3) research into how to manipulate such characteristics in individual specimens; and (4) the production, on industrial scales and using industrial methods, of those same characteristics in materials which go on to form components of devices in widespread application.
With this understanding of materials science, a second reason for the relevance of dirt emerges; for the distinction between dirt and non-dirt is tied closely to ideas about the organization of matter. Douglas’ initial deﬁnition of dirt, for instance, is that it is ‘matter out of place’. As she puts it, this deﬁnition...
... is a very suggestive approach. It implies two conditions: a set of ordered relations and a contravention of that order. Dirt then, is never a unique, isolated event. Where there is dirt there is system. Dirt is the byproduct of systematic ordering and classiﬁcation of matter, in so far as ordering involves rejecting inappropriate elements. This idea of dirt takes us straight into the ﬁeld of symbolism and promises a link-up with more obviously symbolic systems of purity.12 Materials science, insofar as it is the ‘systematic ordering and classiﬁcation of matter’ should, by Douglas’ logic, have a close relationship with notions of purity. In general, contamination can pose a serious danger to the practice of materials science; the possibility of disorder, however introduced, threatens to reverse all of the materials scientist’s work in bringing about order.
In fact, at ﬁrst glance it might appear that dirt is always the materials scientist’s enemy; as we shall see, though, this belief misses many of the Mody: A Little Dirt Never Hurt Anyone 11 subtleties about cleanliness that are key to the practice of the discipline. In certain areas of materials science, particularly those most closely associated with electrical engineering and the semiconductor industry, the rhetoric of cleanliness is employed freely, and entangled in arguments for nanotechnology and miniaturization that verge on technological determinism.13 Note, for example, this quote from the website of Intel (a company that uses bunnysuited clean-room technicians as icons and selling points for its
line of microelectronics products):
Even though the basic material for a microprocessor is sand, a single speck of dust in the air of a fab could ruin thousands of chips. So the environment in which microprocessors are manufactured has to be very clean. In fact the clean rooms in a fab where the chips are made is more than ten thousand times cleaner than a hospital operating theatre. A ‘class one’ clean room, the cleanest of clean rooms, has one speck of dust per square foot.... It’s hard to imagine the absolute cleanliness of a fab.
However hard you try to think how clean a fab is, it’s cleaner.14 One last reason to look at dirt and deﬁlement in materials science, then, is to examine more carefully some of the most public talk about cleanliness in materials science. One purpose of this paper is to show that the systematic expulsion of contamination found in clean rooms can only be understood as part of a wider ﬁeld of materials science and engineering, where ‘unclean’ laboratory practices are often prerequisites for the work done in fabs.
What Counts as Dirt?