The Competitiveness of Nations in a Global Knowledge-Based Economy

Alberto Cambrosio † and Peter Keating ††

“Going Monoclonal”: Art, Science, and Magic in the Day-to-Day Use of Hybridoma Technology *

Social Problems, 35 (3)

June 1988, 244-260

Content

Abstract

Introduction

Local Knowledge and Tacit Knowledge

Hybridoma Technology

The Art of Producing Hybridomas

The Objectification of Procedures

Hybridoma Technique and Its Variants

The Problem of Standardization

From Art to Science

Discussion

References

Abstract

Recent work in the sociology of science has highlighted the local and tacit dimensions of scientific work  Against the widely held assumption that we are here dealing with a form of knowledge largely beyond the control and manipulation of scientists, we will argue that the unsaid is indeed a part of conscious scientific practice - and hence subject to negotiation, discussion, and construction.  Based on a study of the transmission of hybridoma technology, this paper will show that questions of local knowledge, tacit knowledge, and “magic,” far from being ignored by scientific researchers, are explicitly a part of their daily practice.  It will be seen that these questions give rise to a series of social and technical distinctions which are constitutive of scientific work.

Introduction

In recent years, the sociology of science has moved from the study of scientific communities to the problem of the construction of scientific knowledge (Knorr-Cetina and Mulkay, 1983).  This change in interest has resulted in a shift in the scope and object of sociological inquiry as evidenced by the many “laboratory studies” and analyses of scientific controversies, both of which have tended to highlight the contingent elements of scientific knowledge.  Laboratory studies, for example, have described the local dimensions of scientific knowledge and the processes whereby the latter are integrated into the sanctioned corpus of science (Latour and Woolgar, 1979; Knorr-Cetina, 1981; Lynch, 1985; Star, 1983).  Similarly, research bearing on the conduct of scientific controversies has stressed the negotiable elements in these episodes and the role played by tacit understanding in the replication of disputed experience (Collins, 1985; Pinch, 1986).  In this context, the categories of “local knowledge” and “tacit knowledge” have gained central importance.  However, the meaning given to these notions is hardly uniform, and there is indeed some confusion surrounding their use.  To begin with, let us consider the relation between local knowledge and tacit knowledge.

 

Local Knowledge and Tacit Knowledge

While these two terms have often been used as either synonymous or overlapping categories of knowledge, some authors have taken great pains to distinguish them.  Knorr-Cetina (1981:37-40, 127-130), for example, has proposed an analysis of scientific knowledge that distinguishes public knowledge, tacit knowledge, and local knowledge (know-how).  The first two categories are separated from the third on the basis of “availability.”  According to Knorr-Cetina, public knowledge and tacit knowledge refer to scientific knowledge that is “generally

Université du Québec a Montréal; †† Harvard University

* A preliminary version of this paper was presented to the meeting “Senses of Science’ of the European Association for the Study of Science and Technology (Strasbourg, 1986).  We would like to thank Michel Callon for his comments on the final draft.  Research for this paper has been made possible by post-doctoral fellowships awarded to the authors by the Social Sciences and Humanities Research Council of Canada as well as by S.S.H.R.C. grant 410-86.0341.  Correspondence to: Cambrosio, Centre de recherche en evaluation sociale des technologies (CREST), Université du Québec a Montréal, C.P. 8888, Succ. A, Montréal, Québec H3C 3P8, Canada.

244

available” to the scientific community, while local knowledge is composed of the largely inaccessible idiosyncrasies of the individual researcher or laboratory.  Knorr-Cetina’s classification of knowledge corresponds, on the one hand, to the categories of scientific institutions used by sociologists and historians (i.e., field, discipline, institute) and, on the other, to the classification of knowledge used by scientists in their daily practice (i.e., science, art, magic; see also, Fujimura, 1987).

Michael Lynch’s (1982; 1985) notion of tacit knowledge is similar to Knorr-Cetina’s notion of local knowledge.  Lynch, however, has gone further in removing all reference to individuals as possible “repositories” of knowledge.  For Lynch, tacit knowledge, unlike Knorr-Cetina’s local knowledge, is “not anybody’s knowledge.”  In fact, Lynch’s ethnomethodological approach is an attempt to go beyond the description of the unwritten knowledge which circulates within a given scientific discipline in order to describe a more fundamental form of knowledge, incorporated in the practices themselves.  This knowledge is “taken for granted” by the researchers and, although unnoticed, plays a central role in the conscious evaluation of experiments and experimental results, giving rise to what Lynch terms “endogenous critical inquiry” (as opposed to professional sociological inquiry).

H.M. Collins’s work on scientific controversies also relies to a great extent on the notion of tacit knowledge.  And, here again, the notion is given a somewhat different meaning than that attributed to it by Knorr-Cetina and Lynch.  In Collins’s work, the notion of tacit knowledge is intended to provide the key to understanding the author’s emerging “science of knowledge” (Collins, 1987).  In particular, Collins has recently proposed an analysis of the domain of cognition into four components: facts and rules, heuristics, manual and perceptual skills, cultural skills.  The last two elements of Collins’s analysis are members of the category of tacit knowledge, which is the knowledge required to use formal and informal rules but which cannot, without being fundamentally transformed, be verbalized or formalized.

Applied to the analysis of scientific controversies, Collins’s approach results in the characterization of these practices as proceeding along the lines of the “experimenter’s regress,” which is described as the following situation.  Because of the artisanal nature of scientific experiments, a researcher’s competence and the reliability of his/her experiment can only be assessed on the basis of the results obtained.  Since, however, the results can be considered valid only if they are obtained from reliable experiments, any assessment is necessarily flawed by circular reasoning.  Collins argues that scientists are able to break this circle of reason by negotiation, the consensual resolution of which depends upon the prior existence of a network of social relations and therefore an institutionalized “form of life” for the researchers (Collins, 1985).

This model of scientific practice, referred to as “enculturational,” allots tacit knowledge the key role in the production of new knowledge and is, thus, at variance with what Collins terms the algorithmic model defended by researchers when they are asked to account for their activity.  According to this latter model, scientific practice is described as the application of rules (scientific knowledge being composed solely of formal and informal rules).  Insofar as Collins’s model is opposed to the scientist’s model of scientific practice, it is reducible to the distinction between what can be said and what can not: “... the crucial division in knowledge is not the separation between information and heuristics... but between the articulateable and the tacit” (Collins, Green and Draper, 1985:329).  Vincenti’s (1984) typology of knowledge is also based on a dichotomy, this time between explicit and procedural knowledge; the former encompasses descriptive and prescriptive knowledge, the latter prescriptive and tacit knowledge.  Prescriptive knowledge plays the role of watershed between procedural and explicit knowledge.  Tacit knowledge is only a subcategory of the main dichotomy.  Collins’s argument is directed against this kind of typology.

Like Knorr-Cetina’s and Lynch’s conceptions of local or tacit knowledge, Collins’s typology suggests that tacit knowledge is impossible to articulate, and yet it is not clear why this

245

should be so.  If tacit knowledge is that which one knows without knowing how to say it, and this would seem to be what Collins believes, then there are certainly things that are known and cannot be said that can nonetheless be shown and formally transmitted in this manner.  In this respect, Ferguson’s work is quite instructive.  According to this author, there is, in fact, in the area of machine production “an enormous quantity of non-verbal knowledge which has accrued and which has been diffused through published diagrams” (Ferguson, 1985).  Surely the transmission of non-verbal knowledge may in some way count as an instance the articulation of the tacit.

“Rules of thumb” are also problematic for the concept of tacit knowledge.  Collins, for example, has classified this species of knowledge (also referred to as artisanal knowledge, tricks of the trade, etc.) as “heuristics” in recognition of the fact that it can be transmitted in more or less formal fashion.  It, therefore, does not count as tacit in Collins’s scheme.  However, conforming to a widespread tendency in the sociology of work (Jones, 1983; Jones an Wood, 1984; Kusterer, 1979), Collins himself sometimes uses terms like “rules of thumb” a synonymous with tacit knowledge.  One of the reasons for this contradictory use of terms is that it is often assumed that tacit knowledge is impossible to verbalize and that, hence, all non-verbal knowledge is tacit.

However, it is rarely specified why certain kinds of knowledge cannot, by their nature, be enunciated.  One could easily imagine, on the contrary, that some things are left unsaid not because of the impossibility of their being said, but for a variety of other reasons.  For example, an individual’s silence may be the result not of an inability to verbalize, but of the perception of the trivial nature of what might have been said.  Surely, some things are generally accepted as being known to all within a given discipline and hence trivial.  One could, moreover, conceive of knowledge unspoken for fear of sanctions or knowledge unspeakable within a certain conceptual framework (or paradigm or disciplinary matrix).  In other words, the restriction of the tacit to the unsaid does not necessarily entail that the tacit is by nature non-verbal; much that is unsaid clearly has the possibility of being verbalized but remains, for many reasons, unsayable, unthinkable, trivial, secret, or censored (Descombes, 1977).

One of the effects of prescribing a non-verbal, inaccessible, non-transmissible nature to tacit knowledge has been to reinforce the assumption that we are here dealing with a form of knowledge largely beyond the control and manipulation of scientists, that we have entered the unconscious or, at least, non-conscious realm of science.  Against this assumption, we will argue in what follows that the unsaid is indeed a part of conscious scientific practice and hence subject to negotiation, discussion, and (re)construction.  Moreover, we will see that the distinctions (formal, tacit, local) used by sociologists of science, even by constructivists, inadequately describe scientific work.

Based on a study of the transmission of hybridoma technology, this paper will show that questions of local knowledge, tacit knowledge, and “magic,” far from being ignored by scientific researchers, are explicitly a part of their daily practice (for an historical example of this awareness, see also Lawrence, 1985).  It will be seen that these questions give rise to a series of social and technical distinctions which are constitutive of scientific work.  It follows then that the establishment of a typology of knowledge is not only the preoccupation of the sociologist, but is also a recognized issue for scientists.  Rather than give a sociological description of scientific practice with the help of such notions as local or tacit knowledge, this paper seeks to describe the status scientists attribute to such knowledge.

 

Hybridoma Technology

The development of hybridoma technology began in 1975 with originators G. Köhler and C. Milstein, who received the Nobel Prize in 1984.  It is considered a scientific discovery of

246

great importance as well as one of the foundations of the biotechnology industry.  In order to appreciate its status both as a scientific and a technical advance, recall that antibodies, the product of hybridoma technology, in addition to being a key element of the body’s immune system, are an essential diagnostic and research tool in biomedicine.  Prior to 1975, only polyclonal antibodies - that is to say, a mixture of antibodies directed against a variety of antigens or different parts of a single antigen - were available to researchers.  As the cells responsible for the secretion of antibodies were impossible to cultivate, antibodies were harvested from immunized animal sera.  The disadvantage of such a procedure lay in the fact that not only did the sera of immunized animals always contain an inseparable mix of antibodies, but the quality and the composition of the sera varied from one immunization to the other.  Köhler and Milstein’s 1975 discovery consisted of a procedure for the production of an unlimited amount of a single type of antibody.  The technique called for the fusion of cells producing antibodies with a particular type of cancer cells (myelomas).  The resultant hybrid, termed hybridoma, retains both the capacity to produce a specific type of antibody and the immortal character of cancer cells that makes the cultivation of hybridomas possible.  The specific antibodies secreted by the clones of a single hybridoma are called monoclonal antibodies.

Since 1975, hybridoma technology has diffused rapidly through clinical and research biomedical specialties, as evidenced by the exponential growth in publications concerning monoclonal antibodies.  According to  Medicus, by 1984 the number of articles on this technology had increased to over 10,000.  In the space of a few years, hybridoma technology had gone from a highly specialized technique to a standard technique that few biomedical laboratories could afford to be without.

Nonetheless, the production of monoclonal antibodies has remained a largely artisanal technique.  Unlike a number of biochemical procedures, for example, it has yet to be automated.  The animals (rats or mice) must first be injected with the chosen antigen in order to be immunized.  The spleen of the immunized animal is then extracted, and the antibody producing cells are mixed with myelomas in the presence of a chemical (polyethylene glycol or P.E.G.) which promotes cell fusion.  Following fusion, hybridomas are separated from the un-fused cells; this is accomplished by placing the mixture in a solution in which only the hybrids will survive.  After a period of growth in culture, hybridomas producing the required antibodies are selected using a variety of techniques (for example, so-called immunoassays).  The selected hybridomas are then cloned, cultivated, and frozen for conservation.  Subsequently, the hybridomas may be either cultivated in vitro in order to harvest antibodies secreted in the medium or injected into animals where they generate tumors which in turn secrete antibodies into the animal’s internal cavity.  Depending upon the reason for which the antibodies were produced, the latter may be submitted to a series of additional manipulations, including immunochemical characterization of the antibodies and conjugation with radioactive or enzymatic labels.

Hybridoma technology is thus composed of a series of steps which call upon different domains of expertise sometimes loosely identified with different disciplines.  The initial immunization, for example, is most closely related to immunology, while the immunoassays used for the selection of the hybridomas as well as the various procedures used to characterize the antibodies are more closely identified with immunochemistry or biochemistry.  The hybridoma technique invariably draws upon expertise in cell culture, the sterile manipulation of biological material as well as the disciplinary domains related to the antigen of interest (virology in the case of viral antigens, bacteriology in the case of bacterial antigens, etc.).  This break-down of the procedure into a series of domains of technical expertise is not an analytical artifact but corresponds to the way researchers approach hybridoma technology.

Since 1975, hybridoma technology has remained for the most part unchanged (French et al., 1986; Westerwoudt, 1985).  Nevertheless several improvements have been accepted by the

247

majority of researchers, the most significant being: (1) the substitution of a chemical fusion promoter (P.E.G.) for the Sendai virus initially used to promote fusion, and (2) the use of myelomas that do not secrete their own antibodies and that therefore do not interfere with the production of the required antibody.  Other “improvements,” though widespread, are still subject to discussion.  Such is the case with the addition of “feeder cells” to hybridoma culture wells and the use of a serum-free culture medium.  Finally, some innovations have yet to be widely adopted.  Electrofusion (i.e., the use of short electric pulses to promote fusion), despite the many articles crediting it with great advances in fusion efficiency, remains restricted in its application due apparently to high cost and the effectiveness and availability of “traditional” procedures.

 

The Art of Producing Hybridomas

Hybridoma technology is often characterized by practitioners as an art.  What makes it “artisanal” and what does it mean to be such?  In order to answer these questions, we will consider both written sources, such as technical manuals, and informal statements gathered through interviews and laboratory discussions.  We will thus be able to assess what counts as being “artisanal” and to show that this category is indeed a component of scientists’ discourse.

It is well known that knowledge and know-how in hybridoma-related fields have long circulated informally.  According to Goding (1983:3), “Immunochemistry has an oral tradition, and a surprising number of key elements are not easily accessible from the literature.”  The artisanal nature of hybridoma technology is evident in the fact that the training necessary for its manipulation takes the form of an apprenticeship.  As one researcher (winter 1986) remarked, “it’s difficult to learn a technique which is art from a paper.”  Even manuals offering instruction in the technique maintain a similar opinion:

The newcomer to hybridization is well advised to learn the technique in a laboratory which is already practicing fusion.  It has been a frequent observation that newcomers to the technique are relatively unsuccessful initially and obtain many hybrids after some practice, although an experienced observer cannot see any difference between the technique used on the first day and in subsequent, successful experiments.  The best approach is therefore to learn from an experienced laboratory and practice until hybrids are obtained (Zola and Brocks, 1982:4-5).

The artisanal features of the technique are further reinforced by the often unpredictable nature of the results obtained:

The waywardness of the original method, repeatedly commented upon by its originators, is best demonstrated by comparing tests set up and run under identical conditions.  The wide scatter of results is typical: the odds of obtaining hybrid clones in all culture cups or in none is about the same (Fazekas De St.Groth and Scheidegger, 1980:5).

It is quite common for a laboratory to have success for several months and then failure for as long a period.  So the immediate record of a laboratory is more important than the long-term one.  Even within a research group one individual may fail where others succeed and this is a reason for encouraging separate reagents and cell lines (Campbell, 1984:156).

The fusion does not invariably work.  Often a repeat of the fusion process without obvious changes results in successful hybridization (Hybridoma Techniques, 1980:4).

While the rate of successful fusions does indeed increase with time, expertise is not identified with the possibility of succeeding every time.  Rather, expertise is equated with the ability to operate “confidently” in an experimental situation characterized by a high degree of uncertainty (Star, 1983).  This persisting state of uncertainty is said to follow from the interaction of “nature” and a given researcher’s professional trajectory: “the diversity of published

248

approaches [to hybridoma technology] reflects both individual biological problems and previous experience” (Goding, 1983:56).

To be experienced in an art, to have expertise, implies the mastery of specific ways of seeing and doing.  The nature of these skills and the difficulty of their transmission can best be seen in the following example.  After a certain growth period, it is necessary to transfer the hybridomas from a 96-well microplate to another in order to ensure their continued survival.  If they are transferred too soon, there is a possibility they will not survive the transfer; if too much time elapses, there is a possibility they will die in the original microplate wells.  It is furthermore necessary that the hybridomas be left in the original wells long enough to have secreted sufficient antibody to make possible their screening as producers of the desired antibody.  The moment of transfer is thus a crucial step in the procedure, and its determination requires scrutiny of the color of the culture liquid (which contains a pH indicator), the form of the cells, and their size.  However, these parameters cannot be analyzed in isolation.  The decision to transfer depends upon the perception of a synthetic variable referred to as the viability of the culture.  The perception of culture viability allows the experienced researcher to decide when the fused cells should be transferred and requires learning how to look at cells:

As the cells become very dense, they start to look unhealthy, and viability drops (Goding, 1983:67; emphasis added).

Sometimes the hybridomas do not look “happy” after the replacement of HT with normal growth medium (Eshhar, 1985:22; emphasis added).

When inspecting cells by phase microscopy, you get a feeling for just what the cells are doing, and how healthy they are by looking at them.  A good hybrid should look like a little beach ball (an industrial researcher, winter 1986).

You look in the microscope at cells growing: are they healthy or are they not healthy?  You learn that by association.  The professor says: these are healthy, those are not.  You learn by association, without knowing what you are looking at; you learn to know when “it looks good” (an academic researcher, fall 1985).

When “apprentices” are taught the hybridoma technique, the instructor (technician, graduate student, or professor) usually stresses the importance of the variables that have not been written down.  The latter refer more often than not to visual and motor aspects of the procedure which contribute to the learning of a gestalt.  This way of seeing and acting is not restricted to experimental practice per se but is also essential to the understanding of formal written instructions.  To illustrate this, we draw upon the experience of the first author.

At the beginning of this study, Cambrosio undertook a comparison of several different experimental protocols for the production of hybridomas.  He had not yet been able to attend a fusion experiment but relied, to a great extent, on his previous biological training.  While one might expect that it would be relatively easy to determine variations between the protocols, this was true only in a “mechanical” or literal sense; to the untrained eye, the protocols appeared to be arbitrary lists of instructions lacking any overall sense.  The situation changed fundamentally when he was able to attend a training session in the technique.  Once these instructions were embodied in a series of gestures, they became confounded with other factors such as the manual skills of a given person or that person’s degree of familiarity with a piece of equipment.  The comparison between protocols now became possible, each line of instruction evoking shapes, colors, time spans, and gestures that could be compared.

Ethnomethodology has developed the theme of the passage from written instructions to actual experimental practice.  On this view, much of what is important to the understanding of an experimental protocol is not contained in the instructions but is incorporated in the various visual and corporal movements that make up the actual practice.  It is further maintained that the experimental gestures themselves constitute an important part of the reasoning peculiar to specific disciplines (Lynch, 1985).  This fact has not escaped scientists.

249

Scientific researchers perceive the circulation of fellow scientists (most notably “post-docs”) between laboratories as the circulation of “incorporated” knowledge.  In one of the academic laboratories we visited, for example, a “post-doc” had introduced a particular way of shaking test tubes containing recently fused cells which, following centrifugation, had a tendency to stick to the walls of the tube.  This gesture, at first sight somewhat innocuous, was quickly adopted by the other researchers in the laboratory who pointed out that it could make the difference between the success or failure of a fusion.  Bio-industrial laboratories are also well aware of the importance and source of this dimension of scientific practice and, hence, frequently offer post-doctoral fellowships in addition to maintaining consultantships with university researchers.

 

The Objectification of Procedures

The fact that hybridoma technique has features that are widely recognized as “artisanal” has not stifled attempts to objectify the “know-how” contained therein, for not only do written protocols exist, as previously noted, but it has indeed been possible for researchers to learn the technique through such protocols.

This, of course, does not imply that hybridoma “know-how” was transmitted in this way alone.  Insofar as hybridoma technology links together a set of previously existing techniques, it can be said that “learning by doing” happened at an earlier point.  As noted by Collins (1987), the problem here is one of the distribution of know-how within a given field.  Widely distributed know-how is taken for granted and not perceived as know-how.  This point can be related to Vincenti’s (1984:571) distinction between two different modes of circulation for innovations: “diffusion pattern” and “simultaneous pattern”, the difference between the two modes laying, among other things, in “the range of availability of the necessary aptitudes and expertise.”  In what follows, we will examine the ways in which the “tacit” is articulated through the objectification and formal transmission of the technique.

There are essentially four kinds of written documents describing the hybridoma technique.  The first is found in the material and methods sections of scientific papers.  The brevity of these descriptions may be taken as a sign that the information contained therein is destined for members of the same discipline or same technical area.  They are of little use to researchers unfamiliar with the procedure.

A second form in which the technique is described is the experimental protocols produced in manuscript form by university laboratories for internal use.  The protocols are also sent to researchers who request them.  Industrial laboratories tend to produce more formal versions of this type of document (“standard operating procedures”).  However, the informal nature of some university laboratory protocols does not prohibit their transfer to industry.  This is carried out either through transfer of a researcher or as part of a commercial package.  Taggart Hybridoma Technology, developed by an academic scientist and sold under licence by the HyClone Company, provides buyers with a specific myeloma cell line, an experimental protocol, and direct technical assistance through an 800 telephone number.  Research protocols offer a condensation of a particular laboratory’s experience with a technique.  They contain a combination of scientific, artisanal, and idiosyncratic rationales and hence require strict adherence.  As Eshhar (1985:10; emphasis added) has remarked: “Fusion protocols are simple and easy to follow and usually successful, if one sticks to them.”

Books dedicated solely to the detailed description of hybridoma technology are a third kind of description and are generally justified by the importance of this technology for disciplines other than that in which it emerged.  For although the production of monoclonal antibodies draws upon a variety of skills from different disciplines, it is not uniformly accessible to members of all disciplines.  Thus an individual with experience in cell culture (e.g., a virol-

250

ogist) would have a certain advantage in learning the technique over, for example, a biochemist. However, as a technical manual notes:

It is not necessary to have extensive cell culture experience or to be an immunologist to undertake hybridoma work, although it helps... Hybridomas are rather fastidious cells and the chances of producing them and maintaining them are certainly higher if the worker has previous cell culture experience... The most important prerequisite in terms of expertise relates to the antigen type to be used and the assay for antibody against the antigen.  Hybridoma technology is secondary and can be learned, but it is essential to have experience working with the material which is the subject of the project, be it a virus or a peptide, a lymphocyte differentiation antigen or a pathogenic parasite (Zola and Brocks, 1982:7).

As the above quotation clearly indicates, the purpose of such manuals is to allow researchers from other disciplines to acquire the technique without necessarily coming to grips with the theoretical principles (and the practical consequences of those principles) that underlie it.  The technique is thus attributed a secondary or peripheral status.  It is mere technique, a tool for the furtherance of other disciplinary projects.

In other instances, hybridoma technique may be accorded a more central status.  In such cases the technique is no longer the tool of a particular discipline but a vehicle for advanced knowledge as well as a manifestation of the practical mastery of that knowledge. [1]  As one immunologist noted:

There are different strategies to make hybrids very specific against what you want: more and more fine questions are asked and you then need more specific monoclonals.  You may produce hybrids “one by one”, and keep looking, but the more specific you want your monoclonals to be, the longer it gets.  Now, very few people understand how the immune system works and the best way to do hybrids.  People who are not immunologists produce monoclonals by following instructions in the books, in other words, very inefficiently (an academic researcher, fall 1985).

Not only are there different ways of objectifying the hybridoma technique, but the technique may be objectified or formalized to different degrees; the more the technique is formalized (or “packaged”), the more it tends to be seen as “pure technique” and thus less central to the scientific preoccupations of the discipline.  Hence, it cannot be decided a priori whether the hybridoma technique is of a formal or artisanal nature or whether, in fact, it is “only” a technique.  These attributes are determined through practice and thus can serve only as descriptions of the technique within a particular institutional or disciplinary context.  The changing status of the technique, which may be described as socio-technical (Callon and Latour, 1986), appears furthermore to depend upon the shifting relationship between disciplines as well as industrial and academic institutions.

In contrast to the “material and methods” sections of scientific papers, scientific manuals offer what attempts to be an exhaustive description of a “standard” procedure.  Some authors go so far as to seek a reconciliation of the two extremes we have just described by offering both principles and technique for the uninitiated:

I have written this book because 1 believe that previous accounts of the production, and particularly the usage, of monoclonal antibodies have been too dogmatic and inflexible.  “Recipes” have been given which work if followed to the letter, but little attention has been given to the underlying principles... I have therefore tried to emphasize the important variables which make for success or failure in the use of antibodies... I have also tried to point out areas in which the literature gives misleading impressions (Goding, 1983:3).

1. This wavering between a “disciplinary” and a “purely technical” status is reminiscent of the controversy surrounding the introduction of molecular biology into an Australian research institute (Stokes, 1983).  It may also be noted that this ambiguous status is reflected in the fuzzy institutional arrangements surrounding the introduction of hybridoma technology within academic and industrial research centers (Mackenzie, Cambrosio, and Keating, forthcoming).

251

Like the protocols produced for restricted diffusion, manuals represent a form of packaging and, consequently, a form of standardization such as that described by Fujimura (1987).  Presumably, the proliferation of such manuals will result in the reduction of the importance of “golden hands” in the production process.  However, as we have seen, not only are there different degrees of packaging but the relevance of a “standard technique” varies according to a number of factors such as the discipline within which it is to be used.

It should finally be noted that articles bearing on specific features of the technique appear regularly in journals such as the Journal of Immunological Method.  Often, these articles offer improvements in the technique.  However, the status of these improvements is somewhat ambiguous for, as several researchers pointed out, such advances in “protocol” are best confined to manuscript documents for internal use and do not, in themselves, represent much of interest in terms of scientific recognition.

 

Hybridoma Technique and Its Variants

Numerous differences exist between the written procedures used in different laboratories; further, the practical use of these instructions within the same laboratory varies.  This is a fact openly recognized in technical manuals:

There are a large number of fusion protocols in general circulation and most of them work... It has been emphasized throughout this book that the number of variations in procedure is immense (Campbell, 1984:127-31).

An adequate description of this variation, as we already noted, would necessarily call forth both “technical” factors (i.e., the biological problem at issue) and “social” factors (i.e., the previous experience of the researcher).  An individual’s choice of procedure depends, more often than not, on “social” factors relating to the origin of the protocol in question:

There are people who, if they see a protocol originating from an institution or an individual they don’t consider to be “prestigious,” will not try that protocol; it could be the world’s greatest protocol, but they won’t try it (an industrial researcher, winter 1986).

Variants of the hybridoma technique may be loosely classified in the following categories.  There are, first of all, “formal” variations that are duly noted in written protocols.  If limited to the central phase of the fusion procedure, these variations would include such factors as the concentration of the fusion agent, the temperature, the means of sterilization of P.E.G. (autoclave or filtration), and the precise duration of the fusion.  Some of the variants are subject to controversy within the manuals.  “Feeder cells” are a case in point.  Consisting of a cell preparation, they are often added to cultured hybridomas in order to promote their development.  Despite the term “feeder,” little is known about the mechanism of their action (Goding, 1983:71).  Some say they are useless while others declare them to be a major improvement of the technique (Fazekas Dc St.Groth, 1985:5).  Still others maintain that while their benefits may be unknown, they certainly don’t do any harm (Goding, 1983:71; Zola and Brocks, 1982:25).

There are, secondly, informal variants sometimes known as “shortcuts.”  For example, rather than count B cells and myelomas under the microscope in order to obtain a given proportion, say 10:1, it is possible to simply compare the volumes of two pellets of cells.  Given that myelomas are about 10 times larger than B cells, equal volumes of each gives the desired proportion.  The result of experience acquired in the course of many fusions, these “shortcuts” do not appear in written protocols, not even those for internal use, and are transmitted orally.

“Shortcuts” of the latter type may be distinguished from the use of “sophisticated” techniques (sometimes also referred to as “shortcuts”) which may replace certain parts of the procedure.  For example, rather than screening the hybridomas by searching for the required

252

antibody in the cell culture supernatant, it is possible, with the help of DNA probes, to determine the presence of the gene coding for the sought-after antibody.  Using such a technique allows one to screen a much larger number of hybridomas than would otherwise be possible (cf. Vincenti, 1984:562).

Within the private sector, “sophisticated” procedures of the type just mentioned are not often patented but are protected as “trade secrets” because to patent this form of intellectual property would result in diffusion of the knowledge without the ability to control its use.  However, if a “trade secret” remains entirely secret, it loses part of its commercial value.  Fully aware of this problem, a company like Hybritech advertises a “secret” system allowing its employees to rapidly screen large number of hybridomas.

Finally, procedural elements that are perceived of as the most immediate expression of local idiosyncrasies are dismissed as “magic” in opposition to practical scientific know-how.  The use of similar terms has been noted by sociologists of science.  Lynch (1985:108-11) has documented instances of “superstition,” and Fujimura (1987) of “black magic.”  According to one of our informants:

I consider that the actual fusion has a lot of voodoo.  There is a lot of things people do, they don’t know why.  I don’t know why but I just copy what they do and they say: “if you do it differently, it will not work.”  They told me I had to spin the fusing cells with the top open.  Why the top open?  It doesn’t make any difference, this is a small, desk top centrifuge, it doesn’t matter whether the top is open or closed.  I think the history of it is [laughter] that you can’t regulate the speed that well, so initially when people used to do it, they would open the top and see how fast it would spin.  Now people know how to regulate the speed and they don’t really have to look at it anymore, but they leave the top open!  They told me I had to leave the top open, I am supposed to be a scientist, I don’t believe the top has to be open, but I am not going to put it down, because if the fusion did not work, they would tell me it’s because I left the top down (an academic researcher, fall 1985).

Some “magic” is written down while other “magic” is transmitted through personal contact and thus circulates in the same manner as “tricks of the trade.”  Still other “magic” circulates under the guise of reason: if a researcher is forced to spend 10 minutes on the telephone in the course of an experiment and if that experiment is successful, then it is possible that the subsequent protocol will contain the instruction “leave for 10 minutes.”  “Magic,” in turn, may hide experience.  Such is sometimes the case with the optimization of experimental conditions:

We had to do experiments - with different kinds of feeder cells, with different kind of serum - that you never write out but allow you to go some place later and to become “a wizard.”  At the end, you end up with a kind of “voodoo ceremony” with all these experiments you know make a difference, but you don’t know why (an academic researcher, winter 1986).

 

The Problem of Standardization

Despite the fact that it is hardly possible to test all variables and their interactions, technical manuals and laboratory discourse often propose the future reduction of experimental incertitude as much as possible.  However, it is more often the case that methods “become established as soon as they happen to work at all” (Fazekas Dc St.Groth and Scheidegger, 1980:1).  The decision not to standardize is based on a variety of considerations.

First of all, scientists argue that, to the extent that the technique “works,” it is not worth the effort to clarify details that have no direct relation to research objectives.  This distinction between technique as an end in itself and technique as tool often refers, in turn, to prior social distinctions between university and industry and between researchers and technicians.  For example, university scientists may claim that in industry, where routine work is carried out

253

by technicians, the researcher in charge will surely have written up protocols allowing a mechanical reproduction of the technique.  As noted in a technical manual:

Experiments to study all these variables are tedious and relatively uninteresting at a time when investigators are anxious to produce some useful antibodies, irrespective of the efficiency of the process.  Thus, it is not surprising that successful procedures become entrenched, and that dogmatic statements about technical variables are accepted unchallenged.  As the initial excitement wears off, it is to be expected that much work will be done on technical aspects and that the procedures will lose much of their empiricism and mysticism (Zola and Brocks, 1982:4-5).

The prediction advanced in the last part of the paragraph is far from being true for all researchers.  While the move from art to science is indeed possible, so is the reverse:

You can make different parts scientific.  When we do experiments and record what the results were, in a sense we are making things scientific.  But sooner or later, it goes back to art: we know the technique that works, therefore we do it, even though the next generation has never tested it: so it becomes an art.  The value of making it science is not necessarily high, the value of making it work is high (an academic researcher, winter 1986).

Concern over recognition within the network of scientific relations may also militate against the pursuit and publication of technical improvements seen as “minor” or “trivial”:

If you want to publish a technique, you do it in Journal of Immunological Methods or Hybridoma, but they are not prestigious journals, many scientists don’t even get them; it’s not worth your time to write papers for them, so you don’t.  If people want to know your technique, they just call you up, and if they have problems, they send somebody to see how you do it (an academic researcher, winter 1986).

Moreover, optimizing experimental conditions (“experimenting around”) is seen as work suited mainly for doctoral students: “Post docs have to start immediately to write papers, they don’t have the time  (Interview with an academic researcher, winter 1986).  Nonetheless, these perceptions are not universal and vary according to the prestige and centrality of the laboratory.

Disciplinary formation often determines attitudes toward technique.  Some researchers once claimed that if P.E.G. is purified by recristallization, it would lose its fusogenic qualities, which apparently reside in the impurities contained in the commercial lots and not in the P.E.G. itself.  Other researchers argued otherwise (Art to Science in Tissue Culture, 1983).  The researcher who mentioned this controversy to us pointed out that while he was attempting to perfect his own technique, he had chosen to ignore this particular problem and the line of research it suggested.  He explained: “I’m a cell culturist and not a chemist; that’s how you chose.”

Artisanal elements of laboratory work are often held to constitute the “style” of a laboratory, which can be seen as either an obstacle to standardization or as yet another form of it.  Researchers in two different laboratories within the same university distinguished between the two labs on the basis of their styles of equipment procurement and maintenance.  One lab was characterized as possessing “Gucci” equipment (“You know - like Gucci leather”), while the other maintained less fancy apparatus:

Spencer’s lab is known for what they call their “Spencer grade”: it’s low-tech, but it always works.  If they have a broken piece of equipment, they say “it’s Spencer grade”; the way I interpret it is that their equipment is all very well used, they have nothing fancy, but everything is functioning, and it is an excellent place to learn hybridoma technology (an academic researcher, fall 1985; Spencer is a pseudonym).

The decision to use more rudimentary equipment allows researchers in Spencer’s lab to display artisanal dexterity foreign to the “Gucci” lab: “When ‘plating out’ the fused cells I am of the old school, I use a pipette with the finger on it; some people use multiple pipettors, I

254

don’t trust them” (an academic researcher, winter 1986).  It has also allowed members of the laboratory to “trade-mark” their style (see also Traweek, 1984):

We have T-shirts: “100% Spencer grade.”  The philosophy behind this is “do as much as you can with as little as you can,” use old instruments etcetera, that is “Spencer grade” equipment (an academic researcher, winter 1986).

It may be argued that characterizing differences between laboratories, such as the choice of myeloma, in terms of style, itself conceived as a sort of epiphenomenon with regard to research, tends to eliminate the problem of the existence of different and sometimes contradictory prescriptions of the hybridoma technique.  However, questions of style involve both problems of content and form (Goodman, 1978:23).  It is, therefore, worth asking how the researchers themselves account for variations in hybridoma technique.

Some researchers divide protocols into the “reproducible” and “effective” and those which must be assumed to be idiosyncratic and futile.  According to French et al. (1986:345), for example, “A number of fusion protocols use polyethyleneglycol to promote fusion.  We have found the protocol described by Fazekas de St.Groth to be reproducible and effective.”  However, even from a researcher’s point of view, such a solution is not entirely satisfactory since it tends to remove the recognized role played by artisanal elements.

Other researchers prefer thus to distinguish between “rigid protocols” and “minimal type protocols”:

I have “minimal type” protocols.  Hybridoma technology is mostly a question of instinct and experience.  I look at cultures by their color, I don’t do cell counts.  I’m not a very rigid type, but I do know people who have very rigid protocols as far as feeding, splitting, and manipulation of cells goes.  I believe rigid protocols are overdone (an industrial researcher, winter 1986).

In general, however, both scientists interviewed and texts surveyed tend to distinguish between important and accessory parts of research protocols.  Important steps refer to those which are supposed to have a direct and determinant effect on the experimental results.  As such, they are subject to careful experimentation with regard to the parameters they entail.  Accessory steps exercise only a secondary influence on the hoped-for outcome and are therefore considered to be subject to the idiosyncrasies of the laboratory concerned.  A similar distinction is also introduced with regard to the elements whose presence or absence defines the “rigidity” of a given protocol.  The following quote deals with the possibility of including within written protocols instructions concerning the transfer of cells:

We do that in some cases.  The indicator cell line for the assay of T-cells has to be passed every two days, given a supplement every two days: people who tried to stretch that ran into problems.  For other cell lines, I haven’t written it down; some you write down, some you don’t: every cell line is different, so writing it down doesn’t mean a lot.  But for indicator cells it is important.  If necessary, we do it; most of the time it’s not necessary.  Some people are better at that sort of thing than other people.  Depends on how good a farmer you are” (an academic researcher, winter 1986).

When questioned about the lack of consensus as to the criteria for “important” and “accessory” and the fact that laboratories using protocols differing in important steps still manage to produce hybridomas, researchers have recourse to finer distinctions.  Hybridomas themselves are divided into “ready-to-wear” (i.e, directed towards extremely immunogenic antigens and thus, by definition, easy to produce) and “sophisticated.”  Laboratories are divided into “advanced” (“We are a year and a half ahead of everybody else.”) and “ordinary,” or industrial and academic, or distinguished according to disciplinary affiliation wherein the technique has central or secondary meaning for the problematic.  As can be seen, the understanding of research protocols by scientists implies a distinction between the “technical” and the “social” which forms a part - call it a sociology - of the researchers’ practical reasoning.

255

 

From Art to Science

The terms “science,” “art,” and “magic” have so far been used to describe different parts of the hybridoma technique.  In this final section, we examine how these categories have been applied to hybridoma technology as a whole.

“From art to science” is a common expression in areas such as immunology and cell culture (for an example from engineering, see also Schön, 1983).  It also appears regularly in the various essay reviews and technical manuals devoted to monoclonal antibodies.  In spite of what has been said so far concerning the artisanal and sometimes “magical” character of hybridoma technology, this technology is often presented as a decisive step forward in the move from art and/or “magic” to science in the antibody domain:

Prior to 1975, the production of antibodies was considered by some to be a black art practised by immunologists... The uncertainties about the specificity of individual antisera led to many prolonged and acrimonious debates.  All that has now changed (Goding, 1983:1-3).  Serology involving conventional polyclonal antibodies used to be an art bordering on science, and immunologists could be divided into those who believed in immunochemistry and those who believed in “immunomagic.”  While the latter school will always be with us, the discovery of hybridoma antibodies has done much to put serology on a firm scientific basis (Goding, 1983:40).

The theme of a move from art to science is not restricted to university researchers.  It surfaces regularly in the advertising brochures of companies specialized in the sale of scientific equipment and reagents.  A technical bulletin distributed by the HyClone company, purveyors of the “Taggart Hybridoma Technology” mentioned above, bears the title Art To Science in Tissue Culture . The Invitron company publicizes its cell culture products proclaiming that “The Art of cell culture has passed away... The Science of cell manufacturing has arrived.”  According to Invitron, the art of cell culture was characterized by the use of “Ascites Fluid” and “Esoteric Protocols,” whereas the new “Science” requires the use of “Computerized Automation” and “Bioengineenng.”

There are, however, different roads from art to science. In Invitron’s case, it would appear that the route is entirely technical.  Not only is the transformation accomplished through sophisticated technical equipment, but the vision of science advanced is one of a series of technical operations largely devoid of conceptual content.  University researchers, on the other hand, tend to see the transformation conducted at the level of the concepts themselves.  Arguing that the production of monoclonal antibodies has made many artisanal aspects of immunology scientific, Goding claims that:

The old uncertainties of specificity and reproducibility have been replaced by the promise of unlimited supplies of standardized, monospecific antibodies.  Terms like “titre” and “avidity” have become virtually obsolete.  We can now talk about mass and affinity of antibody in a very precise way (Goding, 1983:40).

The degree to which an experimental practice is perceived as artisanal or scientific also depends upon the perceptions and self-perceptions of the discipline in question.  The classification of a technique as either art or science, for example, not only determines how a given technique will be circulated, but is also often an expression of a hierarchy among the laboratories involved in the circulation of the technique.  A researcher told us that previously it had been necessary to visit another laboratory in order to learn the hybridoma technique as applied to T-cells.  Now, however, his own laboratory had become a second possible port of entry into the domain.  The classification of the technique may, in addition, serve as a means of promoting technical changes as “improvements” or devalorizing changes as idiosyncratic “variations” of little scientific impact.  In such cases the application of the label may serve to establish or disrupt scales of credibility, which separate researchers.

Disciplines, too, may be implicated in the process of classification.  As the quotation at the

256

beginning of this section suggests, immunology has often been taxed with having indulged in “immunomagic.”  This often occurs when researchers trained in supposedly “harder” disciplines such as biochemistry or molecular biology reflect on what appears to them to be the more arcane or esoteric procedures in immunology.  Having decided to devote himself to the study of immunology, a successful biochemist we interviewed found that his former colleagues viewed his newfound interest with skepticism bordering on hostility.

This conflict is reflected within immunology, as Goding (1983:40) would have it, by those “who believe in immunochemistry and those who believe in immunomagic.”  Here, however, the dispute is exacerbated by the fact that the relations between science, art, and magic form the basis of the opposition between two schools of thought in modern immunology, the “system” immunologists and the “step-by-step” immunologists.  The latter school would be represented by researchers who restrict themselves to the sequential solution of problems dealing with a restricted number of variables.  The “system” school, on the other hand, would be represented by those concerned with providing somewhat indirect solutions to very general problems such as the causes of cancer.  Because of the opposing views of the relations between theory and practice, evaluation of research becomes especially problematic:

If there are two ways of looking at immunology, what do you do if you have a system immunologist who is a lousy scientist and you cannot verify easily his results and he did a bad experiment?  To account for his results, he usually invents a complicated theory... Some of these people who are bad scientists and who build a house of cards can become very prominent.  In immunology, bad scientists cannot be easily detected (an academic researcher, fall 1985).

In such cases, judgements concerning the reliability of a researcher’s results can only be based on an assessment of the researcher’s long-term performance; this, in turn, raises the question of the “experimenter’s regress” as described earlier.

In the field of industrial production, the classification of experimental procedures as art, science, or magic is determined in part by the demands of commercial success as well as the guidelines of the various regulatory bodies governing activity in a given industrial sector.  Comparing the relative merits of two recent techniques, DNA probes and monoclonal antibodies, on the basis of their possible use in commercial diagnostic kits, one author noted:

Anything which involves a great deal of “art” or extreme complexity will be relatively disadvantaged.  Art particularly is anathema to rational production decisions and to the regulatory and supervisory mechanisms in the health care industry of most countries (Nash, 1985).

The standardization of procedures using monoclonal antibodies has been described in the following mixed metaphor: “a jungle full of pitfalls” (Haaijman et al., 1984).

From the point of view of the university researcher, the perception of his practice of the hybridoma technique as art has certain advantages with regard to his relation with industry.  First of all, it allows the researcher to distance himself from industry by projecting the routine aspects of hybridoma technology as industrial practices.  At the same time, it allows the researcher to reaffirm a relationship of mutual dependence between university and industry in the domain of biotechnology.  Given the fluid nature of the boundaries between industrial and academic institutions in this area (several biotech start-ups stress the quasi-academic climate of the firm), it is no surprise to find that the opposition between art and routine is also used within the industrial sector to distinguish one firm from another or different departments within the same firm.

Finally, the classification of knowledge has many consequences for patenting practices.  It may be argued that the passage from “magic” to “science” in the area of antibody production has opened the possibility of patenting an antibody against a given antigen.  Questions relating to the status and the mode of circulation of patentable knowledge have played a central role in the court battles that have engaged two pioneers in the commercial exploita-

257

tion of hybridoma technology, Hybritech and Monoclonal Antibodies Inc. (Mackenzie, Cambrosio and Keating, 1987).

 

Discussion

In this paper we have examined how researchers using hybridoma technology classify their knowledge and their activities using the categories of “science,” “art,” and “magic.”  We have seen that in the establishment and diffusion of a scientific technique, which may be conceived of as an embedded system of practices, scientists have recourse to many forms of knowledge.  That part which may be considered tacit or local depends upon the network of relations within which the scientists work.  This network is comprised of a system of heterogeneous elements (theories, machines, patents, products).  The articulation of these diverse elements occasions the emergence of the scientists’ categories of knowledge.

As noted in the introduction, recent work in the sociology of science has attempted to construct similar classifications of scientific knowledge and practice using such categories as objective, declarative, procedural, tacit, and local knowledge.  Despite the fact that the sociological classifications are presumably the result of a concerted effort of reflection and analysis, they are not, as we have seen, fundamentally different from the supposedly naive, ad hoc typologies of the scientists.  For although the sociological categories presume to describe dimensions of science overlooked by or invisible to scientists, they invariably make use of the scientists’ categories to achieve this end.  It is true that the use of these categories by scientists is not always consistent and that the boundaries between categories vary among institutions, practices and interests.  But the same may be said of the sociologists’ categories.  We have attempted an empirical demonstration of the inadequacy of these categories, considered fundamental by sociologists, and of the need to relativise them.  In particular, we believe that the existence of “immunomagic” (i.e., that knowledge and know-how that scientists have agreed to drop from discussion for a given period of time) shows that these categories are unable to account for the strategies employed by scientists.

As we have seen, while scientists often present ideal, algorithmic accounts of their work, they also recognize and work with tacit or local dimensions of knowledge whether they be classified as “art” or “magic.”  In many respects, the scientists’ own descriptions of the kinds of knowledge with which they deal on a daily basis are both more precise and more comprehensive than the descriptions offered by sociologists.  Not only are the scientists capable of describing the choices open to individuals and institutions, but they also recognize the many ways “tacit” and “local” knowledge, contrary to the sociological definitions of these terms, circulate among different scientific and technical cultures.  Indeed, contrary to what some sociologists of science argue, a common culture is not a prerequisite for the emergence of a scientific network.

 

References

Art To Science In Tissue Culture 1983 “Cell fusion with P.E.G.: is it reproducible?” Art To Science In Tissue Culture 2:1-2.

Callon, Michel and Bruno Latour 1986 “Les paradoxes de la modernité: comment concevoir les innovation?” Prospective et Sante Hiver:13-25.

Campbell, Ailsa M. 1984 Monoclonal Antibody Technology: The Production and Characterization of Rodent and Human Hybridomas. Amsterdam: Elsevier.

258

Collins, Harry M. 1985 Changing Order: Replication and Induction in Scientific Practice. London: Sage.

Collins, Harry M. 1987 “Expert-systems and the science of knowledge.” pp. 329-48 in Wiebe F. Bijker, Thomas P. Hughes, and Trevor Pinch (eds.), The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. Cambridge, MA: MIT Press.

Collins, Harry M., R.H. Green, and R.C. Draper 1985 “Where’s the expertise?: expert systems as a medium of knowledge transfer.” pp. 323-34 in Martin Merry (ed.), Expert Systems 85. Cambridge: Cambridge University Press.

Descombes, Vincent 1977 L’inconscient malgré lui. Paris: Editions de Minuit.

Eisenbarth, G.S. 1985 “Monoclonal antibodies.” Pp. 31-40 in Murray Moo-Young (ed.), Comprehensive Biotechnology, Vol.4. The Practice of Biotechnology: Specialty Products and Service Activities. Oxford: Pergamon Press.

Eshhar, Zelig 1985 “Monoclonal antibody strategy and techniques.” pp. 341 in Timothy A. Springer (ed.), Hybridoma Technology in the Biosciences and Medicine. New York: Plenum Press.

Fazekas De St. Groth, Stephen 1985 “Monoclonal antibody production: principles and practice.” pp. 1-10 in Soldano Ferrone and Manfred P. Dierich (eds.), Handbook of Monoclonal Antibodies: Applications in Biology and Medicine. Park Ridge, NJ: Noyes Publications.

Fazekas De St. Groth, Stephen and Doris Scheidegger 1980 “Production of monoclonal antibodies: strategy and tactics.” Journal of Immunological Methods 35:1-21.

Ferguson, Eugene S. 1985 “La fondation des machines modernes: des dessins.” Culture Technique 14:183-207.

French, Deborah, Ellyn Fischberg, Susan Buhl, and Matthew D. Scharif 1986 “The production of more useful monoclonal antibodies.” Immunology Today 7:344-46.

Fujimura, Joan H. 1987 “Constructing ‘do-able’ problems in cancer research: articulating alignment.” Social Studies of Science 17:257-93.

Goding. James W. 1983 Monoclonal Antibodies: Principles and Practices. London: Academic Press.

Goodman, Nelson 1978 Ways of Worldmaking. Indianapolis: Hackett Publishing Company.

Haaijman, J.J., C. Deen, CJ.M. Kröse, JJ. Zijlstra, J. Coolen, and J. Radl 1984 “A jungle full of pitfalls.” Immunology Today 5:56-58.

Hybridoma Techniques 1980 EMBO, SKMB Course, Basel. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Jones, Bryn 1983 “Division of labour and distribution of tacit knowledge in the automation of metal machining.” Pp. 19-22 in T. Martin (ed.), Design of Work in Automated Manufacturing Systems. New York: Pergamon Press.

Jones, Bryn and Stephen Wood 1984 “Qualifications tacites, division du travail et nouvelles technologies.” Sociologie du Travail 4:407-21.

Knorr-Cetina, Karin D. 1981 The Manufacture of Knowledge. New York: Pergamon Press.

Knorr.Cetina, Karin D. and Michael Mulkay, eds. 1983 Science Observed: Perspectives on the Social Study of Science. London: Sage.

Kusterer, Kenneth C. 1979 Know-How on the Job: The Important Working Knowledge of “Unskilled” Workers. Boulder: Westview Press.

Latour, Bruno and Steve Woolgar 1979 Laboratory Life: The Social Construction of Facts. Beverly Hills, CA: Sage.

259

Lawrence, Christopher 1985 “Incommunicable knowledge: science, technology and the clinical an in Britain 1850-1914.” Journal of Contemporary History 20:503-20.

Lynch, Michael 1985 Art and Artifact in Laboratory Science. London: Routledge & Kegan Paul.

Lynch, Michael 1982 “Technical Work and Critical Inquiry: Investigations in a Scientific Laboratory.” Social Studies of Science 12:499-533.

Mackenzie, Michael, Alberto Cambrosio, and Peter Keating forthcoming “The commercial application of a scientific discovery: the case of the hybridoma technique.” Research Policy.

Mackenzie, Michael, Alberto Cambrosio, and Peter Keating 1987 “Scientific information and intellectual property: the case of the hybndoma/monoclonal antibody technique.” Paper presented at the 1987 meeting of the Society for Social Studies of Science, Worcester, MA.

Nash, Martin 1985 “On the emergence of commercial products from biotechnology.” Trends in Biotechnology 3:219-22.

Pinch, Trevor 1986 Confronting Nature. Dordrecht, Netherlands: D. Reidel.

Schön, Donald A. 1983 The Reflective Practitioner. New York: Basic Books.

Star, Susan Leigh 1983 “Simplification in scientific work: an example from neuroscience research.” Social Studies of Science 13:205-28.

Stokes, Terry D. 1985 “The role of molecular biology in an immunological institute.” Paper presented at the International Congress of History of Science, University of California, Berkeley.

Traweek, Sharon 1984 “Nature in the age of its mechanical reproduction: the reproduction of nature and physicists in the high-energy physics community.” pp. 94-112 in Claire Belisle and Bernard Schiele (cdi.), Les savoirs dans les pratiques quotidiennes. Recherches sur les representations. Paris: Editions du C.N.R.S.

Vincenti, Walter 1984 “Technological knowledge without science: the innovation of flush riveting in American airplanes, ca.1930-1950.” Technology & Culture 25:540-76.

Westerwoudt, Regine J. 1985 “Improved fusion methods: IV. technical aspects.” Journal of Immunological Methods 77: 181-96.

Zola, H. and D. Brocks 1982 “Techniques for the production and characterization of monoclonal hybndoma antibodies.” pp. 1-57 in John G.R. Hurrell (ed.), Monoclonal Hybridoma Antibodies: Techniques and Applications. Boca Raton: CRC Press.

260