Science Studies and Constructivism

1. Introduction

Science studies and constructivism is much, too much to address in one talk. Therefore, let me first say that I will not be dealing with science studies as such (including for example quantitative research or directly policy oriented one), but only with those studies, which can be seen in the tradition of the "Social Studies of Science". An approach, which also has been labelled "social constructivist approach", "sociology of scientific knowledge" or "new sociology of science" by its proponents (the latter term, roughly twenty years after the beginning, is sometimes still to be found).[2] Despite all differences -to make a long story short- the "Social Studies of Science" shared an ambitious and coherent perspective, namely to relate the knowledge content of science to social factors.

This seems to belong to another era, since the whole approach to science studies has been transformed significantly. It has been replaced by a heterogeneous network, ranging from philosophically uninformed debates about the problem of "reflexivity" to dozens of case studies in the emerging field of technology studies.

The more interesting and innovate contributions, however, are those, which assume a very critical perspective of the entire research programme and try to develop new directions in science studies. These developments can be identified in the recent emphasis on scientific practice and in the actor-network approach.

Both represent a sharp criticism of an overly strong emphasis on social aspects in science studies. The remedies are at hand: Pickering, for example, after having deleted the K of SSK (Sociology of Scientific Knowledge), "since the central topic is practice not knowledge", does not lose time "to delete (...) the first S, since there seems no warrant for assigning causal priority to the social in understanding scientific practice and culture."[3] Likewise, Latour and others who have developed the actor-network approach, employ that kind of anti-social rhetoric and, as a consequence, give up the distinction between social/nature and social/technology.

Also the other half of the old-fashioned label "social constructivism" - namely "constructivism" - has literally disappeared. (A term which has not only been overused, but which has also been used in a highly inconsistent way as Sismondo recently pointed out.)[4] Lynch, in his recent book, speaks of "crisis in relativist and constructivist studies."[5] Knorr-Cetina, in her reply to Sismondo states that "Constructivism (...) is not, however, a world view, or a life-time occupation" and she presumes that "there will be a time when it is more useful to take the questions and run from constructivism."[6] And, let me add the observation that "constructivism" is no longer to be found in most of the book indices.[7]

To sum up, it looks like time for a change within the "Social Studies of Science" has already occured. A change which started with the destruction of basic assumptions, regarded as out of time. This is not only claimed by the main protagonists but also by its more traditional critics within the field.

In this talk I will offer you a radically different interpretation. The shift towards practice and the actor-network approach are, despite their claimed newness, a direct consequence of methodological and theoretical assumptions taken much earlier. These can at least be traced back to laboratory studies conducted in the late 70s and early/ mid 80s. Continuation, not revolution should be the right term to characterize the whole development of the "Social Studies of Science". Or, to put it in Kuhnian terms, what we can observe is not a paradigm shift, but "normal science".

Therefore my argument is twofold. First, I will focus on recent, at first sight brandnew approaches to science studies, namely the shift towards practice and the actor-network approach, pointing out their inherent limitations. Second, I will turn to laboratory studies and focus on underlying, mostly implicit methodological and theoretical principles, which then are criticized. The discussion of both, new and old assumptions and implications are from a sociologist's point of view. Epistemological and political aspects will therefore be of minor importance, since this has been done elsewhere.[8]

In the end, since the main deficits of the recent developments and their antecedents are seen in an insufficient account of modern social theory, I will outline an alternative path. (One path among others, of course.) This remains highly provisional and is to be elaborated further.

2. Social Studies of Science - Recent Trends

2.1. The shift towards practice

One main objective is to focus on all dimensions of scientific practice. Therefore the traditional focus on the social dimensions seems to be too narrow. Entities which previously were not regarded as social and were therefore neglected are included in the analysis. See, for example, Ian Hacking: "I am not to argue for idealism but for a rather down-to-earth materialism. Mine is a thesis about the relationships between thoughts, acts, and manufactures."[9] The emphasis is now on practice: "scientific practice is interesting in its own right", to quote a statement from Andrew Pickering.[10] Therefore it becomes important to find a place where the reality of scientific practice can be traced - raw, unfiltered and without any analytical constraints. Laboratories fulfill this condition, because they allow the setting of relatively clear boundaries between inside and outside and hence prevent from an overload with too much unfiltered complexity of scientific practice.

The amount of equipment and technology, which is relevant in a laboratory setting, is now conceptualized as an actively shaping part of scientific research. This can be seen as a necessary correction of the social voluntaristic emphasis of the early constructivist studies. However, the problem is that equipment and technology are regarded as a direct cause for the production of scientific knowledge. Karin Knorr-Cetina, for example, emphasizes: "The production, measurement and description of the relevant particles takes place in an integrated way in the inside of the apparatus. The surrounding "laboratory" (in quotation marks) is the service station for the machine, which is the true laboratory."[11] In a similar vein, Andrew Pickering tries to attribute a constitutive role to the material world in science. Scientific results are then seen as inherently shaped by what he calls "maneuvers in the field of material agency".[12]

The shortcomings of this approach are at hand. From this perspective, the scientists' cognitive and social inputs are devalued through the conceptual equation with other components. But, first, they are necessary to make an apparatus work before it becomes part of the laboratory equipment. Likewise, knowledge-production in an observed laboratory always remains a cognitive and social process, even if especially the routinized sequences are heavily based on apparata.

2.2. The actor-network approach

Far beyond the theoretically less ambitious shift towards scientific practice, a second emerging trend can be identified in the actor-network approach of the Paris School.[13] Bruno Latour, who without doubt is its most prominent representative, aims at blurring the boundaries of the -usually taken for granted- distinction between the social, on the one side, and technology and nature on the other side. These a priori distinctions have to be substituted, since a network is composed of different actors and their interactions, which all contribute to its strength or weakness.

This approach is not only theoretically ambitious. It also contains a strong ethical component. This new conception of an actively acting nature should be seen as a radical departure from our normal anthropocentrical world view; a world view also reflected in social theory and in the Social Studies of Science, which assign priority to the social. Latour claims: "If we succeed in this emancipation of the nonhumans from the double domination of society and science, it will be the finest result of that perhaps clumsily begun 'anthropology of the sciences'."[14] As a consequence, microbes are introduced as new social actors [15], scallops are understood as beings with ends [16], and even "reflexion" is transferred to nonhuman living beings and technical artefacts.[17]

On the one hand, unfortunately, the origins of these attributions, which lie in human behaviour, are blurred. On the other hand, the network is, in a plainly anti-constructivist way, treated as a kind of objective structure, and Latour himself as its legitimate and independent speaker, but not as its designer. As a result, not only the social, as an analytical category, becomes devalued. But also, nonhuman living beings become totally colonized. Paradoxically, the anti-anthropocentric attempt ends up in a radicalized anthropocentrism, since nonhumans are modelled in strict analogy to human beings.

Under the direction of Latour, who places himself out of his networks, the single components of an actor-network can be moved like chessmen on a chessboard. The emancipatory drive of Latour's approach basically serves to assert his own claims. This can be seen as a classical case of selfexemplification: the selfexemplification of a theory, which regards the enrollment of allies - here: the nonhuman living beings, enrolled by Latour - as the central mechanism for the assertion of scientific claims.

2.3. Summary of 2.

To sum up, both of the developments discussed here - the focus on the totality of the laboratory equipment (here used in a broader sense, including also scientists) and the theoretical radicalization in the actor-network approach - have a common objective. Namely, the complete representation of the interactions of single components. Through the analytically undifferentiated incorporation of social, technical and natural components, the distinction between social and not-social is given up. This leads not only to an overly "social" world view, since under this premise we can observe actions of microbes and direct interactions between men and machines. This "over-sociologization"[18] , as some critics call it, is only one side of the story.

From a sociological point of view it looks somewhat different. Because the term social is used in a purely metaphorical sense (without taking into account any of its characteristics!) these two concepts also represent an ignorance of the social. The way in which basic terms in the social sciences like action and interaction are treated, makes this point more obvious. In a completely desocialized definition of action, the relevant dimensions of intention and meaning on the one hand, normative and cultural aspects on the other hand, are sacrificed. As a consequence, scientists do research in the same way as the sun shines, planes fly, taps lose water and test-tubes break.

Likewise, the underlying concept of interaction is an extremely reductionist one. It merely concentrates on the output-dimension. Interaction, therefore, is seen as the sum of events or of sequences of events. Aspects which lie under the surface and which can be reconstructed only from an internal point of view, like the process-character of interaction and the mutual structures of meaning and expectation within the interaction, are sacrificed. As a consequence, scientists can interact with computers as well as a toaster can interact with a piece of toast.

3. Laboratory Studies as Trendsetter

The actual state-of-the-art in the Social Studies of Science, as described and criticized above, could be seen as a radical departure from previous "social constructivist" assumptions. I would - to repeat myself - rather stress the point that this state-of-the-art can be seen as a direct consequence of methodological and theoretical assumptions taken much earlier. These assumptions can be traced back to the most interesting and successful version of the so-called "new sociology of scientific knowledge": the laboratory studies of the late 70s and early/ mid 80s.

3.1. Methodological assumptions

Laboratory studies have traditionally been criticized from two different angles: First, from a philosophical point of view; second, from others within the wider range of the Social Studies of Science. Philosophers of science vehemently criticized a - from their point of view - overly constructivist approach towards scientific knowledge.[19] Sociologists of science argued that the focus on scientific laboratories is too narrow, leaving out of perspective other social phenomena in science, for example, those on an organisational or disciplinary level, and also those social phenomena between science and other parts of society.[20] Unfortunately, these critiques - too much constructivism and a too selective focus - do not concentrate on what has actually been done in laboratory studies. If one starts from this angle, then the opposite seems to be true: first, a lack of constructivist scepticism and, second, a lack of selectivity in the analysis.

First, laboratory studies pursue a naturalist ideal in respect to their own findings which stands in sharp contrast to constructivist claims. The attempt to document a given and quasi ontological reality reveals a surprisingly unreflective understanding of science, as if laboratory studies were an enterprise to discover some objective truths. (This quasi ontological reality is of course the one of laboratory life, while the reality the observed scientists refer to is being deconstructed.) This very obvious contradiction has been admitted only recently. Lynch, for example, states in his new book that "The earlier laboratory studies used empiricist claims that did not withstand 'reflexive' scrutiny."[21] This anti-constructivist practice - not only in the laboratory studies, but in most of the "constructivist" science studies - explains the emerging debate about "reflexivity", which has to compensate this lack of constructivist scepticism but which in itself is not very far-reaching.

Second, since laboratory studies pursue that naturalist ideal, the comprehensive description of the "infinitesimal details"[22] of laboratory life cannot be done in a selective way, which one allows to ask for more specific questions, for example those relating to scientific change. This ideal implies that a maximum of proximity to the research object is strived for. Additionally, all methodological doubts concerning the problem of "going native" in anthropology are ignored.[23] This raises not only the question of comparability of the research results, but also the question - which seems to me even more serious -, why an unselective representation of scientific practice should not be substituted by the original. Why, and this argument likewise goes against the recent shift towards scientific practice, should social scientists investigate in scientific research, if they have no distinct research questions; why the natural scientists should not do it by themselves?

3.2. Theoretical Assumptions

The theoretical assumptions of laboratory studies represent the second pillar of the recent trends within the Social Studies of Science, namely the shift towards scientific practice and the actor-network approach. Instead of openly stated assumptions, one finds a wide range of metaphors by which the scientists' behaviour is described. This is partly due to the methodological device to give up theory. Knorr-Cetina, for example claims that "constructivism (...) requires nothing less than to keep the analysis sufficiently free from theory".[24] This approach to reality is epistemologically naive and collides with the constructivist notion of the theory-ladenness of any kind of scientific observation. (We find here a striking analogy to the way "constructivism" is practiced under methodological aspects.)

Getting closer to the underlying metaphors, one sees that science becomes, partly with explicit reference to Hobbes and Machiavelli[25], a kind of political arena, in which strategic actors are fighting a relentless fight for improving their position in the field; or that science becomes a perfect economic market, where scientists compete for scare resources[26]; or, alternatively in Latours more recent works, war is regarded as the adequate metaphor for description. Unfortunately, these metaphors not only miss the specific characteristics of the scientific field, but also of those spheres, from which these de-mystifiying descriptions of science are borrowed - especially politics and economics. Or, for example, is there still any serious attempt to describe the logic of political life the way Hobbes and Machiavelli did?

This kind of a metaphorical and theoretically uncontrolled description of scientists' behaviour is accompanied by a highly problematic use of the term social. On the one side, we have the ethnomethodological approach by Knorr-Cetina and Lynch, who specify the term social through the term negotiation.[27] Negotiations between scientists is their way of describing the production of scientific knowledge. On the other side, we find in the social constructivist approach of Latour and Woolgar an oversimplified concept of scientists' behaviour, based on an inadequate concept of action.[ 28] Both, the ethnomethodological approach and the concept of Latour and Woolgar, deserve a more careful critique, also in this regard.

Let me start with the former. The problem with this attempt is that it starts with individual scientists, who construct an always extraordinary and contingent situation out of the blue. Lynch speaks of "the production of social order in situ".[29] The surrounding social context of action, on the other hand, is devalued. The scientific laboratory appears like an unstructered vacuum, without a past and without a future. Likewise, the implicit knowledge of scientists, which is always given in every single situation, has to be ignored. From this perspective, in which voluntaristically acting and interacting scientists take the place of always prestructured and time-depending contexts, one can hardly explain cognitive and social patterns within the laboratory context. Or, as Richard Whitley commented: "Without some 'rules of the game' no result or outcome could be produced."[30]

While the ethnomethodological approach by Knorr-Cetina and Lynch had at least an idea of the dynamics of interaction, we find in Latour and Woolgar an entirely desocialized concept of human action and interaction. Individual actors, who are chosen as the main unit of analysis, are conceptualized as strategically operating, voluntaristically acting and manipulating entities. This oversimplified and reductionist model of man allows for the observation of the scientists' actions and interactions as if they were automata, programmed to pursue their individual interest. The aggregate outcome of individual strategies then can be taken as a proof for what at least in the first edition of "Laboratory Life" is claimed as the social character of knowledge production. The link between these general assumptions about human behaviour and human interaction can be seen in the underlying concept of action. This is freed from all the relevant dimensions of actions mentioned in the beginning. To repeat: Neither the dimension of intention and meaning, nor normative and cultural aspects are taken into account.

At this point our journey into the past ends and we are back to recent trends in the "Social Studies of Science". This, because such a purified concept of action as employd by Latour and Woolgar, can be easily transferred to all kinds of entities. And, as a consequence, men, machines and microbes likewise become acting entities. All differences disappear, and everything becomes possible in the new, enriched world of science studies.

4. Concluding Remarks

In conclusion, I would like to provide an outline for an alternative approach to science studies. To repeat: The new trends in the "Social Studies of Science" are, despite their claimed newness, a direct consequence of methodological and theoretical assumptions taken much earlier. A naturalist ideal of description (as a methodological device) in combination with a purely metaphorical use of the term social (as a substitute for theory) are the basis for the new developments, which focus on directly visible phenomena and tend to delete the term social.

Of course, one could see the most interesting perspectives in an empirically broader research agenda than the actual one. It is obvious that science studies cannot be limited to the study of research practices in scientific laboratories. But since the deficits of the "Social Studies of Science" are basically theoretical ones, I would like to argue for an openness towards social theory.

This does not imply a further journey into the past, neither into the good old times of the Edinburgh-School, which set up the program of the sociology of scientific knowledge[31], nor, further steps backward, into the early days of the sociology of science, associated with Robert K. Merton's research program.[32] However, both programs run into similar difficulties: both fail to specify a transmission mechanism which links the broader macro structures with the micro structures of empirically observable actions of scientists. (With macro structures I am refering to the universal scientific norms in Merton and to the broader social interests in the concept of the "strong program".) Second, both approaches miss an adequate understanding of the societal contexts in which science always is embedded. Merton's studies, on the one hand, are too exclusively concentrated on the social structure of science and therefore neglect the mutual penetration of scientific and other societal contexts. On the other hand, the emphatic reference of the strong program to "society" also remains rather vague. Because of these difficulties, both paths (Merton and the "strong program") can be excluded.

Since the "Social Studies of Science" have become an increasingly inbred culture, and the link with modern social theory has been interrupted, it is worthwhile to look for theoretical advances outside the science studies community. This turns the attention towards Anthony Giddens' theory of structuration[33] and Niklas Luhmann's theory of functional differentiation.[34] Both approaches allow - in contrast to the older concepts mentioned before - a reformulation of the micro/macro-problem and the recontextualisation of science in society. However, since the basic interest of both approaches lies in a general theory of society, the direct application to science studies is of rather limited value, at least in the present situation.

Therefore the New Institutionalism seems to be more promising, since the fruitfulness of this approach for empirical research has already been proven.[35] Neo-institutionalist studies, until now, have been conducted in political[36], economic[37] and educational[38] contexts. They all show that none of their units of analysis can be explained through interest-based strategies of individual actors (or, as Latour would say nowadays: actants). As a consequence, one has to give up the idea that larger units are to be reconstructed from the bottom to the top - for example, social order within a scientific laboratory and in the wider scientific context. According to the New Institutionalism, economic structures are not the result of an indexical logic of production in the quasi laboratory of the factory or of the office of a firm; political life cannot be melted down to decisions taken within the core of political institutions; and education is not the sum of isolated units of instruction in the quasi laboratory of the classroom.

Neo-institutional analyses focus in particular on the relationship between the acquisition of resources and the maintenance of legitimacy and identity. They question, first, whether and how the analysed units manage to be recognized by their social environment and, second, what internal changes result from the interaction of these units with their environment.

If this approach is applied to science studies, one can ask, first, under what conditions science is a recognized institution within society, and therefore granted social legitimacy and external resources. It is not exaggerated to presume that the societal recognition of science implies the incorporation of external standards, especially political and economic ones. Second, one has to look for internal changes which can be seen as a consequence of these processes and which also affect traditional concepts of scientific identity.

Aspects of a changing science within a changing society can then be grasped at different levels of analysis. And under the premise that the Social Studies of Science open themselves towards modern social theory - as an alternative to the ongoing cultivation of inbred concepts - not only the reconstruction of larger networks, but also the observation of scientific laboratories could be an important contribution to science studies in general.

[1] This paper is part of a larger investigation in the theories and methods of the Social Studies of Science, conducted with Raimund Hasse and Peter Weingart. Two papers have already been published: Hasse, R/ Krücken, G/ Weingart, P, 1993, Situazione e prospettive del costruttivismo di laboratorio: una ricerca sociologica, Teoria Sociologica, 2, pp 309-336; Hasse, R/ Krücken, G/ Weingart, P, 1994, Laborkonstruktivismus. Eine wissenschaftssoziologische Reflexion, Rusch, G and Schmidt, S J, (eds) Konstruktivismus und Sozialtheorie, pp 220-262. An english paper will be submitted for publication soon.

[2] For example in Lynch, M, 1993, Scientific Practice and Ordinary Action. Ethnomethodology and Social Studies of Science, Cambridge: Cambridge University Press

[3] Pickering, A, 1992, From Science as Knowledge to Science as Practice, Pickering, A, (ed) Science as Practice and Culture, Chicago: Chicago University Press, pp 1- 29, p 14

[4] Sismondo, S, 1993, Some Social Constructions, Social Studies of Science, 23, pp 513-553

[5] Lynch, M, 1993, Scientific Practice and Ordinary Action. Ethnomethodology and Social Studies of Science, Cambridge: Cambridge University Press, p 103

[6] Knorr-Cetina, K, 1993, Strong Constructivism - from a Sociologist's Point of View: A Personal Addendum to Sismondo's Paper, Social Studies of Science, 23, pp 555-563, p 561- 62

[7] Take for example Pickering, A, ed, 1992, Science as Practice and Culture, Chicago: Chicago University Press or Latour, B, 1987, Science in Action. How to Follow Scientists and Engineers through Society, Cambridge, Mass.: Harvard University Press.

[8] See Bunge, M, 1992, A Critical Examination of the New Sociology of Science. Part II, Philosophy of the Social Sciences, 22, pp 46- 76 for the former, and for the latter aspect Collins, H M/ Yearley, S, 1992, Epistemological Chicken, Pickering, A, (ed) Science as Practice and Culture, Chicago: Chicago University Press, pp 301- 326 and also Collins, H M/ Yearley, S, 1992, Journey into Space, Pickering, A, (ed) Science as Practice and Culture, Chicago: Chicago University Press, pp 369- 389

[9] Hacking, I, 1992, The Self- Vindication of the Laboratory Sciences, Pickering, A, (ed) Science as Practice and Culture, Chicago: Chicago University Press, pp 29-65, p 30

[10] Pickering, A, 1992, From Science as Knowledge to Science as Practice, Pickering, A, (ed) Science as Practice and Culture, Chicago: Chicago University Press, p 6

[11] Knorr- Cetina, K, 1988, Laboratorien: Instrumente der Weltkonstruktion, Hoyningen-Huene, P/ Hirsch, G, (eds) Wozu Wissenschaftsphilosophie? Positionen und Fragen in gegenwärtiger Wissenschaftsphilosophie, Berlin: de Gruyter, pp 315-344, pp 328-29

[12] Pickering, A, 1993, The Mangle of Practice: Agency and Emergence in the Sociology of Science, American Journal of Sociology, 99, pp 559-589, p 581

[13] Callon, M, 1986, The Sociology of an Actor- Network: The Case of the Electric Vehicle, Callon, M/ Law, J/ Rip, A, (eds) Mapping the Dynamics of Science and Technology: Sociology of Science in the Real World, London: Macmillan, pp 19- 34; Latour, B, 1987, Science in Action. How to Follow Scientists and Engineers through Society, Cambridge, Mass.: Harvard University Press; Latour, B, 1988, The Pasteurization of France, Cambridge, Mass.: Harvard University Press; Latour, B, 1988, Mixing Humans and Nonhumans Together: The Sociology of a Door-Opener, Social Problems, 35, pp 298-310 and Latour, B, 1990, Drawing Things Together, Lynch, M/ Woolgar, S, (eds) Representation in Scientific Practice, Cambridge, Mass.: MIT Press, pp 19-68

[14] Latour, B, 1988, The Pasteurization of France, Cambridge, Mass.: Harvard University Press, p 150

[15] Latour, B, 1988, The Pasteurization of France, Cambridge, Mass.: Harvard University Press, p 35

[16] Latour, B, 1988, The Pasteurization of France, Cambridge, Mass.: Harvard University Press, p 167

[17] Latour, B, 1988, The Politics of Explanation: an Alternative, Woolgar, S, (ed) Knowledge and Reflexivity, London: SAGE, p 173

[18] Leydesdorff, L, 1992, The Knowledge Content of Science and the Sociology of Science, Journal for General Philosophy of Science, 23, pp 241-263, p 244

[19] For a recent summary see Bunge, M, 1991, A Critical Examination of the New Sociology of Science. Part I, Philosophy of the Social Sciences, 21, pp 524- 560 and Bunge, M, 1992, A Critical Examination of the New Sociology of Science. Part II, Philosophy of the Social Sciences, 22, pp. 46- 76

[20] See for example Nowotny, H, 1984, Leben im Labor und Draußen: Wissenschaft ohne Wissen? Anmerkungen zu neueren Ansätzen innerhalb der Wissenschaftssoziologie, Soziale Welt, 33, pp 208- 220 or Cozzens, S E/ Gieryn, T F, 1990, Introduction: Putting Science back in Society, Cozzens, S E/ Gieryn, T F, (eds) Theories of Science in Society, Bloomington/Indianapolis: Indiana University Press, pp 1- 14

[21] Lynch, M, 1993, Scientific Practice and Ordinary Action. Ethnomethodology and Social Studies of Science, Cambridge: Cambridge University Press, p 104

[22] Knorr- Cetina, K, 1988, Das naturwissenschaftliche Labor als Ort der 'Verdichtung' von Gesellschaft, Zeitschrift für Soziologie, 17, pp 85-101, p 87

[23] See for example the introduction to Latour, B/ Woolgar, S, 1979, Laboratory Life. The Social Construction of Scientific Facts, Beverly Hills: SAGE

[24] Knorr-Cetina, K, 1989, Spielarten des Konstruktivismus. Einige Notizen und Anmerkungen, Soziale Welt, 40, pp 86-96, p 92

[25] See Callon, M, 1986, The Sociology of an Actor- Network: The Case of the Electric Vehicle, Callon, M/ Law, J/ Rip, A, (eds) Mapping the Dynamics of Science and Technology: Sociology of Science in the Real World, London: Macmillan, pp 19- 34

[26] Latour, B/ Woolgar, S, 1979, Laboratory Life. The Social Construction of Scientific Facts, Beverly Hills: SAGE

[27] Knorr- Cetina, K, The Manufacture of Knowledge. An Essay on the Constructivist and Contextual Nature of Science, Oxford: Pergamon Press; Lynch, M, 1985; Art and Artefact in Laboratory Science. A Study of Shop Work and Shop Talk in a Research Laboratory, London: Routledge and Kegan Paul

[28] See Latour, B/ Woolgar, S, 1979, Laboratory Life. The Social Construction of Scientific Facts, Beverly Hills: SAGE

[29] Lynch, M, 1985, Art and Artefact in Laboratory Science. A Study of Shop Work and Shop Talk in a Research Laboratory, London: Routledge and Kegan Paul, XV

[30] Whitley, R D, 1983, From the Sociology of Scientific Communities to the Studies of Scientists' Negotiations and Beyond, Social Science Information, 22, pp 681-720, p 707

[31] Barnes, B, 1977, Interests and the Growth of Knowledge, London: Routledge and Kegan Paul; Bloor, D, 1976, Knowledge and Social Imagery, London: Routledge and Kegan

[32] Merton, R K, 1973, The Sociology of Science. Theoretical and Empirical Investigations, Chicago: Chicago University Press

[33] Giddens, A, 1984, The Constitution of Society, Berkeley: California University Press

[34] Luhmann, N, 1982, The Differentiation of Society, New York: Columbia University Press; Luhmann, N, 1990, Die Wissenschaft der Gesellschaft, Frankfurt a.M.: Suhrkamp

[35] For an overview see Powell, W W/ DiMaggio, P J, (ed) 1991, The New Institutionalism in Organizational Analysis, Chicago/London: Chicago University Press and Oliver, C, 1992, The Antecedents of Deinstitutionalization, Organization Studies, 13, pp 563-588

[36] See Brunsson, N, 1989, The Organization of Hypocrisy: Talk, Decisions, and Actions in Organizations, Chichester: Wiley; March, J G/ Olsen, J P, 1989, Rediscovering Institutions. The Organizational Basis of Politics, New York: Free Press;

[37] Granowetter, M, 1985, Economic Action and Social Structure: The Problem of Embeddedness, American Journal of Sociology, 91, pp 481-510; Meyer, M W/ Zucker, L G, 1989, Permantly Failing Organizations, London: SAGE

[38] Rowan, B, 1982: Organizational Structure and the Institutional Environment: The Case of Public Schools, Administrative Science Quarterly, 27, pp 259-279; Meyer, J W/ Scott, W R/ Deal, T E, 1981, Institutional and Technical Sources of Organizational Structure: Explaining the Structures of Educational Organizations, Stein, H D, ed, Organization and the Human Services, Philadelphia: Temple University Press, pp151-178