The Invisible Revolution The Construction of Institutional Change in the Fisheries

Petter Holm April 2001

Norwegian College of Fishery Science University of Tromsø Tromsø

FOREWORD When I started writing what would eventually become “The Invisible Revolution,” I thought that it would take a month or so. It turned out to take almost three years. But then, the piece that resulted from the process is quite different from what I had imagined when I started. At that time, my intention was to write a brief summary of some of my published papers. While “The Invisible Revolution” perhaps is a summary of sorts, its 200+ pages hardly makes it brief. The reason for this is quite simple, however. As stated in the regulations for the Ph.D. degree, “[s]everal minor pieces of work may be accepted as part of a doctoral thesis, provided that the content of these works forms a whole.” Since the theoretical perspective of the “minor pieces” has shifted from neo-institutional theory, via constructivist sociology of science, to ActorNetwork Theory, “giving a detailed account of their interrelationship” required a comprehensive discussion. The cost of such a lengthy theoretical discourse is that the empirical descriptions and analyses on which the summary builds are pushed into the background. While the task of the summary is theoretical discourse, however, it is up to the eight papers to contribute the empirical content. This is not ideal, since it is likely – simply because the summary comes before the papers – that most readers will read the summary first and then the papers. In addition, the papers were written before the summary. The empirical analyses are therefore not in all instances perfectly positioned in support of the theoretical arguments. I cannot remedy the latter problem. Or rather, that would have required that I changed the thesis from a collection of minor pieces into a monograph. Time constraints forbid that. But when the former problem is concerned, I have a solution: It is a good idea to read the papers first, and the summary afterwards. Alternatively: Read the papers performing the relevant theoretical perspective before you read the chapter discussing that perspective. This would mean the following reading sequence: Chapter 1; Papers 1, 2 and Chapter 2; Papers 3, 4, 5 and Chapter 3; Papers 6, 7 and Chapter 4; Paper 8 and Chapter 5. A lot of people have contributed in different ways to this thesis. Thanks to Richard Apostle, Peter Arbo, Conner Bailey, Gene Barrett, Jonette Braathen, Pål Christensen, Tony Davis, Sveinung Eikeland, Jens-Eric Eliassen, Erik Oddvar Eriksen, Einar Eytórsson, Narve Fulsås, Hallgeir Gammelsæter, Abraham Hallenstvedt, Bjørn Hersoug, Alf Håkon Hoel, Ingrid Holm, Svein Jentoft, Jahn-Petter Johnsen, Tor Arne Lillevoll, Dan MacInnes, Anita Maurstad, Kevin McCafferty, Tom McGuire, Bonnie McCay, Leigh Mazany, Knut H. Mikalsen, Eva Munk-Madsen, Kristin Berg Nordstrand, Emeli Norrbin, Fredrika Norrbin, John Phyne, Stein Arne Rånes, Håkan

i

Sandersen, Nathalie Steins, Kevin St.Martin, Olav S. Stokke, Jan Sundet, David Symes, Hallvar Tjelmeland, Doug Wilson, Johan Williams, Peter Ørebech, and Bernt Aarset. Thanks to the John D. and Catherine T. MacArthur Foundation for supporting the research project Riding Out the Storm, which started me out on the way that led to this thesis. Also thanks to the Norwegian Research Council (Kyst og Bygdeutvikling) for supporting the research reported here, in particular in the project 121255/111 ‘Norsk fiskerinæring’ som faktum og fiksjon. Finally, thanks to the Norwegian College of Fishery Science and the University of Tromsø for allowing me – indeed making it an obligation – to pursue research as part of my job.

ii

The Invisible Revolution

1. WHY YOU SHOULD NOT BELIEVE IN THE INVISIBLE REVOLUTION

1

ON A DARK WINTER NIGHT IN FINNMARK ...

1

INTRODUCING THE PROJECT

8

THE INVISIBLE REVOLUTION

10

ACCOUNTING FOR THE INVISIBLE REVOLUTION

18

THE SOCIAL CONSTRUCTION OF FISHERIES RESOURCE M ANAGEMENT

22

DECONSTRUCTING SCIENTIFIC KNOWLEDGE

25

2. INSTITUTIONAL THEORIES

27

INTRODUCTION

27

M APPING INSTITUTIONALISM

29

REVIEW: NEW INSTITUTIONALISM IN ORGANIZATIONAL SOCIOLOGY

34

ORGANIZATIONAL INSTITUTIONALISM VERSUS RATIONAL CHOICE INSTITUTIONALISM

40

NEW VERSUS OLD INSTITUTIONALISM

43

VARIETIES OF ORGANIZATIONAL INSTITUTITIONALISMS

44

INSTITUTION AS ENVIRONMENT (MEYER, SCOTT & CO.)

47

Early dualist version

47

Mature structuralist version

55

INSTITUTION AS ORGANIZATION (ZUCKER, TOLBERT & CO.)

58

Structuration

61

The problem with structuration

64

Structuration as a basis for institutional analysis

67

EXPANDED INSTITUTIONALISM (P OWELL AND DIMAGGIO)

69

Institutions shape action – action shapes institutions

69

A return to economic man?

74

Expanded institutionalism as an assemblage

77

INSTITUTIONAL THEORY AND THE INVISIBLE REVOLUTION

79

3. THE SOCIOLOGY OF SCIENTIFIC KNOWLEDGE

87

INTRODUCTION

87

THE ORTHODOX NOTION OF SCIENCE

89

iii

Holm

SSK: A "STRONG" SOCIOLOGY OF SCIENCE

93

EXEMPLAR 1: MATHEMATICS

100

EXEMPLAR 2: REPLICATION

102

EXEMPLAR 3: CONTROVERSY AND DISAGREEMENT IN SCIENCE

103

EXEMPLAR 4: SCIENTIFIC KNOWLEDGE AND SOCIAL INTERESTS

105

SCIENCE AS KNOWLEDGE

108

IS SSK REFLEXIVE?

111

DOES SSK B ELIEVE IN REALITY?

114

SSK AND THE INVISIBLE REVOLUTION

117

4. ACTOR-NETWORK THEORY

125

INTRODUCTION

125

THE SOCIOLOGY OF TRANSLATION

128

THE MATERIALITY OF TEXT

130

THE PRINCIPLE OF EXTENDED SYMMETRY

135

HYLOZOISM

139

HOW FACTS ESCAPE FROM LABORATORIES

144

SOCIOLOGIES

149

THE M ODERN CONSTITUTION

155

P URIFICATION VERSUS TRANSLATION

155

THE HISTORY OF MODERNITY

159

TAKING ANT TO TRIAL

163

ANT AND THE INVISIBLE REVOLUTION

173

ANT AS IRONY

173

THE INVISIBLE REVOLUTION

176

5. INVENTING THINGS THAT DO NOT EXIST

183

THE FICTION OF THE INVISIBLE REVOLUTION

183

THE INVISIBLE REVOLUTION IS A HETEROGENEOUS NETWORK

187

REFERENCES

193

iv

LIST OF PAPERS 1. Holm, Petter, 1995. ‘The Dynamics of Institutionalization: Transformation Processes in Norwegian Fisheries.’ Administrative Science Quarterly 40: 398-422. 2. Holm, Petter, Stein Arne Rånes and Bjørn Hersoug, 1998. ‘Political Attributes of Rights-based Management Systems: The Case of Individual Vessel Quotas in the Norwegian Coastal Cod Fishery.’ In D. Symes (ed.) Property Rights and Regulatory Systems in the Fisheries: 113-126. Oxford: Fishing News Books. 3. Holm, Petter, 1996. ‘Kan torsken temmes? Moderniseringsprosesser i norsk fiskerinæring 1935-1995.’ In E.O. Eriksen (ed.) Det nye Nord-Norge: Avhengighet og modernisering i nord: 109-142. Bergen: Fagbokforlaget. 4. Holm, Petter, 1996. ‘Fisheries Management and the Domestication of Nature.’ Sociologia Ruralis 36(2): 177-188. 5. Holm, Petter, Stein Arne Rånes and Bjørn Hersoug, 1998. ‘Creating Alternative Natures: Coastal Cod as Fact and Artefact.’ In D. Symes (ed.) Northern Waters: Management Issues and Practice: 79-89. Oxford: Fishing News Books. 6. Holm, Petter, 1998. ‘Et Maritimt Mekka?’ Samtiden 2/3: 89-96. 7. Holm, Petter, 2000. ‘Ressursforvaltning som heterogent nettverk.’ In Hallgeir Gammelsæter (ed.) Innovasjonspolitikk, kunnskapsflyt og regional utvikling: 83102. Trondheim: Tapir. 8. Holm, Petter, Bjørn Hersoug and Stein Arne Rånes, 2000. ‘Revisiting Lofoten: Co-Managing Fish Stocks or Fishing Space?’ Human Organization 59(3): 353364.

v

The Invisible Revolution

1. WHY YOU SHOULD NOT BELIEVE IN THE INVISIBLE REVOLUTION

On a Dark Winter Night in Finnmark ... On December 20, 1993, fisher NN from Berlevåg was fined NOK 2000 by the police in Vardø because “On Wednesday 05.05.93 he delivered a fish catch to Ulve Berlevåg A/S without signing the landing receipt” (Vardø herredsrett, 1994). This was a clear breach of the landing regulation, that states that: The buyer shall write out the landing receipt [sluttseddel/bryggeseddel] according to the sales organizations’ rules specified by species immediately after delivery and the receipt shall be signed by the buyer and fisher before the latter leaves the place of landing. (§1, Subsection 1, regulation on landing receipt and delivery journal persuant to the Salt Water Fishing Act)

When NN refused to accept the fine, it was neither because he disputed the facts, nor that he did not know of the regulation. Instead, he maintained that strict adherence to the regulation would be practically impossible. The fisher in question participated in the float line fishery during the summer season in Finnmark. The vessels participating in this fishery depend on coordinating the setting and drawing of their lines, since this must be done simultaneously by all vessels in the relatively restricted fishing grounds in question, and only when the current is favorable. Nevertheless, problems often occur, so that the vessels will land later than planned. In order to avoid missing the next fishing day, a practice had developed in the East Finnmark area of allowing the vessels to deliver their catch outside the normal opening hours of the processing companies. One way of doing this, which did not create problems with respect to the landing regulations, was for the employees at the processing plant to work overtime. Another way, and the reason why NN was fined, was the practice of fishers unloading vessels themselves and storing the fish in the processors’ cold-store until the next morning. The catch was then weighed the next day, and only then was the delivery receipt filled out and signed. Since the vessel and the skipper would by now have left for the fishing ground, the “shore man” would sign on behalf of the skipper. In the line fishery, the shore man, being a full member of the crew, would 1

Holm

stay on shore and, besides acting as a deputy for the skipper in relations with the fish buyer, arrange for the baiting of lines, and sort out various administrative and practical matters in connection with the fishery (Vardø herredsrett, 1994: 8-9). On the day in question, then, an inspector from the Control Agency visited the processing plant in Berlevåg, and found that NN had left the landing area without signing the delivery receipt, and that it had later been signed by the shore man. Since this was in violation of the landing regulation as the Control Agency interpreted it, NN was prosecuted and fined by the police in Vardø. The reason for the Control Agency’s strict interpretation of the landing regulation had to do with its role in monitoring and control of fisheries resource management and in particular with the quota system implemented in 1990. To be able to enforce these regulations, a reliable system of catch records at both vessel and fleet levels had to be established. The landing receipt was one of several instruments in this respect. The signed receipts were to be kept on board the vessel for three months, so that on inspection, they would serve as a basis (together with the catch diary) for checking actual catches against quota allocations. For the Control Agency, it was of crucial importance that the landing regulation was strictly adhered to. The practice that had developed in eastern Finnmark would lead to a delay in the signing of the catch receipt so that it would not follow the vessel, as required. This would hamper control on the fishing grounds. Perhaps more seriously, keeping the catch in the cold-store for a time before actually selling it to the fish buyer would make it difficult for the Control Agency, on inspection, to know which catch belonged to which vessel (Vardø herredsrett, 1994:9). When fisher NN did not accept the NOK 2000 fine, the police brought the case to court. In mid-December 1994, NN together with 18 other fishers appeared in the town court of Vardø (Vardø herredsrett), charged with violating the landing regulation. The scene was set for a confrontation between the fishers and the Control Agency over the interpretation of the landing regulation. As reported by the newspapers, much was at stake for both parties. The fishers argued that a strict interpretation of the landing regulation would seriously hamper the fishery. Moreover, they felt criminalized by the Control Agency. Not only had they not attempted to engage in fraud, as implied by Arne Luther, director of the regional office of the Control Agency, in his testimony to the court (Finnmarken, 10.12.1994:6), but they had simply been following the practice that had developed over many years (Finnmarken, 9.12.1994:6). The representatives of

2

The Invisible Revolution

the Control Agency argued for the need to follow the landing regulation to the letter. In their view, the reference to established practices was not relevant. The landing regulation in question had been adopted in 1990, the same year the quota system had been established. While the Control Agency in the first couple of years had only prosecuted the most serious violations of the regulations, a stricter practice had been introduced from 1993 (Vardø herredsrett, 1994:13). This should also be seen in the context of the changing role of the Control Agency, which from its inception had been responsible for product quality control. From the early 1990s, however, the agency had also been given co-responsibility for control activities in the area of resource management. Hence, the court case was a test not only of a strict interpretation of the landing regulation, but also of the status and power of the Control Agency itself. As seen from the Control Agency’s perspective, Norwegian fisheries were indeed full of criminals and fraudulent practices. The need for a strict interpretation of the landing regulation must be seen in light of the general context during the early 1990s, when, according to the regional director of the Control Agency, fraud was indeed occurring in the interaction between fishers and fish buyers, and, in a single year, as much as 50 000 tonnes of fish were unaccounted for, presumably sold on the black market (Finnmarken, 12.12.1994:6). What was the outcome? In the Vardø court all 19 fishers were found not guilty. In 11 of these cases, the three judges unanimously came to the conclusion that, for a variety of reasons, the fishers were not to blame for the breach of the landing regulation. In eight cases, however, all conforming closely to that of NN described above, the professional judge disagreed and held that the fishers were guilty as charged. In his view, the landing regulation could not be interpreted as allowing the shore man to sign the landing receipt, as this was the skipper’s duty. Nor could there be any doubt that the regulation expressly forbid the skipper and vessel to leave the landing area before the receipt has been signed. The majority, the two lay judges, saw it differently, however. The reason they gave for acquitting the eight fishers was that: ... the fishers’ opportunity to conform to the regulation must be taken into consideration in its enforcement. In view of this, the majority cannot [see] that the fishers’ can be faulted for the way they have acted. We also refer to the fact that the landing receipts were signed as soon as this was practically possible, and that the possible delay cannot have any significance for the authorities’ resource control. (Vardø herredsrett, 1994: 11).

3

Holm

While the prosecutor, the Vardø chief of police, accepted the 11 unanimous acquittals, he appealed the eight cases where the fishers had been acquitted against the professional judge’s vote. The supreme court, giving its verdict in December 1995, agreed with the professional judge Vardø court, and reversed the acquittal. The reason given was that: The majority has, in my opinion, posed too lenient requirements with regard to when the signing [of the landing receipt] must be accomplished. The regulation requires that this must happen before the vessel leaves the place of landing, which must be understood as the buyer’s dock facilities. The prosecutor has argued for the need for clear, strict rules, in particular in order to ensure that the quota regulations are adhered to. I agree with this (...). There is no room for deviation from this requirement, even if it should be more practical for the fishers. (...) The town court’s majority has consequently built on a faulty interpretation of the rules, and the decision is overturned on this basis. (Norges Høyesterett sak lnr 157, snr 81/1995)

This confrontation between the fishers and the Control Agency in the courts is one instance of the drama that forms the subject matter of this thesis: The introduction and institutionalization of fisheries resource management in Norwegian fisheries. The consequences of this process have been far-reaching and can be observed in many different arenas, as demonstrated by the confrontation over landing procedures. For instance, it has forced changes in the fishers’ practices, in the way they carry out their daily tasks. Following the conviction of the Finnmark fishers as reported above, the fishers can no longer unload their catch immediately when they arrive late from the fishing grounds; the skipper can no longer delegate the signing of the landing receipt to the shore man. Unless the fishers can persuade the processor to have his employees work overtime, this means that they will lose fishing days, and maybe that the float-line fishery will become unprofitable – at least this is what they claim. As indicated, such changes in practices will here be seen as indications of changes in institutions. But what are institutions, and how do they change? Without going into a formal definition at this point, we can note that an institution entails patterned activities, practices, or routines (Jepperson, 1991). Sometimes, but not always, such patterned activities are governed by rules. Our example portrays a situation characterized by conflict over the interpretation of rules. The routines were challenged; the practices that the fishers had taken for granted were considered by the Control Agency to be in conflict with the regulations. This example, then, is one of

4

The Invisible Revolution

institutional change rather than of the automatic reproduction of taken-for-granted practices. Situations like this, while only referring indirectly to institutions as routine-reproduced practices, are nevertheless important for the analysis of institutions. When the degree of institutionalization is high, a given social practice will be taken for granted, maybe even to the point of invisibility, by those who are bound by it. This is simply the way things are done. Full stop. No explanation required. The established practice needs no justification and no alternatives are seriously considered. While such situations of course are available for analysis, they will – precisely because they are taken for granted – be somewhat inaccessible. When an institution is threatened, however, the sources of information will overflow as the rationalized myths (Meyer and Rowan, 1991) that account for its origin and function, the values to which it is committed (Selznick, 1957), and the power resources that are available for their defense (Stinchcombe, 1968) are brought into the open. In this context, we are not so much interested in landing practices as an institutionalized pattern of activity in themselves, but in the way these practices are nested within more general institutional patterns. This becomes visible when we consider the arena within which the conflict over interpretation of the landing regulation was fought out: The courts of law. From an institutionalist perspective, the court – an institution or institutional system in its own right – is a mechanism for producing authoritative interpretations of laws and regulations. In the case at hand, the court had to decide what was the correct interpretation of the landing regulation – the fishers’ lenient one or the Control Agency’s strict reading. The recourse to a specialized organizational system devoted to the interpretation of rules is, of course, in itself important in settling conflicts. One would imagine that the weight of such decisions would be considerable even if the parties involved in a conflict were not compelled to adhere to them. As it happens, however, the authority of the court is also tied to the systematic allocation of societal power behind its decisions. The decision of the court as to the interpretation of the landing regulation is backed up by the administration of penalties – in our case whether the fishers will be forced, with the aid of the police if necessary, to pay their fines or not. According to Stinchcombe (1968:184) the concept of “institution” is connected to the values and norms that have a high correlation with power. In other words, institutionalization has to do with the concentration of power specially devoted to some value. This is directly relevant to

5

Holm

the conflict over the interpretation of the landing regulation. The court case was, among other things, a confrontation between different values. The fishers, on the one hand, were defending a landing practice that would enable them to pursue fishing in a “rational” way. The Control Agency, on the other hand, were fighting to end a landing practice that made control difficult. There certainly were values at stake here: Would the fishers be made to bend in the name of resource management concerns? Or would the Control Agency’s inspectors be made to give way in the name of the practical and economic efficiency of the fishery? Both sides in the conflict were able to mobilize considerable support in defense of their position. For instance, both values that came into conflict here were protected in the form of specialized organizational apparatuses committed to them and both were protected by formal legislation. In the case of the resource management concern, the relevant organizational apparatus was the Control agency, an administrative unit under the Norwegian Directorate of Fisheries. In the case of the practical and economic rationality of the fishery, the relevant organizational apparatus was the fishers’ sales organization, the Norwegian Raw Fish Organization (Norges Råfisklag). This organization had been established in accordance with the 1938 Raw Fish Act, which gave organizations controlled by fishers monopoly rights over the first-hand sales of fish. § 2 of the Norwegian Raw Fish Organization’s charter states: “The purpose of the organization is, by way of organized sales, to achieve high and stable prices for fish as well as by-products of fish.” (Norwegian Raw Fish Organization’s Charter, quoted from Christensen and Hallenstvedt, 1990: 225). Admittedly, the Raw Fish Organization played a more peripheral role in the conflict than the Control Agency. The key point of the trial was the interpretation of the landing regulation, the implementation of which was part of the domain of the Control Agency. Since the issue involved the practical organization of transactions between the fishers and fish buyers, the Raw Fish Organization nevertheless was involved. It was the Raw Fish Organization that had established the procedures regulating the transactions between fishers and fish buyers, including the formatting of the landing receipt form. It was also the Raw Fish Organization that was responsible for the collection, certification and aggregation of the catch and landing data reported on these forms. The purpose of these procedures is primarily to implement the price regulations and guarantee the fishers’ economic interests in the transactions. With the institutionalization of resource

6

The Invisible Revolution

management as the primary concern within Norwegian fisheries, however, these procedures and statistics were also included as a part of the authorities’ surveillance and control apparatus (Paper 3). Hence, the Raw Fish Organization, like the Control Agency, has control and surveillance for resource management purposes as part of its organizational objectives. With the former’s strong commitment to the protection of the fishers’ economic interests, however, we should not expect that its interpretation of the landing regulation would be identical to that of the Control Agency. The difference between the two organizations became clearly visible during the trial in the testimony of the representative of the Norwegian Raw Fish Organization. He suggested that resource management control would not be hampered much if the landing receipts were not signed immediately. Since it was the buyers’ receipts that would be returned to the Raw Fish Organization, and would form the basis of its resource control, it would not matter much that the fisher’s copy was not present on the vessel for a day or so. To this, the director of the regional Control Agency office retorted that the Raw Fish Organization’s resource control was an illusion. The only practical way to check is when the landing receipt, signed by the skipper, remains with the vessel at all times (Finnmarken, 9.12.1994: 5). The conflict of values and power between two organizations which we are catching a glimpse of here is an interesting aspect of institutionalization and of the relationship between institutions and individual behavior. A practical way to allocate power to a value, which is an important dimension of institutionalization according to Stinchcombe (1968), is to make it the concern of a specialized organization. Our story includes two examples: The Raw Fish Organization’s commitment to the protection of the fishers’ economic interests and the Control Agency’s commitment to the effective implementation of resource control. Furthermore, the degree to which a value is institutionalized will become visible when it is poised against other values in confrontations like the one that occurred in the court case over the interpretation of the landing regulation.

7

Holm

Introducing the Project We already know the final outcome of this trial of strength. The Supreme Court’s verdict meant that the power of the state was thrown behind the implementation of the landing regulation as interpreted by the Control Agency. This means, according to Stinchcombe’s definition, that resource control and ultimately resource management, became more institutionalized, while the economic interests of the fishers, as protected by the Raw Fish Act and the Raw Fish Organization, become less so. This brings us to a point where the project pursued in this thesis starts to become visible. In one way or another, the eight papers below and this summary are concerned with the establishment and institutionalization of fisheries resource management in Norwegian fisheries. The story they tell is one of a fundamental institutional change. Fisher NN’s fine for having breached the landing regulation and the subsequent confrontation between the Control Agency and the Raw Fish Organization in the Vardø Court in December 1994 is but one instance of this more general process, in which resource management was established as the key concern within Norwegian fisheries and the overriding reason for government intervention in the sector. At the same time, this process meant withdrawal of attention, resources and power from the values that from the late 1930s had been fundamental within the fisheries, namely the economic security of the petty capitalist fishers and the protection of the fisheries as the cultural and economic basis of the settlement along the Norwegian coast. I will refer to the latter system of values and practices as the MSO system after the Mandated Sales Organizations that constituted its basic building blocks. The institutional transition underway during the 1980s and early 1990s – the transition from the MSO system to the resource management system – was indeed a major one. It redefined the fundamental problem of the fisheries, from an emphasis of overabundance of fish and the threat of market crises to an emphasis on the scarcity of fish, vulnerability of fish stocks and the threat of resource crises. It redefined the fishery resource, and turned it from freely accessible common property, into tightly controlled state property. It redefined nature itself, from being strong and sturdy, impervious to whatever man could do to it into something fragile and highly sensitive to human activities. It redefined the identity of the key human actors, the fishers, from being weak and vulnerable, in need of and worthy of state protection, into greedy and rapacious individualists who, if unchecked, would destroy the resource from which they drew

8

The Invisible Revolution

their livelihood. It meant a shift in the locus of governmental intervention, from regulation of marketing and trade, to regulation of harvesting and quotas. It implied new forms of organization, from aiming at self-organization and co-management, to top-down, centralized regulation and control. It meant new power constellations within the fishery sector, from a situation in which the coastal fishers were privileged and empowered, to one in which the industrialists became predominant in the political arena. In short, this was a fully-fledged institutional revolution, a paradigm shift, affecting virtually every aspect and all levels of the fishery. Here is what I want to achieve: To describe and make understandable the transition set off by the introduction of fisheries resource management; to display with indisputable clarity the characteristics of and differences between two institutional configurations – one bound for the great garbage heap of history, the other ascending towards the status of taken-for-granted reality. Why did the MSO configuration self-destruct when resource management entered the scene? How did it happen? What were the consequences? Now you know what I want to accomplish. Do you find it interesting? Perhaps not. Perhaps you think that such a project cannot be much of a challenge, at least if you accept at face value my rendering of the Norwegian fishery sector as the arena for a fundamental institutional transition. But you should not, at least not yet. This is because this perception of the development of Norwegian fisheries can only come as a conclusion of this thesis. Without the support of the papers below, and this introduction, there is virtually nothing that can be mobilized in support of this way of looking at things. So, at this point, if you do not happen to have read some of the papers that are reprinted below, there is virtually no reason why you should accept my interpretation of recent developments in Norwegian fisheries as one of fundamental institutional transition. So, I will have to persuade you, the reader, that I am not breaking down an open door here. Indeed, I’m trying to make you accept that without the work done in this thesis, it would not even be obvious that there was a door that could be opened here at all. This too must be part of my project. In another words, if my project is going to be of any interest, I must persuade you that I have something to offer. Resorting to a jargon that I will introduce later, I want to establish this thesis as an Obligatory Passage Point (OPP) between the reader and the reality of Norwegian fisheries as including a recent fundamental institutional transition. If you know that

9

Holm

already, without the aid of this thesis; if you do not need my work in order to accept the existence of this transition, I will have failed. If this part of reality was obvious without the following 300 pages, I have accomplished nothing. How can I then believe that you will read on from here, much less hope that this thesis will earn me a Ph.D? This means that the success of my project is intimately tied to making you accept, at the outset, the role of a non-believer in the fundamental institutional transition, and then, by mobilizing all sorts of data, descriptions, analyses, and theoretical arguments in support of its existence, have you give up all resistance. An important resource for achieving this is an oxymoron: “The invisible revolution.”

The Invisible Revolution An invisible revolution. How can I ever hope to persuade you of the reality of such a creature? A revolution is supposed to be a dramatic and often violent change of regime, plainly visible to everyone. How can it be invisible? Let us start with the second half of this term, with the grounds for labeling the institutional transition that recently took place in the fisheries a “revolution.” I readily admit that this term is chosen for dramatic effect and that there are things that do not fit the bill. For instance, the transition has not been as quick as your average coup d’état. Instead it stretched over a 20-year period. Our revolution was not as bloody as the Russian or the French revolutions. There was no rolling of heads. No shots could be heard at dawn. No-one was jailed or exiled. Despite this, I maintain that the term revolution captures important aspects of what has happened in Norwegian fisheries after the introduction of fisheries resource management. Like many revolutions, the one triggered by the onslaught of resource management deeply influenced the everyday lives of ordinary people, the fishers and the coastal population. A number of those who used to live from the sea have been forced to move and seek other occupations. Many youngsters, who until recently would have found jobs and settled down in the coastal communities where they were born, or ones like them, now must seek their future in different places, most often in urban settings. The process of becoming a fisher – for the few who can do it – has also

10

The Invisible Revolution

changed drastically. Before resource management, a major barrier for those who wanted to be fishers had to do with the physical and mental hardships of this occupation. Becoming a fisher entailed the hardening of body and mind to tolerate long working hours and grueling physical exertions, long periods away from home and family, great risks to personal safety and health. While many fishers nowadays suffer similar hardships, prospective fishers are also forced to cross barriers of a different kind, like the bureaucratic jungle of rules and procedures that must be negotiated in order to become owners of fishing vessels and holders of licenses and quotas. This means that the identities of the fishers and the fish, as well as the relationship between the two, have been deeply transformed. The fish are no longer common property, that is, a resource governed by a complex but usually informal and weakly institutionalized system of norms. After the introduction of modern resource management, fish has been solidly established as a resource owned, controlled and managed by the state. The state has thus inserted itself in between the fisher and the community on the one hand, and the fish on the other. From free creatures of the oceans, fish have been turned into a commodity, one tradable asset among many in the state’s possession. Similarly, the fishers have been transformed from hunters operating on the edge of modern society, into license holders and quota owners in the midst of modernity. It is thus no longer a truism, as Jentoft (1999) has suggested, that fishers not only fish from boats but also from communities. This is perhaps a truism under common property regimes. After the introduction of modern fisheries resource management, however, it is more to the point to say that fishers not only fish from boats, they also fish from, or perhaps through, states. In this process, what should count as expert knowledge with regard to the fisheries has been redefined. No longer is the knowledge that enables the fishers to carry out the practice of fishing – handling vessels and gear, finding and catching fish – regarded as expertise. Even if you are a successful fisher, that does not qualify you as a fish expert. After the breakthrough of fisheries resource management, the fish expert is unequivocally a scientist, usually a natural scientist trained in fisheries biology and oceanography. It is no longer the practical knowledge of how to find and catch fish that equals expertise, but the command of scientifically certified methods and models designed for measuring and predicting the size of fish stocks (cf. Pálsson, 1991).

11

Holm

Given the substitution of the centralized state in place of local communities as the institutional locus of the fishery, and the substitution of fishery scientists for fishers as the holders of expert knowledge, it should come as no surprise when I claim that the introduction of fisheries resource management has also transformed the Norwegian fisheries as a political economy. The MSO system constituted the Norwegian fishery sector as a petty capitalist sector, with smallscale fisher-owners as the most important economic and political actors. With the introduction of fisheries resource management, the fishery sector has gradually been transformed into an industrial capitalist system, with relatively large-scale industrial processors as privileged economic and political actors. As we shall see later (Papers 1 and 3), this change process can be portrayed as a reconstruction of the overall organizational map of the sector. A number of old organizations have been disposed of and replaced with a few larger organizations – particularly within fish processing and exports. In addition, there have been a series of changes in legal and financial arrangements that have affected the distribution of responsibilities, resources and relationships among key interest groups. An important reason for and corollary of this are the changing relationships between sector interests and government agencies. With the introduction of fisheries resource management, the mode of government intervention in the fisheries has changed from one of delegation of responsibility and co-management in a corporatist tradition into one that relies much more on top-down hierarchical control. All in all, the outcome of these organizational changes is a fundamental shift of power distribution between the key interest groups within the fisheries. Put somewhat crudely, we can say that the fishers have lost and the processors have gained. Power has moved from sea to land. In addition, there has been an overall shift of focus and ideology, so that concern for traditional, small-scale owner-operators has now become more peripheral, while concern for the interests of large-scale industrial actors – whether at sea or on land – has become more central. Is this a revolution? Perhaps this choice of label exaggerates the violence and abruptness of this process. Nevertheless, I will maintain that it accurately describes the drama and the importance of the change processes involved.1 What we have here is a fundamental political,

1

Merriam-Webster's Collegiate Dictionary’s (online version) definition of “revolution” includes “an activity or movement designed to effect fundamental changes in the socioeconomic situation”; “a fundamental change in the way of thinking about or visualizing something: a change of paradigm (the Copernican revolution)” and “a changeover in use or preference especially in technology (the computer revolution, the foreign car revolution).”

12

The Invisible Revolution

institutional, conceptual and ideological transition, a regime shift that turned the power relation in the sector upside down, a reconstruction of fishing as a livelihood and a way of life for thousands of people. If it can be accepted that the changes set off by the introduction of modern resource management in the fishery are revolutionary in scope and scale, this does not make it easier to argue convincingly for the appropriateness of the first half of our oxymoron, namely that this revolution has been invisible. On the contrary, how can a revolution be invisible? Would it not be more reasonable to expect that something as fundamental, large-scale and violent as a revolution, involving dramatic power shifts and, one would think, fierce resistance and struggle, will by necessity be highly conspicuous and almost impossible to ignore, even for those who would want to? This would indeed seem reasonable. Nevertheless, the organizational, political and ideological changes that followed the introduction of scientifically based resource management in Norwegian fisheries have received scant attention. During the late 1980s and early 1990s, when I first started to understand that the sector was going through fundamental changes, I was struck by the lack of attention this process received. The Fishermen’s Association hardly recognized that the sector was changing and that their power bases were being eroded. The officials at the Fisheries Ministry were uninterested, preoccupied with the ongoing process of EU adaptation. As an illustration, consider the story about the destruction of the Main Agreement. This agreement, signed by the Government and sector interests in 1964, established the legal and organizational framework for negotiations between the state and the fishery sector, represented by the Fishermen’s Association, for fishery subsidies. In line with the industrialization doctrine in vogue at the time, the Government had intended the Main Agreement as a program for structural rationalization, with investment support and buy-back programs as the main instruments. In practice, however, it became an income support scheme. During the 1964-1990 period the subsidies amounted to 19 per cent of the value of the catch, of which nearly 80 per cent was allocated to income support (Holm, 1991). In this way, the Main Agreement established a system whereby the state buffered the fishery sector from fluctuations in fish stocks and markets. Instead of an instrument for rationalization and industrialization, as the Government had intended, the Main Agreement had been turned into an economic safety-net for the fishermen. Instead of

13

Holm

reconstructing the sector, the agreement helped protect and fortify the established institutional order (Holm, 1991). As we shall see later (Papers 1 and 3), the Main Agreement became an important part of the institutional configuration that I call the MSO system. A striking feature of this system, of which the reconstruction of the Main Agreement itself is an important example, is the considerable amount of influence and political resources it gave to the fishers. Here you have a government determined to industrialize and rationalize the fishery, and, as part of that policy, to break up the established economic and institutional structure of the sector. What was the result? A subsidy system that produced almost exactly the opposite of what had been intended (Holm, 1991). Now, the details of the explanation I have proposed for this are given in the papers below and need not concern us here. What is important in this context is the considerable capacity of the MSO system to mobilize against external threats, indeed, to turn them to its own advantage, as in the case of the Main Agreement. Interestingly enough, this feature of the MSO system was again confirmed in the early 1980s, in connection with an attempt to have the Main Agreement and the subsidy system abolished. The attack came from within the Ministry of Fisheries itself, and was headed by an economist by the name of Bjørn Brochmann. From the perspective of bioeconomics, he argued that the fishery subsidies, rather than improving the efficiency and profitability of the fisheries, which was its formal purpose, contributed to overcapacity and thereby undermined that very goal. The logic was as follows: In the fishery, both market prices and catches will fluctuate. In good years, some of the profits will be converted into larger vessels and more efficient gear. In poor years, however, the subsidies will allow the fishers to survive economically. Thus, the Main Agreement set the fishery rolling down a one-way street, heading relentlessly towards growing over-capacity and dependenc on subsidy (Brochmann, 1980; 1981). What Brochmann did was simply to apply the logic of the resource management model to an institution created within the MSO system. If, as everyone should know, you add subsidies to an open-access fishery, the result will inevitably be even more overcapacity, and an increasing, negative resource rent. From that viewpoint, the Main Agreement was not only inefficient, but counterproductive and irrational, and should hence be reformed (Holm, 1991). It should be pointed out that Brochmann’s views not only were solidly founded in bioeconomics, but also in the reality of the Norwegian 200-mile exclusive economic zone,

14

The Invisible Revolution

established in 1977. In fact, the Norwegian subsidy system had not been irrational under Mare Liberum, since it helped build capacity and hence allowed Norwegian fishers to participate in the race for fish in international fishing grounds. With the 200-mile limit, however, the most important fishing grounds for Norwegian fishers had come under Norwegian jurisdiction and, at least potentially, under explicit management and control. Brochmann’s attack on the Main Agreement can be regarded as a first attempt, aided by bio-economic analysis, to point out that the old institutional system, with the Main Agreement as a case in point, no longer fit the realities of the fishery sector – as defined in relation to the new oceans regime – and that it therefore had to be reformed. Nevertheless, Brochmann became immensely unpopular in the fishery sector. There was an uproar of protest and indignation. Brochmann and his ideas were thoroughly battered in public meetings and newspaper articles. The reactions in the sector were so strong that the attempt to reform the Main Agreement was cancelled, and the subsidies were allowed to continue as before. Brochmann himself left the Ministry of Fisheries a few years later, allegedly because of his unsuccessful attempt to have the subsidy system changed. So far, the story of the Main Agreement conforms exactly to what we have learned to expect when it comes to important institutions. When they are attacked, those who gain from them mobilize in their defense. But the end of the story is different, and far more interesting. Almost 10 years after Brochmann’s first attack on the Main Agreement, I was engaged in a research project, sponsored by the Fishermen’s Association and the Norwegian Raw Fish Organization, to examine the consequences for the Main Agreement and fishery subsidies of the adoption of the Free Trade Agreement on fish and fishery products within EFTA. This agreement, concluded in 1989, abolished all tariffs and quantitative import restrictions on fish and fish products within the EFTA region from July 1, 1990. As I soon found out, it also meant that all subsidies not directed towards structural measures were to be cancelled by the end of 1994. Income support, the main instrument within the Main Agreement system, was no longer allowed. Hence, roughly 70 per cent of the fisheries subsidies in 1989 would have been illegal according to the EFTA agreement (Holm and Johnsen, 1990). This was the end of the Main Agreement. In my presentation of the result of the project to the Fishermen’s Association, I concluded that the Main Agreement was about to become obsolete.

15

Holm

The interesting thing here is that this conclusion, which by the way has proved to be entirely correct, came as something of a surprise to the representatives of the fishers’ organizations. They had not been informed of, much less been a party to, the negotiations leading to the EFTA agreement. When the agreement had been adopted, they needed an outside researcher to find out what it meant. A key institution within the MSO system was sold cheaply on the open market, and its major constituents were paralyzed, excluded from the process and unable to mobilize for its defense. Where the fishers’ organizations 10 years earlier had effectively immobilized and disposed of Bjørn Brochmann as he attacked the subsidy system, they now watched helplessly as the Main Agreement was transformed from the institutional apex of the MSO system into an empty shell. It must perhaps be granted that the demolition of the Main Agreement is a special case. The other parts of the reform process that followed in the wake of resource management – the adoption of a new fish export law, the cancellation of the Processor Act, the adjustments to the Raw Fish Act, the changes in the Fishing Border Act, the reinterpretation of the rules pertaining to the funding of the Fishermen’s Association, and the introduction of a vessel quota system for the coastal fishery (Papers 1, 2, and 3; cf. Holm and Mazany, 1995) – were not undertaken to the accompaniment of the same stunning silence and immobility on the part of the affected interests as was the dismantling of the subsidy system. While these events occasioned considerable political activity and mobilization, the focus nevertheless remained quite narrow. More generally, it is remarkable that these pieces were never seen in context, that the reform process as a whole was never noted and made into an item on the sector-political agenda. A revolution was going on, but the parties that were about to be disempowered and dispossessed seemed unable to see it, much less mobilize against it. Let me return for a moment to the notion of “revolution.” In opening his book about the scientific revolution, Steven Shapins writes: “There was no such thing as the Scientific Revolution, and this is a book about it” (Shapin, 1996:1). What this refers to is the fact that those who lived in the early 18th century, when the modern institution of science had its breakthrough, neither experienced the events they lived through as a revolution nor labeled it as such. The conceptualization of these events as constituting a scientific revolution was invented much later, by historians in the 20th century. What Shapin denies, then, is the existence of the scientific revolution

16

The Invisible Revolution

as an explicit definitional characteristic of a specific situation by the actors who lived and enacted it. But Shapin does not, and cannot, deny the existence of the scientific revolution as a heavily institutionalized thought object in 20th century scholarly discourse and popular belief. The relevance of this for my project should be obvious. In the case of the fisheries, we also have a situation in which those who have experienced the introduction of resource management have not thought of it as a revolution. Just as in the case of science, the notion of the breakthrough of fisheries resource management as an event of revolutionary dimensions is introduced after the fact, as a reinterpretation of past events. In this way, I could have made Shapin’s words my own: “There was no such thing as the resource management revolution, and this is a thesis about it.” As a matter of fact, I have done something very similar to this, since my key term, “the invisible revolution”, carries out much of the same work as Shapin’s phrase. But not exactly the same. A key difference between the scientific revolution and the resource management revolution is that the latter, in contrast to the former, is not a wellestablished thought object in academic discourse. The qualification, “invisible,” not only refers to the sector’s lack of acknowledgement of the revolutionary effects of the introduction of resource management, but also to its absence in academic discourse. When I started my work on the institutional change process in Norwegian fisheries, and gradually came to realize that a fundamental transition was happening, the most astonishing thing was that this had not already been made into a well-researched phenomenon in fisheries social science. Much like the economic actors and their representatives, the anthropologists, sociologists and political scientists working on fisheries-related research in Norway did not seem to recognize that the sector was changing, or at least did not seem to realize the scope of these changes. True enough, they shifted around their research in accordance with the transition. Where they previously had been preoccupied with the corporatist organizational system of Norwegian fisheries, they now began working on resource-management related issues. The reasons for and institutional context of this shift of focus were not made research issues on their own terms, however. Even at the Norwegian College of Fisheries, which should be ideally positioned to register and analyze sectoral change processes like this, it seemed to pass unnoticed, at least as measured in the basic teaching agenda. Resource management issues were included; that was not the problem. It was added on

17

Holm

as a separate topic, however, presented isolated from and as if it had nothing to do with the political and institutional organization of the fisheries sector.

Accounting for the Invisible Revolution Have I convinced you of the reality of an invisible revolution? Perhaps not yet. What we have so far is a label, “the invisible revolution,” and a set of empirical observations that seem to fit it. I have argued, referring to the papers reprinted below, that a dramatic and fundamental institutional and political transition has taken place in the fisheries. I have insisted that this transition was systematically linked to the introduction of fisheries resource management but that, despite its revolutionary character, it has received surprisingly little attention. Even if the papers that back up these propositions are deemed to be incontrovertible – but you must, of course, be the judge of that – the whole notion of an invisible revolution in the fisheries still seems a bit far-fetched, doesn’t it? The tension that resides at the heart of our oxymoron, between the conspicuousness of a revolution and the claim that it was invisible, is far too strong. The whole concept – and thus my project – is about to implode. Something is surely lacking here. What can it be? How can the notion of an invisible revolution in the fisheries be stabilized? The simple answer here is “an explanation.” If the notion of the introduction of fisheries resource management as an invisible revolution is going to hold as a plausible version of reality, indeed, as the most plausible version, I will have to supply a convincing account of why such a conspicuous thing as a revolution has remained invisible to so many for such a long time. What should such an explanation look like? Consider again the example of the Main Agreement as reported above. When Bjørn Brochmann attacked the subsidy system, it was because, within the new situation as established by the 200-mile EEZ and made understandable by bio-economic theory, it had become blatantly irrational. In Brochmann’s analysis, the basic realities of the fisheries, as they now appeared, made the subsidy system into an instrument that produced almost exactly the opposite result to what had been intended. As bio-economic theory showed, subsidizing the fishery could only lead

18

The Invisible Revolution

to overcapacity and inefficiency. Regarded as investments in the fishery, the monies spent through the subsidy system could only give negative returns. Subsidizing the fishery through the Main Agreement thus represented one of the most stupid and counter-productive policies imaginable. The only possible conclusion, which Brochmann drew and tried to have implemented, was that the subsidy system, established by political agreement and thus open for change, had to give way because it confronted a harder piece of reality, the nature and dynamic properties of fish stocks under exploitation. As we already know, Brochmann did not succeed. His account of the reality and logic of the fishery was rejected by the fishery sector. It would take another ten years for this account to gain credibility and become the officially accepted rendering of the situation in and problems of the fishery. I shall have more to say about this in a little while. The important thing now – the reason this story is given as a clue to the explanation of our basic puzzle of an invisible revolution – is that it provides us with an illustration of the logic of institutional transitions. It is a well-known proposition within institutional theory that institutions are taken-for-granted and that they come with accounts that make sense of their origin and function (Powell and DiMaggio, 1991; Jepperson, 1991). I would add, following Douglas (1986), that these two aspects of institutions, which would seem to be contradictory, are closely knit together. The taken-for-grantedness of institutions cannot simply be seen as an absence of conceptualization and theory, a complete lack of imaginable alternatives. On the contrary, it is better understood as a result of intense work, rhetoric and persuasion, in which the construction and dissemination of credible accounts play an important part. The ultimate achievement here is to naturalize the account, to make it lose its fictional character, and instead have it accepted as simply the true representation of the nature of things. The best story is no story at all, only facts and reality. Brochmann’s account of the logic of Norwegian fisheries is a nice example of how such stories can be set up, and illustrates the type of work they are designed to do. It is told in a realist mode, as a story about the shape of reality. In doing this, it borrows support from the bioeconomic model of the meeting of the fish, dictated by their biology, and the fisher, dictated by the logic of rationality and market forces. Within this story, there is no space for the Main Agreement and the subsidy system except as misconception and irrationality. It becomes a misunderstanding, a historical blunder, and the time has come to lay it to rest.

19

Holm

If one accepts Brochmann’s account, taking his description of the situation in the fisheries at face value, doing away with the Main Agreement is not a revolution. A revolution is a clash of societal forces, a struggle for power between contending interests, values and perspectives. Within Brochmann’s account, there are no such things, because no-one in their right mind, no person who remains sane and rational, could possibly defend the old order. There can be no quarrel with the laws of nature and the logic of economic behavior. Everyone must agree, at least if they can be made to understand the language of bio-economic analysis, that subsidizing the fisheries is stupid, as counter-productive as wetting your pants in order to stay warm on a cold winter night. In other words, the new institutional order in the fishery, constructed around fisheries resource management, comes with accounts, definitions of the world, scientific stories, which, once they are accepted as true, simply obliterate all grounds for resistance. When the transition has taken place, it cannot be regarded as a revolution. It is simply an adjustment to reality, a necessary, rational and long overdue clean-out of historical garbage. Here is the explanation I propose for the invisibility of the resource management revolution. It is a revolution; I still insist on that. But the power of the new institutional order is such that, once installed, it will rewrite its own history. Instead of a revolution, a paradigmatic shift, a fundamental change of perspectives and power relations, what emerges is a rational adjustment to reality. The integrity, logic and hence the reality of the old institutional order have been lost. No support for it can be mobilized, because its accounts, the stories that used to make it blend into reality, have suddenly been revealed as stories, as pure fiction with no credible reference to the real world. Before the new accounts of reality have taken hold, an attempt to establish a new order will appear as revolutionary. Brochmann’s attack on the Main Agreement in the early 1980s was indeed perceived in this way. The stories that made sense of the Main Agreement were still largely intact at this time. Important themes were the Main Agreement as the bulwark protecting the fisheries from fluctuations and uncertainty; the agreement’s significance as a symbolic acknowledgement of the importance of the fisheries in the Norwegian economy and society; the role of the sector in providing jobs and upholding settlement in the coastal areas. Additionally, the idea of fish as a scarce resource, the starting point of Brochmann’s analysis, was at the time not

20

The Invisible Revolution

well entrenched in the fisheries. On the contrary, the institutional structure of the fisheries was built on and promoted the opposite notion, namely that the fundamental problem in the fisheries was the overabundance of fish (Papers 1 and 3). Even though the view of fish as vulnerable and prone to overexploitation was well known, as demonstrated locally by the decimation of the AtlantoScandian herring stock during the 1960s, this view of things had not worked its way into organizational structures, decision making systems and practices within the fisheries. Precisely this was going to change during the next 10 years or so. Gradually, the ideas and definitions integral to fisheries resource management took hold and were turned into hard reality. This happened through the construction of a decision-making system for setting and allocating total quotas (i.e., the regulatory system comprising, among other things, ICES, the Regulatory Council, the Sea Border Committee, and the Ministry of Fisheries). It happened through the construction of legal instruments for distributing fishing rights among individual skippers and vessels (i.e., licenses and individual quotas). It happened through the establishment of a monitoring and control system capable of securing reasonable adherence to the rules and regulations. Some of the elements in this system had been in existence before the 1980s, and some of them would not be completed until the 1990s. It seems fair, however, to point out the 1980s as the decade during which these elements were linked together to such an extent that the theory of fish as a scarce resource became an operative fact of life in the fisheries (Holm and Mazany, 1995). So, at the next crossroads, when the EFTA Free Trade Agreement came up and offered an opportunity to confront the Main Agreement once more, the basis for resistance had gone and it fell like a house of cards. At this point in time, the credibility of the accounts that would have made it possible to turn this attack on a key institution within the MSO system into a crisis, an attempted revolution, had weakened dramatically. While the EFTA Free Trade Agreement and the demise of the Main Agreement certainly could be seen as an attack on the fishers’ power, it was difficult to defend on those grounds alone. The new version of reality, in which fish is scarce and subsidization irrational, had taken hold and was accepted, also by the fishers’ representatives. Hence, there was no revolution, only an adjustment to reality. The old institutional configuration had to be left behind because its time had passed. Fighting for the old order was like fighting for

21

Holm

irrationality. There could be no revolution, because the opposition was not another interest groups, but reality itself. This account surely is an adequate explanation for the invisibility of the resource management revolution. The trick is not that the two institutional alternatives facing one another were so similar that no-one noticed the transition between them. On the contrary, the invisibility of the revolution stems from the winning side’s success in projecting its contender as so far off in irrationality that it lost all sense of reality. Did we really believe that? Incredible! Yesterday’s reality has become fantasy and fiction. The past has been left behind; it is as if it never existed. This explanation – I’m sure you’ll agree – will do the job. It has the right shape, so to speak, to prevent the conspicuousness of a revolution reducing to absurdity the claim of its invisibility. There is a revolution. But since the winning side gets to rewrite history, it will not appear that way. Instead of a revolution, what is on display is rational adaptation to reality. No more, no less.

The Social Construction of Fisheries Resource Management Focusing primarily on the outcome of our invisible revolution, the central theme of this thesis could perhaps be reformulated as the analysis of the social construction of fisheries resource management. With the aid of Hacking’s (1999) analysis of social constructivism in his book The Social Construction of What?, we can go one step further in detailing what such a project should comprise. Hacking (1999:1) starts by noting that a lot of different things are said to be constructed, among others brotherhood (Clawson, 1989), child abuse (Hacking, 1991), facts (Latour and Woolgar, 1986), gender (Lorber and Farell, 1991) nature (Eder, 1996), quarks (Pickering, 1984), reality (Berger and Luckmann, 1967), and technology systems (Bijker, Hughes and Pinch, 1987). The titles of these works largely conform to the formula “The Social Construction of X,” where the items listed above fill in for the X. Hacking notes the great diversity of the things that are said to be socially constructed; child abuse seems to be very different from reality, as well as from gender and brotherhood. So what is it that all these things have in common

22

The Invisible Revolution

that make it meaningful to talk about them all as socially constructed? Hacking’s answer to this question is that the point of claiming that a certain phenomenon is socially constructed is to raise consciousness of it and open it up to debate and perhaps also to change. According to Hacking, social construction work is critical of the status quo, holding that: “X need not have existed, or need not be at all as it is. X, or X as it is at present, is not determined by the nature of things; it is not inevitable” (Hacking, 1999:6). Moreover: “X was brought into existence or shaped by social events, forces, history, all of which could well have been different” (Hacking, 1999:7). This is clearly also the point when I suggest that fisheries resource management should be seen as socially constructed. Just like social construction work on brotherhood, gender, facts, or the scientific revolution, a social construction thesis of fisheries resource management takes as its starting point a protest that this phenomenon is taken as inevitable and naturally given, instead of resulting from social, political and historical processes. Hacking goes on to say that, if everyone agrees that the X in question is contingent and the result of historical and political struggle, there is not much point in saying that it is socially constructed. So, a precondition for talking of social construction is that X is not generally seen as constructed: “In the present state of affairs, X is taken for granted; X appears to be inevitable” (Hacking, 1999: 12). Hacking’s point here is well illustrated by my rendering of fisheries resource management above. The main thesis is that it is socially constructed; it is an institutional and political system of the same kind as the institutional and political system it replaced, the MSO system. In order for this thesis to be of interest, however, the socially constructed nature of fisheries resource management must not be acknowledged. Hence, the invisible revolution. This means, then, that the notion of the invisible revolution, as promoted here, hinges crucially on the degree to which fisheries resource management is in fact taken-for-granted. But is it not fairly inevitable, you will perhaps ask, that fisheries resource management is a social institution? Can there be anyone who will doubt that? Well, the social and institutional dimensions of resource management are quite self-evident if we focus primarily on the latter part of the concept, management, involving a complex of rules, procedures and practices regarding the manner in which a given fish resource is utilized, including how the catch effort is restricted, whether by input or output controls; the way the regulation measures affect the distribution of licenses, quotas and wealth among fishers and communities; how affected parties are involved in

23

Holm

management decision-making, the degree of centralization; what types of monitoring and control systems are in place; the legitimacy of the regulations and the management system as such, and so on. All of this is clearly social, institutional and political through and through. A social construction thesis with regard to fisheries resource management will be self-evident, and therefore of no particular interest. But this changes drastically if the emphasis is moved from the management of the fisheries resources towards the fisheries resources themselves. A thesis that focuses not on the construction of management, but on the construction of fisheries resources has a much larger attention-grabbing potential. While it is obvious that management is social and political in nature, this can hardly be said of the fisheries resource, the fish, the fish stock, or the ecosystems the fish stock forms an integral part of. In order for a social construction thesis with regard to fisheries resource management to be of interest, then, it must include the notion that not only management but also the fishery resources themselves are in some way socially constructed. Such a notion, I think you will agree, is not trivially true and widely accepted. On the contrary, such a perspective has to confront great challenges and come to terms with important dilemmas. In order for the structure and dynamics of fish stocks to influence management decisions, they must first be represented in the places where such decisions are taken. Within the present resource management system in Norway, such representation is to a large extent taken care of by fisheries biologists. Information about the status of fish stocks, how healthy they are, how they stand up under pressures exploitation, how much can be harvested without risking their depletion, how the stocks can best be harvested, etc., is generated by fishery science. A social construction thesis of fisheries management, then, opens up the question of how we are to understand the knowledge produced by natural science, in this case fishery biology and stock assessment. Can we take the knowledge claims of natural science at face value? Does this knowledge represent the true nature and behavior of fish stocks? Or should we reject such knowledge claims as arbitrary and severely underdetermined by the true nature of fish stocks? Can it be argued that it is better to understand the knowledge claims of science with respect to fisheries management as socially and politically determined, instead of simply corresponding to nature?

24

The Invisible Revolution

Deconstructing Scientific Knowledge Let me recapitulate the chain of arguments that have brought us here. The notion of an invisible revolution in the Norwegian fisheries as suggested above rests on the acceptance of fisheries resource management as socially constructed. In order for this to be non-trivially true, this must mean that not only management itself, but also the fishery resources are socially constructed. Since fisheries resource management relies on the natural sciences to produce the knowledge required, this means that the “invisible revolution” thesis implies that knowledge produced by natural science, or at least that currently produced by fishery scientists for stock assessment purposes, can be seen as socially constructed. Now it becomes clearer why you should not be too eager to accept the notion of an invisible revolution. In order to do that, you need to accept an understanding of science that goes radically against the self-understanding of most scientists, indeed, that goes against the dominant perception of science within the modern world. You will, in other words, be committed to the constructivist side in the ongoing “science and culture wars” (cf. Gross and Levitt, 1994; Sokal, 1996a; 1996b; Sokal and Bricmont, 1998; Koertge, 1998; Edge, 1999). If you end up in this position, you had better do so with your eyes open. The constructivist side of the frontier is not a very hospitable place. There is no recourse here to the everyday rationality that is normally so useful for keeping faith in the reality of reality. In this strange place, you cannot even trust in the success of science. Even if it works in practice, that doesn’t have to mean it’s true. In addition, of course, there is the constant threat of ambush from the science warriors. Is this a place you really want to be? You will have to decide that for yourself. My main mission here is to warn you of the consequences of your choices. If you choose to believe in the invisible revolution, you will inevitably end up on the constructivist side of the fence. But I don’t mean to be too pessimistic. If, in spite of this, you choose to side with the invisible revolution, that does not necessarily mean that you have committed yourself to a life in irrationality and unreason. There may be a good life to be had in the inhospitable landscape of constructivism. But you need to bring adequate tools and resources. I’m talking metaphorically, of course. What you need to survives are resources in the form of concepts, ideas, analytical schemas, methodologies and theoretical perspectives. It is only if you are equipped with adequate

25

Holm

theoretical armament you will withstand the sniper shots of the science warriors. It is only if you apply the appropriate theoretical lenses that you will be able to remain sane without the calming perspective of everyday rationality. We have finally arrived at the issue that is going to occupy the next four chapters, namely how one can account for knowledge claims backed up by science without accepting them as simply corresponding to reality. How can we analyze science and scientific knowledge as social institutions? Which theoretical perspectives will allow us to survive the attempt to reject the notion that science speaks the truth? Which theoretical tool will do the trick, thus allowing us to make the invisible revolution visible? In each of the next three chapters, a contender for the job is evaluated. Chapter 2 takes neo-institutional theory to the task. Chapter 3 looks at the sociology of scientific knowledge (SSK). And Chapter 4 brings in actor-network theory (ANT). In Chapter 5, the basic arguments and findings are summarized, and we sit down in order to find out which candidates have passed and which have failed.

26

The Invisible Revolution

2. INSTITUTIONAL THEORIES

Introduction My research on Norwegian fisheries gradually turned into an attempt to understand the origin, function and dynamics of the sector’s institutional structure. This structure was intimately linked to a particular type of organizations, the Mandated Sales Organizations (MSOs), as exemplified by the fishers’ sales organizations and the exporters’ trade organizations. The way the MSO form swept across and transformed the entire fishery sector during the 1930s and 1940s was very much reminiscent of the waves of institutional reforms that are typically portrayed by neoinstitutionalists. The MSO system was, in a similar way, swept away by the introduction of fisheries resource management 50 years later. Paper 1, “The Dynamics of Institutionalization,” hence tackled the analysis of these processes from a neo-institutional perspective. While this paper does not talk about a resource management revolution, its description of the transition between the two sector systems easily lends itself to such a label: The emerging sector system has brought new practices, new power structures, and new ideologies. Where practices are concerned, the fish trade has been deregulated, while harvesting must face new constraints. The processors have been let loose to compete in international markets. To prevent them from destroying their resource base, the fishermen have been surrounded by strict bureaucratic controls. Where power structures are concerned, the coastal fishermen have lost most of their privileges. The processors, liberated from the tight reins imposed by the MSO system, have become more influential. But the new sector system has also created important new power positions for previously marginal actors, in particular the state and the scientific community. Where ideology is concerned, the MSO myth has given way to the tragedy of the commons. The image of the vulnerable fisherman exploited by merchants has been replaced by the image of the rapacious fisherman who must be restrained by science and bureaucracy. (Paper 1: 416)

With the aid of neo-institutional theory, then, there was no problem seeing the sector transition as institutional through and through. Here is no talk about confrontation between Society and Nature. Instead, the two sector systems each represent different configurations of practices, power structures and ideologies. One institutional system clashes with another institutional system. 27

Holm

Through the perspective of neo-institutional theory, the resource management revolution starts becoming visible. We can note, however, that the “Dynamics” wanted to legislate between different versions of institutionalism. The paper expressed a clear preference for a more ambitious, expanded institutionalism, one that could include power, interests and agency, thus transcending the dichotomy postulated between “technical” and “institutional” factors: The social practices that show themselves as technical, natural, and self-evident are the most heavily institutionalized and should therefore take center stage in institutional analysis (Zukin and DiMaggo, 1990; Powell, 1991), but there is a tendency within institutionalist thinking to exclude them from analysis. This is done through the division between ‘technical’ and ‘institutional’ environments. The former refers to rational, market-like phenomena, in which efficiency considerations have primacy, while the latter refers to the realm of rules, regulations, and rituals, in which legitimacy considerations dominate (Meyer and Rowan, 1977; Meyer and Scott, 1983; Scott and Meyer, 1991). This division of the world into one ‘hard’ and one ‘soft’ part excludes some of the most important phenomena in modern societies - market forces, competition, professionalization, and science - from institutional analysis. (Paper 1: 417)

The paper hence positioned itself in a specific way within the field of institutional analysis. While it approved of an ambitious institutional program, one which wants to take on a large chunk of reality, particularly phenomena that do not look like they are man-made, it charged another version of institutionalism, one that concentrates on phenomena that are obviously socially constructed, with leaving too much space to the rationalists. The author of the “Dynamics” also believed that neo-institutional theory, if correctly developed, will not only be able to tackle the analysis of reasonably hard and rational things like market phenomena and competition, but even the hardest and most rational of them all: Science. Can such a faith in the capabilities of neo-institutional theory be substantiated? Is it possible, with the aid of neo-institutional theory, to deconstruct science? As we recall from chapter 1, this is the requirement for making the invisible revolution visible again. Only if we can perceive resource management, grounded in scientific knowledge, as wholly institutional, can we see the introduction of resource management as a revolution. The author of paper 1 clearly believed that institutional theory can and should take on science, and that institutional theory could produce the invisible revolution. In this chapter, we shall try to find out whether he was right or

28

The Invisible Revolution

wrong. In order to do so, we must find out what institutional theory is, where it belongs in the larger theoretical landscape, where the different versions of it depart from each other, and what it can and cannot do.2

Mapping Institutionalism DiMaggio and Powell (1991a:1) have observed that while institutionalism purportedly represents a distinct approach to the study of social, economic and political phenomena, “... it is often easier to gain agreement about what it is not than what it is.” One important reason for this state of affairs is that there are several more or less distinct schools of institutionalist thought. Imposing order by mapping the different traditions, their development and relationships is not an easy task, however, since there is no agreement as to how the different schools of thought should be classified. DiMaggio and Powell (1991a), for instance, suggest that, in addition to the sociological institutionalism to which they themselves subscribe, there are three new institutional schools. These are the institutional economics of Coase (1937; 1960), Williamson (1975; 1985) and North (1981), the positive theory of institutions or rational choice institutionalism of Shepsle (1986), Riker (1980) and Ostrom (1990), and the international regime theory of Young (1986), Keohane (1988) and Krasner (1988). Immergut (1996) and Hall and Taylor (1998) give a slightly different characterization of the field. They distinguish among three different analytical approaches under the label of “new institutionalism.” First, there is rational choice instititutionalism. Even though Hall and Taylor (1998) suggest that it “in principle” might be warranted to identify new institutional economics as a fourth school, they argue that it overlaps so heavily with rational choice that they can be treated as one. Second, we have the sociological institutionalism that arose within organizational theory. The key contributors to this school are Meyer and Rowan (1991), Meyer and Scott (1992a) and Powell and DiMaggio (1991). Third, there is historical institutionalism within political science, to which March and Olsen (1989) and Thelen and Steinmo (1992) are key contributors.

2

As already suggested by the quotations from paper 1, I have been particularly interested in one particular brand of neo-institutional theory, that developed within organizational sociology. This chapter will focus on this version of institutionalism and only discuss other versions in order to identify its boundaries. 29

Holm

While the classifications proposed by DiMaggio and Powell (1991a), Immergut (1996) and Hall and Taylor (1998) are somewhat different, they all reinforce the impression created by the qualifier “new” in front of institutionalism, i.e., that there are considerable differences between these perspectives and the old institutionalism. Others have come to different conclusions. Nee (1998), for instance, starts out with the contrast between institutionalism in the Durkheimian tradition, which is based on methodological holism, and that of Weber, which starts with methodological individualism and context-bound rationality. In his view, part of the tradition that calls itself “new,” namely the sociological institutionalism of Meyer, Scott & Co. belongs to the old institutional tradition, whereas schools of thought that for DiMaggio and Powell (1991a) and Hall and Taylor (1998) are miles apart, namely rational choice institutionalism, historical institutionalism and DiMaggio and Powell’s (1991a) version of sociological institutionalism, are one and the same. In Nee’s view, only the latter perspectives, based on methodological individualism, are entitled to call themselves “new institutionalist.” Grendestad and Selle’s (1995) map of the new institutionalist landscape turns out to be almost a mirror image of Nee’s. They exclude from consideration the “calculus approach” (Hall and Taylor, 1996:939) of new institutional economics and rational choice. Instead they draw attention to the strong similarities between the (cultural approach to) institutionalism in organization theory and sociology, represented primarily by DiMaggio and Powell (1991a) and March and Olsen (1989), and the institutional theory developed within cultural theory, in which the key contributors are Douglas (1986), Wildawsky (1987), and Thompson (1992). It should be noted here that the cultural theory of Douglas and Wildawsky approaches the Durkheimian tradition, featuring methodological holism and neo-functionalist analysis (Douglas, 1986). In this categorization, then, the distinction between old and new institutionalism fades out and the Weberian tradition blends into the background. While Grendestad and Selle (1995) may disagree with Nee (1998) on the location of authors like DiMaggio and Powell, and draw diametrically opposed conclusions as to which tradition is entitled to the honorary label “new institutionalist,” they seem to be in complete agreement with regard to where the fundamental line of division should be drawn, namely between Durkheimian holism and Weberian individualism. So a number of different classifications of institutionalism are available. Let me now propose one that tries to locate institutional theories and their varieties in a larger theoretical

30

The Invisible Revolution

landscape. As a start, we need a map in which a few central landmarks are clearly marked. The most important of these landmarks are those overwhelming figures of Durkheim and Weber. Let us begin with Durkheim, since he was the one who defined sociology as the science of institutions: “… one may term an institution all the beliefs and modes of behavior instituted by the collectivity; sociology can then be defined as the science of institutions, their genesis and functioning” (Durkheim, 1982: 45). To Durkheim, then, institutions originated at the level of the collectivity. The collectivity cannot be reduced to the sum of the individual members of which it is composed. Hence, he insisted that institutionally defined modes of action and ways of judging are crystallized outside the individual, and are “independent of the particular individuals considered separately” (Durkheim, 1982: 45). Durkheim was a methodological holist and regarded society as sui generis, an emergent phenomenon not reducible to the aggregate properties of individuals. With Nee (1998: 4), we can perhaps compare the relationship between society and individuals in Durkheim’s theory with that between a tree and its leaves: “Leaves come and go according to the seasons, but it is the tree’s branches and trunk that over the years shape their pattern and distribution.” This position can be contrasted to the methodological individualism of Weber, who maintained that: “When reference is made in a sociological context to a state, a nation, a corporation, a family, ... or to similar collectivities, what is meant is ... only a certain kind of development of actual or possible social actions of individual persons” (Weber, 1968: 14; cf. Pope, Cohen and Hazelrigg, 1995). While Weber thus assigned ontological primacy to individuals and their activities where Durkheim had insisted on societal facts as the rock bottom of sociological explanation, his theory nevertheless was one of social institutions. Individual rationality and choice must, according to Weber, be understood within the institutional framework of a given society: customs, conventions, norms, religious and cultural beliefs, households, kinship, ethnic boundaries, organizations, community, class, status groups, markets, law, and the state (Nee, 1998). Now, the distinction between holism and individualism does not exhaust the differences between Durkheim’s and Weber’s approaches to the study of institutions. This pair of founding fathers also differ as to the kind of existence they assigned to the building blocks from which society is built. Durkheim’s first and “most basic” rule of method is “to consider social facts as things” (Durkheim, 1982: 60). “A thing,” said Durkheim (Ibid: 36), “is any object of knowledge which is not naturally penetrable by the understanding.” In other words, things stand in opposition 31

Holm

to ideas (ibid:35-36). To Durkheim, then, the stuff of society has a separate existence prior to and independent of people’s consiousness of it. In explaining even such a seemingly subjective and eminently private act as suicide, for instance, Durkheim proceeded by “disregarding the individual as such, his motives and ideas” in order to seek directly its social causes (Durkheim, 1951; cited in Pope, Cohen, and Hazelrigg, 1995: 230). Durkheim, at least in his methodological writing, was a materialist and a realist. (Pope, Cohen, and Hazelrigg, 1995: 232; Lukes, 1982). In contrast, Weber insisted that behavior must be understood in terms of individual motives and subjective meanings: “A correct causal interpretation of a concrete course of action is arrived at when the overt action and the motives have been correctly apprehended and at the same time their relation has become meaningfully comprehensible” (Weber, 1968: 12). With his emphasis on verstehen, Weber came close to the hermeneutical-historical tradition, as against the classicist and positivist position of Durkheim (Guneriussen, 1996: 103-105). If Durkheim was a materialist, Weber approached an idealist position. Now, this attempt at labeling is very rough and will not hold up to closer scrutiny. For instance, while Durkheim took a materialist/realist position in his early methodological writings, he moved towards an idealist position in later works, in particular in his sociology of religion (Durkheim, 1995; Guneriussen, 1996: 293). And Weber sought to combine the hermeneuticalhistorical method with positivist causal explanations and came, for instance, in his work on economic sociology, quite close to a materialist/realist position, explaining market phenomena based on notions of individual interest-seeking behavior (Guneriussen, 1996: 104; Pope, Cohen and Hazelrigg, 1995: 233-234).

32

The Invisible Revolution

Idealism

Materialism

Durkheim

Holism

Individualism

Weber

Figure 1: The two defining dualisms of social science, holism versus individualism and idealism versus materialism, as established by Durkheim and Weber.

Had this been an essay in the history of social science, I would have had to go into details here. It is not, however, and the brief remarks already made must suffice. The point is that Durkheim and Weber between them mark out a space within which the main brands of current institutionalist perspectives can be located. So the two dimensions on our map of possible institutionalisms are: First, that between individualism and holism; second, that between idealism and materialism. Figure 1 shows the resulting map, with Durkheim’s and Weber’s positions tentatively suggested. We can also note that the structuralisms of Radcliffe-Brown and Parsons belong in the upper right-hand square, while neo-classical economics can be found as far down and to the right as it is possible to come. On the idealist side, we can perhaps give the Hegelianinspired hermeneutics of Dilthey as an example of idealist holism, while Garfinkel’s ethnomethodology resides in the lower left-hand corner (cf. Guneriussen, 1996). As this suggests, Durkheim and Weber between them mark out a landscape within which most varieties of social theory can be located. The next issues, then, are where we find 33

Holm

institutional and neo-institutional theories, and, in particular, where the variety of neo-institutional analysis that has developed within organizational sociology is located. We begin with a short review, focusing on some of the key contributions.

Review: New Institutionalism in Organizational Sociology During the last decade or so, we have witnessed a renewed interest in institutions across the social sciences. Nevertheless, no agreement has emerged about what should count as an institution, how they are created, how they change, or about the proper methods for studying them. Here, we shall concentrate on the particular variety of institutionalism that has informed part of my work, namely the new institutionalism of organizational analysis. While the Carnegie School’s critique of rationality (Simon, 1976; March and Simon, 1958) is often mentioned as an important precursor of the institutionalist agenda within organizational sociology (Immergut, 1998), this brand of new institutionalism is usually traced back to the contributions of Meyer and Rowan (1991), Meyer and Scott (1992a), Zucker (1991, 1983), and DiMaggio and Powell (1991b). This work continued the long-term movement within organizational theory of “opening up the organization;” perceiving it as an “open system” interacting with and formed by its environment (Scott, 1992c). However, the interest in institutions represented a break with the dominant perspective within organization theory on the nature of the interdependencies between the organization and its environment, in which efficiency concerns were postulated as the driving force, and variations in formal structure were regarded as rational adaptations to environmental conditions (Barley and Tolbert, 1997). The early new institutionalist contributions pointed out that this perspective, focusing on the management of complex relational networks and the exercise of coordination and control, had neglected legitimacy concerns as an alternative source of formal structure (Meyer and Rowan, 1991; Scott and Meyer, 1992; DiMaggio and Powell, 1991b). Instead of regarding formal structures in instrumental and efficiency terms, Meyer and Rowan proposed that, in modern societies: … the elements of rationalized formal structure are deeply ingrained in, and reflect, widespread understandings of social reality. Many of the positions, policies, programs, and procedures of modern organizations are enforced by public opinion, by the views of important constituents, by knowledge

34

The Invisible Revolution

legitimated through the educational system, by social prestige, by the laws, and by the definitions of negligence and prudence used by the courts. Such elements of formal structure are manifestations of powerful institutional rules which function as highly rationalized myths that are binding on particular organizations. (Meyer and Rowan, 1991:44)

An organization, then, is not free to develop the structures that fit its unique technological mix and exchange relationships. Instead, it is suspended in a web of values, norms, rules, beliefs, and taken-for-granted assumptions. Such cultural elements provide blueprints for organizing by specifying the structures and procedures that are appropriate for a particular type of organization (Barly and Tolbert, 1997). Following this program, institutionalists typically seek out organizations that display features that cannot be readily explained from a rationalist perspective. A good example is the work of Meyer and colleagues on educational organizations (Meyer and Scott, 1992a). From a rational perspective, schools are highly dysfunctional. They do not control and coordinate their work processes well, and they are unable to buffer activities from their environments. This conclusion is something of a paradox, however, since schools are spectacularly successful judged by the large and stable allocation of money to them; the stability of their programs and personnel; their infrequent failures; and the generally high level of participant and constituent satisfaction with them (Meyer, Scott and Deal, 1992: 48-49). The paradox is resolved by looking at school organizations as reflections of institutionalized rules in their environments. From such a perspective, it becomes easy to explain the remarkable structural homogeneity of schools within the American school system. Schools, and the individuals within them, acquire an understanding of educational processes and division of labor from participating in the same institutional environment, sharing the same educational “culture.” To a large extent, this homogenization of organizational structures works through the adoption of institutionalized rules that define what a school is: ... some of these rules are generalized cultural beliefs (such as definitions of teacher roles, and such categories as reading and mathematics), some are requirements enforced by occupational associations (such as tenure rules), and some are mandated by state and federal legislation (such as certification and accreditation requirements). Schools conform to these rules because it is adaptive for them to do so: Their survival and resources depend upon their conformity with institutional requirements. (Meyer, Scott and Deal, 1992: 55)

35

Holm

Following this pattern, a typical new institutionalist study focuses on the effects of institutionalized rules and standards on individual organizations within a sector (Scott, 1995: Chapter 6). For instance DiMaggio (1988) has studied how variation in the institutional environment – between scholarly book publishing and public television – is associated with differences in internal organizational structures and processes, and Fligstein (1990) have showed how changes in the institutional environments of the largest American industrial firms are reflected in their strategies and structures. Another variety of this type of study is one that focuses on the diffusion of a given institutionalized form within a given population of organization. Thus, Tolbert and Zucker (1983) studied the diffusion of the civil service reform in US municipalities; Fligstein (1985) analyzed the spread of the multidivisional form among large US firms; Scott and Meyer (1991a) traced the rise of training programs in firms and organizations; and Røvik (1998) looked at the way institutionalized forms like co-worker counseling, management by objectives and total quality management “travel” in and out of organizations. Most of these studies focus on the diffusion of a given institutionalized form, assuming that little happens to it as it is adopted by the organization, and indeed that the core activity of the organization largely remains unaffected by its introduction. This is closely connected to the analytical dichotomy between technical and institutional environments, between the structures and processes related to coordination and control of technical activities and exchanges on the one hand, and the organization’s symbolic and ritualistic communication of its conformity and commitment to institutionalized standards and values, on the other. The demands on the organization from its institutional environment will often be internally contradictory as well as in conflict with the demands made by its technical environment. Instead of choosing between them, or attempting to reconcile them, the organization will try to have it both ways, by decoupling structural elements from each other and from activities, and by relying on “the logic of confidence and good faith,” and “ceremonial inspection and evaluation” (Meyer and Rowan, 1991: 57-60). Such ideas, closely related to the concepts of loose coupling (Weick, 1969; 1976) and organized anarchies (Cohen, March and Olsen, 1972), reverbate strongly in Brunsson’s (1989) work on organizational hypocrisy and Røvik’s (1992; 1998) concept of the “multistandard organization.” A radical separation is maintained between the world of action, rationality, machines and things on the one hand and the world of myths, rituals, talk and ideas on the other. The institutionalized bits

36

The Invisible Revolution

and pieces that stand in the center of these analyses belong to the latter world. As such they are not confronted with the hard “realities” of organizational life when they are adopted by organizations. While the thrust of the new institutionalist contribution has taken the institutional environment as an independent variable and studied its effect on individual organizations, there are also a number of studies that focus on the structures and processes of the institutional environment itself. DiMaggio and Powell’s (1991b) paper on “Institutional Isomorphism and Collective Rationality in Organizational Fields” is important in this respect. The paper starts out in the typical new institutionalist rejection of rationalist theory, suggesting that the real question is not about variation among organizations, as conventional theories of organization will have it, but how the striking homogeneity of organizational forms and practices can be explained. The answer, DiMaggio and Powell suggest, lies in a shift in the main force that drives organizational change, from competition and the need for efficiency, to the pressure towards homogeneity that arise within highly structured, constraining organizational fields. Referring to Gidden’s (1979) concept of “structuration,” DiMaggio and Powell suggest that the institutional definition of fields consists of four parts: … an increase in the extent of interaction among organizations in the field; the emergence of sharply defined interorganizational structures of domination and patterns of coalition; an increase in the information load with which organizations in a field must contend; and the development of a mutual awareness among participants in a set of organizations that they are involved in a common enterprise. (DiMaggio and Powell, 1991b: 65)

Within such institutionally defined fields, there will be strong pressures towards homogenization. This process is, according to DiMaggio and Powell, best captured by the concept of isomorphism, defined as a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions (1991b: 66). The authors then go on to identify three mechanisms of “institutional isomorphic change.” These are, first, mimetic isomorphism, by which organizations seek to avoid uncertainty by copying or modelling themselves on organizations that are perceived to be particularly successful, legitimate, or “modern.” Second, we have coercive isomorphism, resulting from pressures exerted on organizations by other organizations upon which they are dependent. The main example here is of 37

Holm

course the state and its agencies, which exert coercive pressures through legislation. Third, there is normative isomorphism, which stems primarily from professionalization. Professionalization is an important source of isomorphism, both by means of its cognitive base, produced by university specialists and transmitted by formal education, and the growth and elaboration of professional networks through which new models diffuse rapidly. The three mechanisms identified by DiMaggio and Powell (1991b) have received considerable attention in the later work of new institutionalists. The first, mimetic isomorphism, is of course identical to the conformity drive that leads organizations to (passively) adopt the organizational forms that are institutionalized within an organizational field. In this context, however, and particularly since DiMaggio and Powell founded their argument on Gidden’s concept of “structuration,” it is appropriate to point out that the organizations themselves actively contribute to the institutionalization of the organizational forms they adopt. Hence, Whitley (1992) emphasizes firms’ strategies and choices in the emergence of distinctive “business receipts,” that is, the particular ways of organizing, controlling and directing business enterprises that become established as the dominant forms of organization in different societies. And Fligstein (1990:6), accepting DiMaggio and Powell’s definition of fields, emphasizes how these are set up to benefit their most powerful members. DiMaggio and Powell’s (1991b) second isomorphic mechanism, that of coercion, has not in itself received much attention from new institutionalists. As Zucker (1987: 444) remarks, this mechanism does not fit easily within the new institutionalist program of taken-for-granted scripts, since the use of sanctions indicates that other attractive alternatives exist and are considered. However, under the label of the coercive mechanism, DiMaggio and Powell mentioned the state as an important source of isomorphic pressures, through legislation, allocation of resources, and imposition of taxes (Scott, 1995: 1993-95). The importance of the state in structuring organizational fields is so great that it is treated in almost every study of institutionalization, be it the educational system (Meyer and Scott, 1992a), organization of large firms (Fligstein, 1990), double entry bookkeeping (Carruthers and Espland, 1991), or the definition and enforcement of property rights (Campbell and Lindberg, 1990). However, new institutionalists have also taken up the construction of the state and its institutionalization as the key actor in Western society, for

38

The Invisible Revolution

instance, in the work of Thomas and colleagues (Thomas and Meyer, 1984; Thomas et al., 1987). The third mechanism mentioned by DiMaggio and Powell (1991b), professionalization, has received some attention from new institutionalists. DiMaggio (1991), for instance, has studied the construction of US art museums as a professional project, and Scott (1992a) has looked at the role of professions in health care organizations. As is the case with the role of the state in the structuration of organizational fields, the role of professionals in field structuration is so pervasive that it is touched upon by a number of studies, for instance, Fligstein (1990), in his analysis of the changing “conception of control” within the business sector, and Brint and Karabel (1991) in their analysis of the transformation of American junior colleges. The role of professionals, in particular in the position of business consultants, also figures heavily in Czarniawska and colleagues’ work on the “translation of organizational change” (Czarniawska and Sevón, 1996). What can we, after this all too brief and eclectic review, say about the common core of the new institutionalism in organizational analysis? We can note that this program has a particular empirical focus, as suggested by the label “organizational.” It does not want to take on all types of institutions, but concentrates on formal organizations in the public and economic sectors. We can note that it is heavily dominated by Americans, with a small Scandinavian branch. It is also easy to see that this school of thought represents a reaction against methodological individualist and rationalist perspectives, which are particularly predominant within organizational theory and business economics. Hence, DiMaggio and Powell identified the new institutional school in opposition to rational choice and new institutional economics: The new institutionalism in organization theory and sociology comprises a rejection of rational-actor models, an interest in institutions as independent variables, a turn toward cognitive and cultural explanations, and an interest in properties of supraindividual units of analysis that cannot be reduced to aggregations or direct consequences of individuals’ attributes or motives (DiMaggio and Powell, (1991a: 8).

This would suggest that this branch of institutionalism is part of the broader movement in social theory, as against rationalism and methodological individualism and towards cognitive and cultural factors. We also recognize a social constructivist mode of analysis, as suggested by Hacking (1999; Chapter 1 above). That is, organizational institutionalists want to examine taken39

Holm

for-granted phenomena within organizational life and demonstrate that they are socially constructed. Where do these identifiers place organizational institutionalism on our map? A preliminary diagnosis – indicated in particular by its anti-rationalism, its interest in institutions as independent, collective phenomena, and its preference for cognitive variables – would be that it belongs to the idealist-holist, upper left-hand corner. Let us now take a closer look at the boundary-defining work that secures this position.

Organizational institutionalism versus rational choice institutionalism A stong feature of the new institutionalist perspective within organizational sociology is its strong rejection of rationalist theories. A typical new institutionalist piece identifies its particular contribution in contrast to that of a rationalist alternative, primarily neo-institutional economics, or rational choice institutionalism. Within the latter traditions, institutions are perceived of as frameworks of “rules, procedures and arrangements” (Shepsle, 1986), or “prescriptions about which actions are required, prohibited or permitted” (Ostrom, 1986). Individual actors, with preestablished interests, are taken as a starting point and institutions are understood as solutions to problems of collective action. Institutions are regarded in instrumental terms, as the results of human design. While such institutions may strike back at individuals, regulating their behavior, they do not fundamentally affect their identities or interests, which remain pre-established and are taken as naturally given and universally constant. DiMaggio and Powell (1991a) sum up the contrast between the new institutionalism in organization theory and rational choice and new institutional economics in terms of the following four points. First, while institutions within the rational school are defined as (technical) rules or conventions, emphasizing the “regulative pillar” of institutions, as Scott (1995) would have it, new institutionalists in organization theory only accept as institutions conventions that take on a particular character. This is in line with the old sociological institutionalism, in which an organization only becomes an institution when it has lost its utilitarian character, and has been infused “with value beyond the technical requirements of the task at hand” (Selznick, 1957: 17).

40

The Invisible Revolution

Second, new institutionalists within organizational theory have a very different view of rationality and decision-making than institutionalists belonging to the rational school. Whereas the latter may accept that there are limits to rationality, for instance, in the form of cognitive limits, time and resource constraints, uncertainty, etc (Williamson, 1975), they retain the idea of actors as rational decision-makers and strategists (Immergut, 1998). The new institutionalists within organizational theory, in contrast, depart much more radically from the idea of a rational actor. Instead of the actor as a decision-maker, who weighs different alternatives against each other in terms of their (probable) outcomes, the actor is for new institutionalists much more sensitive to the requirements of the situation in which s/he finds her/himself. Starting from the taken-for-granted expectations that are associated with a given situation, actors will conform to what they believe is the appropriate or prescribed course of action (March and Olsen, 1989: Chapter 2; Immergut, 1998; DiMaggio and Powell, 1991a: 10). Third, institutionalists within the rational school tend to see institutions as efficient solutions to collective-action problems. In March and Olsen’s (1989) terms, they tend to assume that history is efficient and hence that existing institutions are optimal responses to the problems they were created to solve (cf. Perrow’s, 1986, critique of Williamson, 1975). For new institutionalists, in contrast, institutions are slow to change. This follows from the premise that institutions are grounded in taken-for-granted accounts and cognitive categories. Further, institutions are often created in path-dependent processes, and will thus not be equilibrium solutions to given problems. This is not only because institutions constrain choice and prevent actors from pursuing certain courses of action, a point which rationalists will readily admit. In addition, sociological institutionalists will emphasize that institutions constitute the criteria by which actors discover their preferences and interests; indeed, they constitute who the actors are, be they individuals, firms or states. Fourth, rational choice institutionalists and new institutional economists adhere to the litany of methodological individualism, and will hence view institutions, at least in principle, as aggregate consequences of individual behavior. New institutionalists in organization theory, in contrast, are primarily interested in institutions as independent phenomena, phenomena that cannot be explained by reference to some lower, fundamental level of analysis (March and Olsen, 1989). New institutionalists are primarily interested in macro-level phenomena – structures and

41

Holm

processes that are industry-wide, national, international and global in scope – and focus on homogeneity, repetitiveness and diffusion rather than differentiation, innovation, and change (DiMaggio and Powell, 1991a:9).

Idealism

Holism

Materialism

Organizational institutionalism

Individualism Rational choice institutionalism Figure 2: Organizational institutionalism defines itself in radical opposition to rational choice institutionalism, located in the materialist/individualist corner, thus positioning itself in the idealist/holist square.

These four points establish an insurmountable gulf between organizational institutionalism on the one hand, and rationalist choice institutionalism, located in the lower right-hand corner of our map, on the other, as suggested in Figure 2. The former’s twin attack on individualism and materialism, summed up in the rejection of the naturalistic version of the rational actor, suggests that this type of perspective is safely located in the holist/idealist upper left-hand square.

42

The Invisible Revolution

New versus old institutionalism Powell and DiMaggio (1991: 11-19) spent considerable effort in substantiating the claim that there is a “considerable gulf” between the old and the new institutionalism. Whatever the new institutionalism is, no-one should be in doubt that it is different from the old one. Now, within organizational sociology, old institutionalism is identified with the work of Philip Selznick (Powell and DiMaggio, 1991:12; cf. Perrow, 1986: Chapter 5), in particular his TVA and the Grass Roots (Selznick, 1980), and Leadership in Administration (Selznick, 1957). As argued by Perrow (1986:157) the conceptual framework of this brand of institutionalism is structuralfunctionalist, “indicating that functions determine the structure of organizations and that structures can be understood by analyzing their functions.” This, of course, places the old institutionalism at the holist end of our map (Figure 3, p. 44). In the same way as Parsons’ version of structuralfunctionalism, the old institutionalism is also materialist. To see this, we can start with the key distinction within the old institutionalism between rational, means-oriented, efficiency-guided organizations and value-laden, adaptive and responsive institutions. According to Selznick, to institutionalize is “to infuse with value beyond the technical requirements of the task at hand” (1957: 17). Powell and DiMaggio (1991) here point out that within this perspective, rooted in a Freudian ego psychology via Parsons’ action theory, institutionalization is understood within a moral frame of reference as a process in which value commitments are transmitted to members through socialization and internalization. As against this “hot” imagery of values and affection, the new institutionalism – reflecting the cognitive turn in the social sciences – emphasizes taken-forgranted scripts, routines, and classifications: “Normative obligations ... enter into social life primarily as facts” (Meyer and Rowan, 1991: 42). Instead of institutionalization as a process involving values, goals, interests and commitments, connecting to actors’ material positions in society and lending itself to causal analysis, new institutionalists emphasize cognitive and ideational phenomena: Scripts, concepts, myths, ceremonies, and rituals. Where the old institutionalism was realist and materialist, the new one is idealist (Figure 3, p. 44).

43

Holm

Idealism

Holism

Materialism

Organizational institutionalism

Old institutionalism

Individualism

Figure 3: Organizational institutionalism defines itself as neo-institutional, in sharp opposition to the old institutionalism, which is located in the materialist/holist square.

Varieties of Organizational Institutitionalisms While organizational institutionalisms’ quarrel with rational choice and neo-institutional economics pushes it towards the holist upper half of our map, its quarrel with the old institutional school pushes it towards the idealist left-hand half of our map. But our attempt to grasp organizational institutionalism does not end here. While the boundary work that intends to set organizational institutionalism apart from the functionalist and rationalist versions of institutionalism creates the impression of strong common features in organizational institutionalism, there is in fact great variation and tension among those pursuing neo-institutionalist analysis here. As we shall see, the question can also be raised whether the different versions of organizational institutionalism should be seen as one or several distinct schools of thought. Here, we should note that the lines delimiting these boxes in Figures 1, 2 and 3 are dotted, indicating that there are no impenetrable barriers to be found here. Furthermore, it does not make much sense to think that these lines can be located

44

The Invisible Revolution

in precise, absolute positions. Instead, such positioning can only be relative. Labels such as holism, individualism, idealism and materialism are used to create boundaries between one’s own position and others’. While broad differences in location are possible to chart, no precise metrology exists that allows precise positioning. Now, the version of organizational institutionalism that I opted for and applied in my analyses of the MSO system was the more ambitious “expanded institutionalism” of Powell and DiMaggio. Despite these authors’ attempt to establish one unified institutionalist agenda, their own version of it remains quite distinct from (at least) two other schools of thought. These are, first, the macro-oriented institutionalism of Meyer and his co-workers, and, second, the ethnomethodologically inspired institutionalism of Zucker and her co-workers. The three schools all seem to share a starting point somewhere in the upper left-hand idealist/holist square, focusing on institutions as cognitive and cultural constructs that have existence independent of the activities and perceptions of individual actors. In elaborating institutional theory, however, they take different directions. As a preliminary sketch (Figure 4, p. 46) we can say that Meyer & Co., after a dualist start, move further up towards the left, ending up in an extreme version of idealist structuralism. Zucker & Co, inspired by ethnomethodology, work their way downwards, towards the subjectivist left-hand corner. DiMaggio and Powell, in contrast, want to go downwards to the right, bringing them closer to the individualist/ materialist square. What follows now is a more detailed discussion of these movements, their motives, and their difficulties.

45

Holm

Idealism

Materialism

Holism Meyer & Co. Zucker & co

DiMaggio & Powell

Individualism

Figure 3: The relative movement of three different schools of organizational institutionalism. Meyer & Co. (Institution-as-environment) go towards the idealist/holist corner; Zucker & Co. (Institution-as-organization) go towards the idealist/individualist corner; DiMaggio and Powell (Expanded institutionalism) go towards the materialist/individualist corner.

46

The Invisible Revolution

Institution as Environment (Meyer, Scott & Co.)

Early dualist version The typical neo-institutional piece explicates its particular contribution in contrast to a rationalist alternative. In the Meyer/Scott version of organizational institutionalism, the contrast with rationalist theories and perspectives takes a specific form. In order to see this, we shall go back to the paper which launched this tradition, namely Meyer and Rowan’s “Institutionalized Organizations: Formal Structure as Myth and Ceremony” (Meyer and Rowan, 1991, originally published in 1977). The authors here start by contrasting two radically different ways of understanding and theorizing the rise and function of rationalized formal structure. According to conventional theories, which Meyer and Rowan trace to Weber, “rational formal structure is assumed to be the most effective way to coordinate and control the complex relational networks involved in modern technical or work activities” (Meyer and Rowan, 1991: 42). Here, then, is the rational theory, against which Meyer and Rowan’s neo-institutional alternative is pitted. The problems with the rationalist theory, they go on to say, is that it is premised on organizations actually functioning according to their formal blueprints. As empirical research shows, however, there are great gaps between formal and informal organization: structures are only loosely linked to each other and activities; rules are often violated; decisions often remain unimplemented, and so on. Formal organizations are endemic to modern societies and it is important to be able to explain why they arise: Such an explanation should account for the elaboration of purposes, positions, policies, and procedural rules that characterizes formal organizations, but must do so without supposing that these structural features are implemented in routine work activity. (ibid: 43)

Enter institutional sources of formal structure. In elaborating this alternative account of the rise of formal structures, Meyer and Rowan start with Weber’s “neglected” work on the legitimacy of rationalized formal structure. In this view, formal structures are not only means to ends, but also “manifestations of powerful institutional rules which function as highly rationalized myths that are binding on particular organizations” (Meyer and Rowan, 1991: 44). Instead of

47

Holm

seeing formal structures as established solutions to the problems organizations face when trying to realize their goals, Meyer and Rowan propose viewing them as reflections or manifestations of “instititionalized rules,” or “rationalized myths.” What kinds of creatures are these? With reference to Berger and Luckmann (1967), Meyer and Rowan define instututionalized rules as “classifications built into society as reciprocal typifications or interpretations” (ibid: 42). But we should note here that Meyer and Rowan are not much interested in the way such “reciprocal typifications” emerge out of interaction; how they, in Berger and Luckmann’s terminology, get to be externalized and objectified. Instead, Meyer and Rowan focus on such rules and myths as predominant properties of modern societies, after institutionalization and rationalization have already occurred. Berger and Luckmann portrayed the relationship between individual and society as dialectical, an ongoing process in which man is a social product and society is an ongoing human production. In Meyer and Rowan’s organizational institutionalism, it is the first moment of this dialectical relationship that gets all the attention, at least during the early phase. In this way, the institutional perspective of Meyer and Rowan portrays the formation of formal structure as a top-down process, where the causal arrow goes from macro society to micro organization: “institutional theories in their extreme form define organizations as dramatic enactments of the rationalized myths pervading modern societies” (ibid: 47). Let us, before we go any further in the analysis of this version of institutionalism, locate it in our map. As a starting point, we can note that the rationalist theory of formal structure against which the institutional theory is portrayed belongs to the materialist-individualist lower right-hand square. The emphasis here is on technical problem-solving related to work-processes and taskoriented exchange relations between the organizations and their environment. Formal structures grow out of such processes as efficient solutions to practical or technical problems. The institutional theory of Meyer and Rowan positions itself diagonally opposed to this, in the idealist/holist upper left-hand square. The emphasis here, as already mentioned, is overwhelmingly on the way society, in the shape of institutionalized rules or rationalized myths, influences organizational forms and individual behavior. The fundamental theory of how society is constituted, as based on the phenomenology of Berger and Luckmann (1967), is idealist. The shared, taken-for granted typifications have a cognitive basis; shared knowledge forms the building blocks of society. Reality itself is constructed out of ideational stuff like shared typifications, scripts and rationalized myths.

48

The Invisible Revolution

Now, this version of institutionalism has frequently been criticized as being “restricted” or “narrow.” The main thrust of this criticism, to which we will return in much more detail later, is that it focuses exclusively on the way already established institutions, understood as macro-constructs, shape and constrain action, while disregarding the processes by which institutions are created and change. For instance Friedland and Alford (1991: 243-4), have claimed that institutionalists “ ... have not begun to study the way institutional arenas are patterned in the way that they are or the conditions under which new institutional forms develop.” In the same vein, Barley and Tolbert say: With the passage of time, Meyer and his colleagues have come to treat institutions primarily as exogenous to organizational action (Meyer and Scott 1983; Scott and Meyer 1987; Sutton et al. 1994; Scott and Meyer 1994). Their more recent work associates institutional pressures with the demands of centralized authorities or regulatory agencies and, only secondarily, with widespread beliefs, practices, and norms. Consequently, their research programme has focused on the causes and consequences of conformity and on the manner in which the environment ‘interpenetrates the organization’ (Meyer and Rowan 1977). The problem with this approach is that it depicts an institution as somehow distinct from those who comply and, more importantly, from the act of compliance itself. The result reifies the notion of institution. (Barley and Tolbert, 1997: 95)

In other words, the institutional program of Meyer and Scott is restricted because it only has something to say about how institutions affect individuals and organizations, but remains silent about the way rules and myths come to be institutionalized and taken for granted. I believe that this criticism is to the point, except that I would have addressed it to the early works of Meyer, Scott & Co. rather than the later ones, as Barley and Tolbert do. I will return to this point below. Nevertheless, I do not think that this problem represents the only restriction and is perhaps not even the most serious one, on organizational institutionalism in Meyer & Scott's version. To see this, we can start by exploring how, according to Meyer and Rowan (1991), institutional and rationalist explanations of formal structure are related to each other. While Meyer and Rowan in their article greatly emphasize the institutionalist perspective, they do not argue that institutional explanations can completely replace rational explanations in accounting for the rise of formal structures. On the contrary, they propose that institutional and rational explanations are complementary, and that they between them can do the job. The key idea here is that all organizations, to different degrees, “are embedded in both relational and institutionalized contexts and are therefore concerned with both coordinating and controlling their 49

Holm

activities and with prudently accounting for them” (ibid: 54). Some organizations use clearly defined technologies and produce outputs that can easily be evaluated. In such cases, markets will develop, and customers may gain considerable rights to inspection and control. When this happens, efficiency often determines success. Other organizations, such as schools, R&D units, and government bureaucracies, use ambiguous technologies or produce outputs that are difficult to appraise. Here, internal participants and external partners may call for institutionalized rules that can promote trust and confidence in outputs: Thus, one can conceive of a continuum along which organizations can be ordered. At one end are production organizations under strong output controls … whose success depends on the management of relational networks. At the other end are institutionalized organizations whose success depends on the confidence and stability achieved by isomorphism with institutional rules. (Meyer and Rowan, 1991: 55)

According to this perspective, then, institutional and rational theories of organizations are not based on fundamentally different ontological and epistemological grounds, as Barley and Tolbert would insist. Instead, institutional and rational theories are two equally valid perspectives that can be used to account for the way organizations adapt to two different types of demands from their environments. In later versions, these will be referred to as institutional and technical environments respectively. Meyer and Rowan’s criticism against rationalist explanations, focusing on the logic that applies to an organization’s adaptation to its technical environment, we now discover, is not an all-out charge of inadequacy, as is the impression one may get from a superficial reading, but a more modest warning that they should not be applied outside their proper domain. This theory thus becomes explicitly and radically dualist, allowing for institutional phenomena and explanations as well as for technical phenomena and explanations. Referring this back to our basic map, it means that Meyer and Rowan’s perspective combines two diagonally opposed positions: One institutional perspective located in the idealist/holist upper left-hand corner, and one rationalist perspective located in the materialist/individualist lower right-hand corner. It must be pointed out here that this perspective, as far as I can see, is truly dualist, and that there is no attempt – in this early version – to show how the two sides of reality come together or grow out of some more fundamental, common process. On the contrary, the core feature of this perspective is exactly that the technical and the institutional remain maximally different and stay strictly apart from each other. Thus, in their 1977 paper, Meyer and Rowan go on to discuss how organizations can simultaneously handle demands from their institutional and

50

The Invisible Revolution

technical environments. The point here is that organizations will be at the intersection of two radically different and conflicting demands: Categorical rules conflict with the logic of efficiency. Organizations often face the dilemma that activities celebrating institutionalized rules, although they count as virtuous ceremonial expenditures, are pure costs from the point of view of efficiency. (Meyer and Rowan, 1991: 55-56)

Meyer and Rowan propose that organizations can resolve, or perhaps rather escape, this dilemma by means of devices like decoupling, applying the logic of confidence and good faith, and by way of ceremonial inspection and evaluation. Without going into detail here, we can note that these are techniques that allow the organization to insulate what it has to do in order to conform to demands from its institutional environment from what it has to do in order to conform to the demands from its technical environment. In this way, the organizations are interpenetrated from two sides, both their technical and their institutional environments, yet the two will never meet within the organizations. Now, the dualist perspective proposed by Meyer and Rowan (1991) was not a freak accident, quickly left behind and forgotten as it perhaps should have been. Instead, it was elaborated and extended in later work and has become something of a trade-mark for organizational institutionalism. Thus, besides Meyer's and his co-workers’ attempts to sort out the logic of organizational sectors characterized by weak technical environments, exemplified by US public sectors like education (Meyer and Rowan, 1992; Meyer, 1992a, 1992b, 1992c; Scott and Meyer 1994b; Meyer, Scott and Strang, 1994, Meyer et al, 1994) health and aging (Scott, 1992a, 1992b; Meyer, 1994a), and local government (Meyer and Scott, 1992b), Meyer and Scott have also explicated, in a series of publications, how the distinction between technical and institutional sources of organizational structures can be used for analytical purposes (Scott, 1991; Scott and Meyer, 1991b; Meyer, Scott and Deal, 1992; Scott and Meyer, 1992; Meyer 1994b). In this line of work, it gradually became clearer how one can distinguish between technical and institutional environments and organizations’ dealings with them. Thus Scott and Meyer (1992) defined technical sectors as “those within which a product or service is exchanged in a market such that organizations are rewarded for effective and efficient control of the work process” (ibid: 140). In contrast, institutional sectors are “characterized by the elaboration of rules

51

Holm

and requirements to which individual organizations must conform if they are to receive support and legitimacy from the environment” (1992: 140). Where technical environments exercise output controls over organizations, institutional environments reward organizations for establishing correct structures and procedures (Scott, 1991: 167). At the same time, “institutional” and “technical” are no longer defined as dichotomous states, as the extreme points of one and the same scale, so that the presence of one of them necessarily would mean the absence of the other. Instead, the two are seen as two independent “dimensions along which environments vary” (Scott, 1991: 168). In this way, organizations having strong technical environments do not necessarily have weak institutional environments (as general manufacturing organizations do), or the other way around (as schools and churches). Some organizations, like banks and hospitals, would confront strong demands from technical as well as institutional environments, while others, like restaurants and health clubs, would meet strong demands from neither (Scott and Meyer, 1991b: 124). This move, it seems, is yet another way to ensure that the two fundamental but contradictory spheres of reality, the technical and the institutional, can coexist and do their work in shaping organizations, but never really confront each other. The only concession here is an admittance that it will often be difficult to distinguish empirically technical from institutional rules and procedures: “This is because those who formulate institutional rules strive to make them appear technical in nature” (Scott and Meyer, 1991b: 124), believing that this will enhance their legitimacy. But such tactics on the part of organizational people are quickly countered. To reveal whether a procedure in reality has technical merits or if it is ritual or ceremonial in nature, one must check the existence of “criteria for evaluating outcomes independent of those that assess conformity to the specified procedures” (ibid: 124) Technical rules and procedures, this seems to say, are real and can be independently verified; institutional rules and procedures are socially constructed and as such cannot be evaluated objectively. Now we can return to the issue of the way this perspective makes up for a restricted institutionalism. It is easy to see that its sharp dualism makes it restricted in another sense than that implied by Barley and Tolbert. In their critique, which takes Meyer and Rowan’s reference to Berger and Luckmann at face value, the main problem is that they only focus on one half of the dialectical relationship between individual and society. The adequate theoretical foundation for analyzing institutionalization as a process was there, this suggests, but it was not activated in the empirical analyses. But this is not quite right, I believe. The main problem with Meyer and

52

The Invisible Revolution

Rowan’s institutionalism is not a failure to activate Berger and Luckmann’s theory as to how institutions arise out of action. No, the problem is that Meyer and Rowan saw task-oriented action and exchange processes from the perspective of a conventional theory of rational behavior. The way organizational structures arise out of action and interaction on the micro-level was not left out by oversight, as suggested by Barley and Tolbert, but was simply defined as outside the realm of the institutional. The institutionalism of Meyer and his followers was indeed partial. In order to cover the full range of organizational phenomena one wanted to study, the form and function of formal organizational structure, it had to be complemented with a theory of rational action. This dualism imposes a division of organizational phenomena of two radically very different kinds. Some phenomena, i.e. rules and procedures disconnected from output evaluation and efficiency criteria, belong to the realm of the institutional and can appropriately be studied by way of institutional perspectives. Other phenomena, those that lend themselves to independent output controls and efficiency evaluation, belong to the realm of the technical and must be analyzed by way of rationalist theories. This means that a great deal of the heavy duty organizational phenomena, like market phenomena, technology systems, control and coordination activities linked to production, all those activities that in some way or another are grounded in scientific knowledge, are outside the scope of institutional analysis. What remains for institutional analysis are the softer, nonrational things, like myths, ceremonies, symbols, decisions that are never implemented, structural features that really are only for display; in short, superficial structural elements that may be important for legitimacy purposes, but otherwise represent “pure costs from the point of view of efficiency” (Meyer and Rowan, 1991: 56). While institutionalists are free to study things that are put up for display but do not really work, they have no business entering the real and rational world of technology, markets and science. How can we understand such a self-imposed restriction of the institutional program? A charitable interpretation, perhaps, is to see it as a very timid first approach in a setting completely dominated by rationalist approaches with imperialist pretensions. In this way we can perhaps understand the sub-text of Scott’s (1987) paper on the “The Adolescence of Institutional Theory.” In its first, immature state, this would seem to suggest, institutional theory was only allowed to play with the things belonging to the “soft underbelly” of organizational life (Scott and

53

Holm

Meyer, 1994a: 118), phenomena that are not that important anyway and that the older, more mature rationalists have not bothered with. When it comes of age, however, institutional theory will take on more important and demanding stuff. Perhaps this interpretation is appropriate for Meyer, Scott & Co., since they – as we shall see shortly – indeed have moved (grown up?) from the extreme dualist position to one that also opens up “technical” environments and factors for institutional analysis. But this cannot be the whole story since the radically dualist version of organizational institutionalism still lives on, in particular in parts of the Scandinavian branch of neo-institutionalism (e.g. Røvik, 1992; 1996; 1998; Brunsson, 1989). This suggests another interpretation of the dualist position, one in which the rationalist side after all takes priority over the institutionalist side. Despite the focus on institutional phenomena within the dualist perspective, one sometimes gets the feeling that it harbors a deep skepticism against them. While rationalism is routinely set up as the villain in this literature, its purported heroes – institutional phenomena – are portrayed with a severe character flaw. Systematically contrasted with rational and material phenomena, as decoupled from technical imperatives and exchange processes, they end up as mere ritual, nothing but windowdressing, and empty formalism. In this version of neo-institutional analysis, the focus is explicitly on the travel of ideas, malleable and easily transportable phenomena that enter organizations, stay around for a while, and then exit again, apparently without leaving any marks. Institutionalized standards in this conceptualization live exclusively in the world of ideas, and remain sharply distinguished from the world of materials, things and machines, a world which, although systematically under-theorized, in the undertext comes out as much more stable and governed by rationality. The theory of institutions then becomes a theory of fashions, superficial garments that organizations put on, and then take off again, without affecting what they are and how they function in any substantial way. The purpose of this move seems not to be to celebrate institutional phenomena and explore their significance, but to expose and ridicule them. Hence, the soft institutional phenomena, the questionable nature of which is indicated by such labels as myth, ritual, talk, hypocrisy, or fashion, are contrasted against the much more reliable and honest realities of machines, materials, money, efficiency and rationality. This brand of institutionalism, despite its initial rejection of rationalist theories, ends up as a revealer of empty myths and superficial talk on a rationalist basis.

54

The Invisible Revolution

Mature structuralist version While the dualist version of organizational dualism still has supporters in Scandinavia, Meyer and Scott have recognized its shortcomings and have tried to reposition themselves. Thus, Scott and Meyer (1994a) explicitly distance themselves from the radical dualism they had established in earlier work between the “institutional” and the “technical”: We acknowledge that by concentrating much of our empirical work on schools and professional service organizations, we have contributed to the misconception that institutional arguments pertain only to certain (limited kinds of) organizations. We no doubt inadvertently reinforced this error by distinguishing in our work between “institutional” and “technical” organizations and environments (Meyer and Rowan, 1977; Meyer and Scott, 1983). Our intent was to differentiate between different bases of regulating organizations – by attention to processes versus outcome controls, not to suggest that some organizations operate outside the framework of institutional forces … (Scott and Meyer, 1994a: 119)

Doing away with such “misconception” and “error,” Scott and Meyer now want to join forces with those who want to “expand the range of institutional analysis to incorporate the full range of organizations” (ibid). While this phrasing is reminiscent of Powell and DiMaggio’s expanded institutionalism, as we shall see later, this should not be taken as a move towards this program. On the contrary, Scott and Meyer’s rejection of Powell and DiMaggio’s version of institutionalism is sharp: In contrast to the arguments of DiMaggio (1988, 1991) and others who see the creation of institutions as an occasion for reviving ‘agency’ – the rational actor endowed with a preexisting set of interests to be defended and advanced by creating new cultural rules – we think the most interesting and useful forms of institutional theory depend on showing the collective and cultural character of the development of institutional environments. (Scott and Meyer, 1994a: 4)

As suggested by this quote, Meyer and Scott want to anchor their new, expanded, non-dualist institutionalism more firmly in the “collective and cultural character” of institutions, that is, in the holist/materialist upper left-hand corner of our map. Let us now have a closer look at the way they do this. Unsurprisingly, this positioning starts by evoking the contrast with a rationalist, purportedly dominant alternative: Most explanatory models produced by social theorists in recent decades take for granted much of the organizational and ideological individualism of modern society. Two foundation stones underlie these

55

Holm

models: Society consists essentially of individual actors, and social activity ordinarily involves the purposive behavior of individuals. (Meyer, Boli and Thomas, 1994: 11)

Rejecting such conceptualizations, Meyer et al., go on to suggest that rather than reifying actors and actions as naturally occurring, real entities, they should be taken as socially constructed. “Culture,” they say, “includes the institutional models of society itself” (ibid: 17). Culture in this usage is more than vague ideas about the moral and natural environments, but is what constitutes society, with actors, intentions, interests and all. “Culture has both an ontological aspect, assigning reality to actors and action, to means and ends; and it has a significatory aspect, endowing actor and action, means and ends, with meaning and legitimacy” (ibid: 17). Meyer et al. refer to Berger and Luckmann as a major source of this viewpoint. Nevertheless, they are quick to set up a defense against the subjectivist and voluntarist versions of phenomenology: This is not to say, however, that society is a matter of continuous negotiation, communication and exchange of meaning. Symbolic interactionism and the social-psychological variants of phenomenology employ an institutional view of social behavior, but they depict the meanings and rules constructing actors and action as operating at the same level of analysis as the actors affected by them. The local interplay between interaction and its meaning, whereby actors (usually people, in these formulations) continually discover and construct who they are through looking-glass feedback processes, allows for no level of reality external to the phenomenological situation itself. This tradition of research ignores the extraordinary power of exogenous institutionalized definitions of reality. (ibid: 19)

The lower left-hand idealist/individualist corner of our map is clearly not a place where Meyer et al. want to spend much time. Instead, they establish themselves in a structuralist position, granting independent reality to the structures themselves. Institutionalized rules, they claim, are “located in the legal, social scientific, customary, linguistic and epistemological, and other ‘cultural’ foundations of society.” (ibid:18). There is no dialectic relationship between action and structure, individual and society here. Meyer & al., going in the opposite direction to Giddens (1984; see below), do not allow the individual actor any foundation for “acting otherwise.” The individual actor cannot really be distinguished from the cultural script and institutional definitions from which they emerge: “Actors enact as much as they act: What they do is inherent in the social definition of the actor itself” (Meyer, Boli and Thomas, 1994: 18). The individual “is an institutional myth

56

The Invisible Revolution

evolving out of the rationalized theories of economic, political, and cultural action” (ibid: 21). Meyer and colleagues thus resolutely head towards the extreme holist/idealist corner: A fully institutional analysis, then, calls attention to the extent to which, both in the modern world polity and throughout much of Western history, the cognitive and moral frames of activity at all levels are anchored in the broadest institutional (societal and world) levels. Actors and action are illuminated by universalistic lights. (ibid: 19)

In the dualist version, the “technical” realm had been resting on a realist ontology and a rationalist epistemology, whereas the “institutional” realm had been described with the aid of an idealist ontology and a relativist epistemology. In repositioning themselves, Meyer et al. redefine the basis for the technical realm and move it wholly into the realm of the institutional. What appear to us as “technical” are merely institutional myths disguised as reality itself. Meyer et al. call attention to the capacity for rationalization within the Western cultural account, creating “an extraordinary array of legitimated actors reified as purposive and rational – individuals, associations, classes, organizations, ethnic groups, nation-states” (ibid: 26-27). In sharp contrast to the earlier dualist version, Meyer et al. now emphasize that rational processes and “technical” environments cannot be taken at face value as having “real” existence independent of institutional definitions. The task of institutional analysis is not to reproduce and reify such taken-for-granted views, but to step outside them and show how they are part of the Western cultural configuration. The “technical” collapses into the “institutional,” which, at least in its most essential form, rests on ideational constructs on the most general of levels. An organizational institutionalism with such a structuralist and idealist leaning will certainly not be restricted in the same way as Meyer & Co.’s early dualist version. All the “technical” phenomena that were formerly defined as falling outside the realm of institutional analysis can now be reimported. Indeed, such phenomena now take on particular importance precisely because they follow a key pattern of institutionalization – that of rationalization – within the Western cultural account. In this version, the criticism of institutionalism as “restricted” is clearly no longer valid. From the perspective Meyer & Co. now propose, it certainly will be possible to tackle the “social practices that show themselves as technical, natural, and self-evident,” i.e. “market forces, competition, professionalization and science” (Paper 1: 417).

57

Holm

Nevertheless, it would seem that this move throws Meyer et al. back into the criticism launched in particular by Zucker, Tolbert, and others, of being blank as to the processes of institutionalization. In their structuralist version of institutionalist analysis, Meyer et al. leave virtually no scope for actors outside of the cultural scripts that define them. Their identities, their goals, their perceptions, their rationalities, their interests, their repertoire of action, the opportunity situations they find themselves in – all these and more are institutionally defined. If the rationalist perspectives organizational institutionalists use as a counterpoint can be said to rest on an undersocialized conception of human actors, actors in this version of institutional theory suffer from severe over-socialization (Wrong, 1961; Granovetter, 1985). If there ever was a perspective that could rightfully be blamed for portraying humans as “cultural dopes,” this must be it. Such a perspective may allow for an account of how established institutions work back towards the individual level, constituting and constraining practices and action. However, it will have very little to say as to how institutions emerge; how some practices are institutionalized while others are not; why certain ideas become taken for granted while others stop being taken for granted; why institutions sometimes change and sometimes are abandoned; why the Western cultural account, with its emphasis on rationalization, becomes dominant and pushes back alternative cultural configurations.

Institution as Organization (Zucker, Tolbert & Co.) There is a variety of institutionalism that is oriented towards the lower left-hand corner of our map. This is the ethnomethodologically inspired organizational institutionalism of Zucker and Tolbert (Zucker, 1983; 1987; 1988; 1991; Tolbert and Zucker, 1983; Tolbert, 1985; Barley and Tolbert, 1997). Let us take a closer look at it, starting with a brief presentation of ethnomethodology. The inventor of ethnomethodology, Harold Garfinkel, influenced by Shutz’ phenomenology, regarded social order as a practical accomplishment, as the outcome of interaction, not a precondition for it (Garfinkel, 1967; DiMaggio and Powell, 1991a: 20; Collins, 1988: Chapter 8). Taken-for-granted knowledge cannot sustain order by itself, since it will never

58

The Invisible Revolution

be perfectly shared. While people ordinarily take it for granted that meaningful activities are going on, this assumption relies on their willingness to avoid searching for a secure or “objective” basis for interaction. A key concept here is “indexicality“, which Collins (1988: 276) explains in this way: Any particular item of significance is an ‘index’ of what lies beyond it. If we shift our attention to what lies beyond, then the same problem repeats itself: there is a further context, which is again taken for granted, and so on.

The entire social world is a set of indexicalities. For ordinary interaction to succeed, however, this lack of a truly shared or objective basis has to be ignored or smoothed over. This means that meaningful cognitions do not guide social behavior. Instead, they are imposed on social events in retrospect (Collins, 1981: 991). For Garfinkel, ethnomethodology is the study of the procedures people use to remedy the indexicality of social life. These procedures include the et cetera assumption, referring to things in the world in a short-hand mode of expression; waiting for clarification, indicating that people do not expect that everything should be clearly understood at the moment they hear or see it; and offering accounts, that is, supplying excuses or reasons that make sense of breaches in interaction (Collins, 1988: 282-3). The ethnomethodology of Garfinkel entered new institutional work through the work of Zucker and Tolbert (Zucker, 1983; 1987; 1988; 1991; Tolbert and Zucker, 1983; Tolbert, 1985). Zucker (1991) took over Garfinkel’s definition of institutionalization as the extent to which practices have ready-made accounts. Acts performed by actors occupying a specified position or role, for instance, within the context of formal organizations, are high on institutionalization. In contrast to the main current of organizational institutionalism – that pursued by Meyer, Scott & Co. – Zucker and her followers regarded the organization itself as an institution (Zucker, 1987). In accordance with the ethnomethodological program, the focus here is on the micro-construction of institutions, how they arise from interaction within organizational settings. Where mainstream institutionalists see institutions as pre-given structures, says Zucker, her own version focuses on institutionalization as an ongoing process. It may of course be debatable to what extent Zucker pursued a strict ethnomethodological program in her institutional analysis of organizational processes. To me, it seems that she was much more willing than Garfinkel would have been to allow institutions to take

59

Holm

on an objective, exterior existence, independent of the individual sense-making activities from which they emerge. This happens through the importation of Berger and Luckmann’s version of phenomenology: Institutionalized acts, then, must be perceived as both objective and exterior. Acts are objective when they are potentially repeatable by other actors without changing the common understanding of the act, while acts are exterior when subjective understanding of acts is reconstructed as intersubjective understanding so that the acts are seen as part of the external world (see Berger and Luckmann 1967 on reification and objectivation). (Zucker, 1991: 85)

This move seems to allow a bridge over the individualist/holist divide. But it also goes way beyond Garfinkel’s ethnomethodological project, to which the idea that the constructions produced by individuals’ sense-making activities can take on independent external and objective existence is foreign. Nevertheless, Zucker, in her empirical research, was more interested in institutionalization as a process than as an outcome, and thus in practice could remain quite close to an ethnomethodological program. The problem of dealing with institutions as a product as well as a process was not directly confronted. In a recent paper, Barley and Tolbert, while recognizing the work of Zucker & Co. as a major stream of institutional work, nevertheless suggest that it has not “directly investigated the processes by which structures emerge from, or influence, action” (Barley and Tolbert, 1997: 96). To rectify this, they turn to Giddens’ theory of structuration (Giddens, 1976; 1979; 1984). Below, I first present Giddens’ theory of structuration and discuss some of its problems. I then review Barley and Tolbert’s attempt to develop practical tools for institutional analysis on this basis.

60

The Invisible Revolution

Structuration In the introduction to The Constitution of Society, Giddens (1984) complains about the current state of affairs in social theory, which he describes in terms of a “yawning” divide between subjectivism, represented by hermeneutics on the one hand and objectivism, represented by functionalism and structuralism on the other. This is the problem Giddens aims to solve. Within structuration theory, the dualism between subject and social object is reconceptualized as a duality of structure: If interpretative sociologies are founded, as it were, upon an imperialism of the subject, functionalism and structuralism propose an imperialism of the social object. One of my principal ambitions in the formulation of structuration theory is to put an end to each of these empire-building endeavours. The basic domain of study of the social sciences, according to the theory of structuration, is neither the experience of the individual actor, nor the existence of any form of social totality, but societal practices ordered across time and space. (Giddens, 1984: 2)

Giddens is trying to transcend the question of the primacy of individual or society, agency or structure, by focusing instead on social practices. It is social practice that constitutes individuals as actors, and also embodies social structure. Actors and structures, while not the same thing, are two aspects of the same thing, namely social practices (Craib, 1992: 34). While the ongoing activity of daily life occurs as a flow of intentional action (Giddens, 1984: 8), which includes not only discursively reasoned but also unconsciously motivated activities, Giddens stresses that unintended consequences of action have a vital role to play in that they may systematically feed back to become the unacknowledged conditions for further acts: Thus one of the regular consequences of my speaking or writing English in a correct way is to contribute to the reproduction of the English language as a whole. My speaking English correctly is intentional; the contribution I make to the reproduction of the language is not. (Giddens, 1984: 8)

With Bhaskar (1989:27) one could perhaps label this the “duality of praxis, ” the corollary to the “duality of structure.” In other words, “praxis is both work, that is, conscious production, and (normally unconscious) reproduction of the conditions of production, that is society.”

61

Holm

While Giddens in this way emphasizes the importance of unintended consequences of action, he is at pains to distance himself from the notion that action is in any way determined, which is a consequence of the functionalist treatment of unintended consequences (Giddens, 1984: 12-14). This is fundamental to structuration theory, in which the social world is produced by the active doings of subjects. Giddens thus stresses that agents are always able to act otherwise: To be able to ‘act otherwise’ means being able to intervene in the world, or to refrain from such interventions, with the effect of influencing a specific process or state of affairs. (...) An agent ceases to be such if he or she loses the capability to ‘make a difference’, that is, to exercise some sort of power. (Giddens, 1984: 14)

Action, then, logically involves power. Giddens proposes that power, which characterizes all action, is exercised through the utilization of resources. “Resources are structured properties of social systems, drawn upon and reproduced by knowledgeable agents in the course of interaction” (Giddens, 1984: 16). One can perhaps talk about “a duality of power,” in which the resources drawn upon in attempts “to make a difference” are at the same time reproduced and distributed by way of unintended consequences of these activities, as structural properties. This brings us finally to Giddens’ concept of structure, the core of structuration theory. As already mentioned, he does not grant structure an independent existence, external to actors or action. At this point, Giddens rejects the functionalist notion of structure, as a kind of patterning of social relations, sometimes naively conceived of as the skeleton of an organism. Instead, he borrows from structuralist and post-structuralist thought and suggests that structure can be seen as an intersection of presence and absence, and that the underlying codes can be inferred from surface manifestations (Giddens, 1984:16). It is perhaps easiest to understand what this means by way of a linguistic analogy. Giddens suggests that the relationship between structure and action is like that between language and speech. Thus, while speech (like action and interaction) is always spatially and temporally located, presupposes a subject and is oriented towards another person, language (like structure), has no socio-temporal location, is characterized by the absence of subjects, as well as a subject-object dialectic (Craib, 1992: 41-42). Structures are only present in their instantiation, “in the constituting moments of social systems” (Giddens, 1979: 64):

62

The Invisible Revolution

[This] implies recognising the existence of: (a) knowledge - as memory traces - of ‘how things are to be done’ (said, written), on the part of the social actors; (b) social practices organized through the recursive mobilisation of that knowledge; (c) capabilities that the production of those practices presupposes. (Giddens, 1979: 64)

Commenting on a transcript of a “a strip of interaction” in a courtroom between a public defender, a district attorney and a judge, Giddens (1984:330) notes that such exchanges become meaningful for the participants only by the tacit invocation of institutional features of the system of criminal justice. The mutual knowledge each of them draws upon and assumes to be held by the others, consists, in addition to an awareness of the tactics of proper procedure, of a vast amount of knowledge about what the legal system is, the normative procedures of law, the roles of prisoners, advocates, judges, etc: However, by invoking the institutional order in this way - and there is no other way for participants in interaction to render what they do intelligible and coherent to one another - they thereby contribute to reproducing it. Moreover, it is essential to see that in reproducing it they also reproduce its ‘facticity’ as a source of structural constraint (upon themselves and upon others). They treat the system of justice as a ‘real’ order of relationships within which their own interaction is situated and which it expresses. And it is a ‘real’ (i.e., structurally stable) order of relationships precisely because they, and others like them in connected and similar contexts, accept it as such - not necessarily in their discursive consciousness but in the practical consciousness incorporated in what they do. (Giddens, 1984: 331)

This brings us back to the notion of the “duality of structure,” implying that the structural properties of social systems are both the medium and the outcome of the practices they organize. Structures and actors do not constitute two independently given sets of phenomena, but represent a duality. Giddens’ theory of structuration addresses the notion of institution directly in this way: The most deeply embedded structural properties, implicated in the reproduction of societal totalities, I call structural principles. Those practices which have the greatest time-space extension within such totalities can be referred to as institutions. (Giddens, 1984: 17)

While institutions are practices, their extension in time and space suggests that they are closely linked to the notion of structure. Thus, Giddens talks about “institutional analysis” as having to do with the study of structures and structural properties (Craib, 1992: 52). Structural 63

Holm

properties are categories providing the basis for classifying different forms of institutions. More specifically, Giddens has proposed that there are three different but interconnected institutional dimensions: signification/communication, domination/power, and legitimation/sanction (Giddens 1984: 29). Within each of the three dimensions, the “knowledgeable capacities of agents” (displayed in interaction) are related to structural features through something Giddens calls “modalities.” It is through these modalities – interpretive schemes in the case of signification/domination, facilities in the case of domination/power and norms in the case of legitimation/sanction – that the “duality of structure” expresses itself. In other words, the structures of signification correspond, in the institutional sphere, to symbolic orders and modes of discourse; the structures of domination correspond to political and economic institutions; and the structures of legitimation correspond to legal institutions (Giddens, 1984: 31). It is on the basis of these suggestions that Barley and Tolbert (1997) go on to build their version of organizational institutionalism. I shall return to this attempt in a moment. Before that, however, we need to go back to the starting point and ask to what extent Giddens in this theory actually achieves his goal, namely transcending the schism between voluntarism and determinism.

The problem with structuration There are some problems connected with structuration theory as a foundation for institutionalist analysis. While Giddens, like the institutionalists, is preoccupied with the “problem of order,“ or in Giddens’ own terms, the “binding” of space-time in social systems (Giddens, 1984: 17; Craib, 1992:155-65), he is not willing to grant those features of social practices that secure such ordering of practices – structures or structural principles – the ontological status of real, independently existing entities. Instead, Giddens (1979:64) talks about structures as having a “virtual existence” outside time and space, and that they only materialize when they are instantiated in social practices. Let us have a closer look at this argument. In his treatment of structural constraint (Giddens, 1984: 174-180), the main point is that structures are not only constraining but also enabling. Here, Giddens argues against all “objectivist“ notions of constraints, which he takes to imply a notion of “causation more or less

64

The Invisible Revolution

equivalent to the operation of impersonal causal forces in nature.” (Giddens, 1984:174). According to Giddens: Structural constraints do not operate independently of the motives and reasons that agents have for what they do. They cannot be compared with the effect of, say, an earthquake which destroys a town and its inhabitants without their in any way being able to do anything about it. The only moving objects in human social relations are individual agents, who employ resources to make things happen, intentionally or otherwise. The structural properties of social systems do not act, or ‘act on’, anyone like forces of nature to ‘compel’ him or her to behave in any particular way. (Giddens, 1984: 181)

It would seem that Giddens acknowledges that some types of material forces – earthquakes and the like – do indeed act independently of human agency. It would be easy to multiply the examples here, for instance, hurricanes, floods, the effects of El Niño, and the like. And, it would seem that in the case of less acute examples, like global warming and the fluctuations of commercially exploited fish stocks, while people do indeed try to do something about them, they contain elements of “brute material forces“ (cf. Cohen, 1989: 217) in the same way as earthquakes do. But this line of reasoning would be contrary to Giddens’ intent here, which is to marginalize those types of phenomena in social theory. Giddens does not admit physical objects and material forces into the social world, at least not as constituting entities. This role is strictly reserved for knowledgeable human beings. Here it is worth noting that Giddens’ usage of the concept of “praxis” departs crucially from the Marxist tradition, in which it is tied to labor, or work in the physical world (Craib, 1992:166). While Giddens explicitly distances himself from the idea that practices are discursive in nature, he never leaves the interpretive framework. Thus, practices are grounded in practical consciousness, which, although usually not given discursive expression, nevertheless is a type of knowledge (Giddens, 1984: xxiii). What role, then, is the material world allowed to play in structuration theory? One possible point of entry for the material world into structuration theory would seem to be in the form of resources. Resources, together with rules, are in structuration theory the constituents of social structure. While rules are knowledge of how to engage in social practices (Giddens, 1979: 67) or “techniques or generalizable procedures applied in the enactment/reproduction of social practices” (Giddens, 1984: 21), resources are the “media through which power is exercised as a 65

Holm

routine element of the instantiation of conduct in social reproduction” (1984: 16). Here, Giddens distinguishes between allocative resources, which are “forms of transformative capacity” generating command over objects, goods or material phenomena, and authoritative resources, which are types of transformative capacity generating command over persons or actors (Giddens, 1984: 33). A reasonable interpretation of this would be that allocative resources are material objects, while rules make these objects socially relevant (through processes of signification and legitimation (Giddens, 1984: 15). While Giddens acknowledges that resources may have a material or “real” existence, he nevertheless rejects the notion that this is an important aspect of what resources are:

Some forms of allocative resources (such as raw materials, land, etc.) might seem to have a ‘real existence’ in a way which I have claimed that structural properties as a whole do not. In the sense of having a timespace ‘presence’, in a certain way such is obviously the case. But their ‘materiality’ does not affect the fact that such phenomena become resources, in the manner in which I apply the term here, only when incorporated within processes of structuration. The transformational character of resources is logically equivalent to, as well as inherently bound up with the instantiation of, that of codes and normative sanctions. (Giddens, 1984: 33)

In other words, resources too, like structures and power itself, only exist as they are instantiated in social practices. This means that time-space embodiment is not an essential aspect of what (allocative) resources are, and that “material existents” (Giddens, 1979: 104) do not have an independent role in power relations (cf. Barbalet, 1987; Layder, 1987). As this should demonstrate, Giddens exspends considerable effort to deny that structures can take on a “real” existence, prior to and independent of their instantiation in practices, and to deny that material existents can have a role as co-constituents of the social world. An important reason for these two closely linked standpoints seems to be the presupposition that all such “objectivist” notions would imply a commitment to the “regularity principle,” in other words, the notion that general laws, akin to the causal laws in nature, operate in society. When Giddens protests against “objectivist” social science, it is because it – inappropriately – adopts the natural sciences as an ideal, and hence treats “the order of society” as if it were of the same kind as “the order of things.”

66

The Invisible Revolution

Giddens protests against this position because of the “double hermeneutic” – there is a “mutual interpretive interplay between the social sciences and those whose activities compose its subject matter” (Giddens, 1984: xxxii). No such phenomenon exists in nature, according to Giddens. This position is connected with his notion of power and the dictum that agents always can “act otherwise,” a view he formulates as a contradiction to the idea of social constraints operating like “forces of nature” (Giddens, 1984: 15). In structuration theory, both agency and structure are instantiated in social practices. But this makes it difficult to distinguish between the two, particularly since Giddens, in connection with his discussion of power, insists that agents are never determined by structures: they always have the capacity to “act otherwise.” The distinction between agency and structure collapses, since structures, to the extent that they are constraining, are so only because agents allow them to be (Craib, 1992: 150-1). The problem with Giddens’ theory of structuration, then, is that, while he is setting out to overcome the dualism between social subject and social object, or in his own words, turning it into a duality, he comes down on the side of the subjects, thus falling down into the lower, individualist half of our basic map (Figure 1, p. 33).

Structuration as a basis for institutional analysis In Gidden’s model of structuration, the realm of action is connected to the institutional realm through something called modalities. In Barley and Tolbert’s (1997) attempt at developing an institutional analysis on the basis of structuration, they suggest two modifications of Giddens’ model. The first is the substitution of the notion of “modalities” with that of “scripts,” defined as “observable, recurrent activities and patterns of interaction characteristic of a particular setting” (Barley and Tolbert, 1997: 98). Barley and Tolbert go on to say that “[t]he notion of a script usefully substitutes for Giddens’ more abstract notion of modalities because scripts can be empirically identified, regardless of the type of actor or level of analysis in which a researcher is interested” (ibid: 98). The second transformation is to turn Giddens' model, which Barley and Tolbert argue is basically static, into a more dynamic model capable of linking actions to the maintenance and change of institutions. They propose a sequential or recursive model, in which

67

Holm

there are four stages. Within this model, at stage one, institutional principles are encoded in the scripts used in specific settings. These scripts, and thus the institutionalized structural principles, are in the second stage enacted when they are utilized in practical action. In this way, structural principles translate into patterned activities. During the next two stages, the process works the other way around. In the third stage, actions influence scripts either by way of replication, in which case scripts remain the same, or by revision, in which case scripts change. In the last stage scripts are linked back to structural principles by way of objectification and externalization of the patterned behavior in question. These stages have brought the process back to the starting point, from which the whole sequence of encoding, enacting, replication/revision, and externalization/objectification can start over again. Despite the two modifications Barley and Tolbert undertake in order to make Giddens' theory more practical, the model that results stays quite close to the original. I also think that the major problem of of Giddens’ theory of structuration, i.e. that the structural side of things is granted too little independence from acting and thinking subjects, is not solved in Barley and Tolbert’s version. While they suggest that the scripts are empirically observable, it seems from the examples given that they cannot be observed directly as such, but can only be inferred from action and interaction sequences. One example of such scripts is taken from Barley’s (1986) study of interaction patterns in radiology departments. ‘Direction seeking’, for instance, was a script that embodied the institution of medical dominance. Although the content of specific instances of direction seeking varied widely, each was characterized by the same interactional plot: (1) a technologist inquired about an appropriate course of action; (2) a radiologist provided the technologist with an answer; and (3) the technologist acted accordingly. Because the technologists enacted the script long after they had learned to run the scanner and, hence, generally knew the answers to the questions they asked, the script served primarily to validate the radiologists’ dominance as professional experts. (Barley and Tolbert, 1997: 98)

Here, the only access to the script is through the observation of the actual actions performed. The only empirical intake to the scripts as well as the structural principles they embody is through the realm of action. The same can be seen when Barley and Tolbert go on to flesh out a methodology for studying structuration. This methodology (ibid: 103-111) focuses overwhelmingly on the charting of flows of action and interaction. No allowance is made for identifying scripts and

68

The Invisible Revolution

structural principles on their own, as taking on objective, external form independent of action, say, as encoded in textual and material form, or built into machines or buildings. While this perspective may have something to say about how institutions arise from action, it is fairly restricted in accounting for the way institutions gain independence from the level of action and work back on the actors. Despite the efforts to transcend the schism between voluntarism and determinism, action and institution, this line of work falls into the lower half of our map (Figure 1, p. 33).

Expanded Institutionalism (Powell and DiMaggio) Powell and DiMaggio (1991) have proposed another neo-institutional perspective, labeled an “expanded institutionalism.” As this suggests, Powell and DiMaggio are critical of the restricted dualist perspective of Meyer, Scott & Co. As we have seen, Meyer and Scott too have accepted that the dualism of their early work represents a problem. Nevertheless, the remedies Scott and Meyer have proposed take a different direction than those suggested by Powell and DiMaggio. Indeed, there seems to be considerable tension between these two camps, indicated for instance by Scott and Meyer’s charge that expanded institutionalism is an attempt at reviving “a rational actor endowed with a preexisting set of interests” (Scott and Meyer, 1994a: 4). In this view, Powell and DiMaggio have committed treason and joined forces with the arch-enemy, the rationalists. We shall come back to a discussion of the adequacy of this interpretation in a little while. Before that, however, let us take a closer look at what Powell and DiMaggio have actually proposed.

Institutions shape action – action shapes institutions The proposal of an expanded institutionalism starts out with a critique of the dominant neoinstitutional perspective – that of Meyer & Co. – as being “restricted” or “narrow.” Powell

69

Holm

(1991), for instance, characterizes the interpretation of institutionalist arguments in the organizational literature in the following way: Institutional arguments have become associated (inappropriately, in my opinion) with notions that organizations are not concerned with task performance; instead organizations are viewed as manipulators of appearances, seeking only legitimacy. Others have used institutionalism as a residual explanation, useful for explaining outcomes when more instrumental explanations proved inadequate. The individuals who inhabited institutionalized fields were regarded as unsure of their interests or unable to realize them, thus making them susceptible to mimetic influences. Institutional arguments and perspectives such as rational choice or population ecology were portrayed as oppositional. (Powell, 1991: 189)

While Powell is quick to add that this is a misguided characterization – one that highlights specific findings at the expense of the basic arguments – he nevertheless seems to agree that such a misreading has some basis in institutionalist work. In the same vein, DiMaggio and Powell claim that: Up until now, it is fair to say that the new institutionalism has been most attentive to processes of legitimation and social reproduction. We have emphasized that organizational environments are composed of cultural elements, that is, taken-for-granted beliefs and widely promulgated rules that serve as templates for organizing. (DiMaggio and Powell, 1991a: 27-8)

But institutions, as this restricted institutionalism suggests, are not only constraints on human agency. They are also products of human action. What is lacking from the institutionalist agenda is hence the analysis of agency, power, interest and change (DiMaggio and Powell, 1991a: 28). This picks up on DiMaggio (1988:4), who argued that institutional theory had limited itself to “... precisely those aspects of organization that are unaffected by the particular interests of politically conceived actors.” Powell and DiMaggio recommend that institutionalists untie themselves from such unnecessary restrictions and instead pay more attention to how “... political and institutional forces set the very framework for the establishment of economic action; these processes set the limits of what is possible” (Powell, 1991: 187). Institutionalists, according to Powell: ... need to develop arguments as to why outcomes at some given point in time cannot be understood in terms of the preference of actors existing at that same point in time, but must be explained as the product of previous choices, that were shaped by institutional conventions and capabilities. This requires much

70

The Invisible Revolution

more sophisticated analysis of how institutions, by precluding some options and facilitating others, shape individual identities and public discourse. (Powell, 1991: 189)

In an expanded institutionalism, then, institutions are products of action, while, at the same time, institutions shape individual interests and choices. The shift of perspective proposed by DiMaggio and Powell would have direct implications for the focus and analysis of institutionalist case studies. As already mentioned, the typical neo-institutional piece is a diffusion study, describing how some institutionalized organizational form becomes fashionable and works its way into a given organizational field. The causal arrow is outside ⇒ in; from environment, seen as a reservoir of institutionalized bits and pieces, into organizations, which gain legitimacy if they can display the appropriate set of institutionalized bits and pieces. In institutionalist studies, the relationship between organization and environment is ruled by the logic of appropriateness. When the standard that for the moment is seen as “modern” has been set by the environment, the only available response for organizations that want to survive is to conform (Oliver, 1991; Scott, 1995). In this standard neo-institutional picture, as remarked by Zucker & Co., there is not much room for explaining how institutionalized forms arise in the first place. In the early dualist perspective, one source of establishment of institutionalized forms could have been the realm of the “technical,” so that practices arising from rational exchange processes gradually tear themselves loose and live on in the realm of the institutional (cf. Tolbert and Zuker, 1983). But this is somewhat unsatisfactory, since it would imply extreme deference towards rational mechanisms, and in effect define institutionalism as a theory of yesterday’s organizational solutions. In the later structuralist perspective, a given institutional pattern at one level is typically “explained” by its correspondence to another such pattern at some higher level. But where does this pattern come from, then? Another, even higher, more universal level? In the end, this perspective has no way of explaining where institutions come from. All we get is a rerun of the lesson we have already learned from Weber, that the Western cultural account is preoccupied with the rational. The expanded institutional perspective solves the problem of how institutions arise by looking at them as outcomes of political projects, projects in which actors with purposes and

71

Holm

interests try to have new institutions set up or established ones changed. This does not mean a reintroduction of the rational actor as a basic ontological starting point, as in neo-classical economics. On the contrary, in expanded institutionalism, such actors can be seen as outcomes of institutional processes. Powell (1991: 188), for instance, argues that a “key institutional insight” is that individual preferences, choices, and indeed self-definition, are constituted by institutional structures. Here, the expanded institutionalist perspective dovetails with the Weberian tradition’s broad conceptualization of the relationship between institutions and individual behavior, and in particular its focus on the way institutions affect the identities, self-images, preferences and rationality of actors (Hall and Taylor, 1996; Immergut, 1998). The rational actor in this picture is clearly not a naturalized, universal entity outside history, but a historical and institutional product. Nevertheless, expanded institutionalism assumes that such institutional products can tear themselves loose and gain some independence from the processes in which they are created. They can therefore also serve as a basis from which new institutions can be constructed or old institutions changed or defended. Powell wants institutionalists to demonstrate “how political and institutional forces set the very framework for the establishment of economic action; these processes define the limits of what is possible” (Powell, 1991: 187). In this way, the expanded institutionalism of Powell and DiMaggio reintroduces rationality, power and politics into institutional analysis. Instead of systematically shying away from such issues, perhaps because they were seen as (rightly) belonging to the domain of the rationalists, they can now be confronted head on. Such issues represent a major focus of expanded institutionalism. In making this point, Powell reaches back to the old institutionalism and quotes Stinchcombe’s (1968:107) definition of an institution as “a structure in which powerful people are committed to some value or interest” (Powell, 1991: 191). With regard to the treatment of politics and power expanded institutionalism is also most critical of the established neo-institutionalist tradition. Hence, Powell faults neo-institutionalists for having neglected “the myriad of ways in which professional groups and organizations can shape the institutional environment in which they operate” (Powell, 1991: 201). While expanded institutionalism thus regards institutions as products of actions, this should in no way be seen as a return to a rationalist focus on results of action as somehow contained within the intentions of the actors. Quite the contrary; the emphasize here is on path dependency

72

The Invisible Revolution

and unintended consequences (Hall and Taylor, 1996: 941). Here, Powell refers to the work of economic historians Brian Arthur (1990) and Paul David (1986) on path-dependent processes, or “the surprising manner in which historical small events can become magnified by positive feedback” (Powell, 1991: 193). Borrowing the concept of “punctuated equilibrium” from evolutionary biology (cf. Krasner, 1984), Powell suggests that institutional change, when it occurs, is likely to be episodic, “highlighted by a brief episode of crisis or critical intervention, and followed by longer periods of stability or path-dependent development” (Powell, 1991: 197). My own attempt at contributing to neo-institutional theory was directly inspired by and a response to DiMaggio and Powell’s call for an “expanded institutionalism.” Paper 1, “The Dynamics of Institutionalizaton,” starts with the point that new institutionalists “cannot explain how institutions are created and how they change,” and goes on to argue for an expanded institutionalist agenda, i.e. one that includes all types of behavior, also those driven by interest and power (Paper 1: 398). The paper suggests that institutionalists should take seriously the insight that institutions, while they are frameworks for action and therefore taken-for-granted, are also products of action and thus constructed for some purpose. In order to handle both these aspects of institutions, the paper proposed the concept of “nested systems,“ by “drawing a disctinction between actions guided by the established institutional order, on the one hand, and actions geared toward creating new or changing old institutions, on the other” (p. 399). The parallels to DiMaggio and Powell’s perspective are obvious and explicit. To the extent that the concept of “nested systems” contributes anything new, it must be that the number of levels in the analysis can be extended almost ad infinitum, and that the feedback mechanisms between level can be of different kinds and will often give important unintended and unexpected consequences. Expanded institutionalism does not look for the essence and final source of institutional processes in only one direction, be it upwards to the macro level of universal ideas, like Meyer and Scott, or downwards into the minds of individuals, like Zucker and Co. Instead, expanded institutionalism, being driven to an extreme in the nested systems concept, is more pragmatic and finds institutional processes scattered among several levels of action, levels which are constantly being produced and reproduced in relation to each other.

73

Holm

A return to economic man? The expanded institutionalist program wants to span both of the two dichotomies making up our basic map. On the individualist-holist dimension, it sees actors, their interests and perceptions as institutionally shaped, while at the same time leaving individual actors some room for making choices and changing institutions. On the materialist-idealist dimension, it wants to keep ideational and cultural stuff in sight at the same time as the material world is kept firmly in place, in particular as linked to human affairs through power, interests and rationality. While the expanded institutionalist program thus must be characterized as dualist, this is a different type of dualism than the one we found in the early work of Meyer & Co. There, as we saw above, the main axis was constructed along the diagonal of our map, contrasting the “institutional,” belonging in the holist-ideational upper left-hand corner on the one hand and the “technical,” belonging in the individualist-materialist lower right-hand corner on the other. An important point within this perspective was to maximize the distance between technical and institutional realms, making phenomena belonging to each side radically dissimilar and forever separated. The expanded institutionalist program is dualist in an entirely different way. Here, institutional phenomena are located in the interplay between ideas and interests, and between individuals and collectives. This is reminiscent of the phenomenology of Berger and Luckmann, in which man is a social product at the same time as society is a human production. With such a perspective it becomes possible to keep in focus both moments of institutionalization: how institutions are products of action and how institutions set the framework of action. Instead of two opposed and utterly different realms, both serving as separate causes for empirical phenomena, as in Meyer and Scott’s early dualism, Berger and Luckmann propose a dialectical analysis, in which both sides of the postulated dichotomies can reciprocally cause each other. Nevertheless, there remain some problems connected with the dualism of the expanded institutionalist program. What ontological status does this program assign to actors and institutions, respectively? Are they, as Giddens would have it, different expressions of the same reality, or are they grounded in two basically different types of existence? The same type of question can be asked concerning the materialist – idealist dimensions. What is the relationship

74

The Invisible Revolution

between interests and ideas? What ontological realms do they belong to, and how do they come together? In short, where exactly does expanded institutionalism settle down in our basic map? In answering these questions, we note that the expanded institutionalist program puts a strong emphasis on the reintroduction into institutional analysis of agency, interests, power, rationality, and related issues. This stands out as the most important expansion DiMaggio and Powell have to offer as against the established new institutionalist agenda. It is the absence of such issues that DiMaggio and Powell note when they describe the “restrictedness” of new institutional analysis. And it is the re-introduction of these issues into the analysis that is going to allow new institutionalists more leverage on such phenomena as institutional creation, reproduction and change. One interpretation of this is that the expanded institutionalist agenda goes too far in granting individual agency and interests existence independent of institutions. Friedland and Alford have thus criticized the expanded institutionalist program for harboring an “institution-free conception of interest and power” (Friedland and Alford, 1991:244), and Scott and Meyer have similarly blamed DiMaggio and Powell for making an attempt at reviving a “rational actor endowed with a preexisting set of interests to be defended and advanced by creating new cultural rules...” (Scott and Meyer, 1994a: 4). In this view, as mentioned above, Powell and DiMaggio’s expanded institutionalism jumps right out of the true institutional home square (the holist-idealist upper left-hand corner), and lands in the lower right-hand square of neo-classical economists and the rational choice crowd. I don’t think that this is a reasonable interpretation, since it would have to ignore so much of what Powell and DiMaggio have explicitly said about expanded institutionalism, for instance their acknowledgement of institutions as independent phenomena, not wholly reducible to aggregates of individual behavior; the point that agency and interests are institutionally shaped; the emphasis on actors in pursuit of their interests as they see them within historically and institutionally defined contexts; the admonition that institutionalists pursue historically detailed analyses in which decisions at one point may both set the stage for and create barriers to later decisions; and their explicit rejection of rationalist explanations of institutions, as found, for instance, in rational choice and institutional economics. Against the background of the considerable work done by Powell and DiMaggio to distance the expanded institutionalist program from rationalist approaches, it seems far-fetched to postulate an identity between them.

75

Holm

This said, there are indeed aspects of the expanded institutionalist program that would seem to suggests that it has more in common with theoretical perspectives belonging to the individualist Weberian lower half of our map than with the upper Durkheimian half. While Powell and DiMaggio did not seek support in Weber for expanded institutionalism, it has much in common with various institutional traditions that are explicitly neo-Weberian, namely historical institutionalism within political science (Thelen and Steinmo, 1992; Hall and Taylor, 1996; Immergut, 1998), and Brinton and Nee’s (1998) sociological version of neo-institutionalism. We can note that from both these perspectives, Powell and DiMaggio’s version of organizational institutionalism is cited with approval, while Meyer, Scott and Co.’s version is written off as Durkheimian structuralism that is of little value (Hall and Taylor, 1996; Nee, 1998). We should note, however, that even if we accepted that expanded institutionalism is neoWeberian, and thus compatible with some form of methodological individualism, this is not the same as locating it in the extreme rationalist corner. As Guneriussen (1996: Chapter 12) has argued, there are several distinct versions of methodological individualism. The operative variable here is the conceptualization of the “individual,” going from a biologically given entity at one extreme to a socially and institutionally defined entity at the other. If we plot this into our map, the first belongs to the (lower) materialist right-hand side while the latter belongs to the (lower) idealist left-hand side. Guneriussen points out that the naturalistic concept of the individual on the right-hand side is rooted in Classicist or Enlightenment thought, while the concept of the individual as socially embedded on the left-hand side is rooted in Romantic thought (Guneriussen, 1996: 301-305; cf. Bloor, 1991: Chapter 4). The Classicist model of the individual, as found for instance in rational choice and neo-classical economics, is the rational actor, naturally pre-given, primarily calculative and instrumental in orientation. The Romantic model of the individual, as found for instance in the historical-hermeneutical tradition, to which Weber belongs, is the expressive, sense-making individual who is much more sensitive to the concrete historical and societal context in which s/he finds herself or himself. If the expanded institutionalist perspective is neo-Weberian, then, it is much closer to the Romantic conceptualization of the individual than to the Classicist conceptualization. But even if we thus avoid the conflation of Weber, methodological individualism and Classicist rationalism as parts of the same package, it still does not seem quite satisfactory to

76

The Invisible Revolution

classify the expanded institutionalist program as neo-Weberian. Powell and DiMaggio’s commitment to a view of institutions as independent variables and as phenomena not reducible to aggregates of individual behavior remains difficult to fit into this interpretation. There is, in short, very little in the expanded institutionalist perspective that suggests that individuals are ontologically privileged or that individuals, interests and rationality are less socially constructed than are institutions.

Expanded institutionalism as an assemblage But if we cannot simply place the expanded institutional program somewhere on the lower Weberian half of our map, where does it properly belong? One possible answer is that it tries to avoid being locked into one particular position. Perhaps its “expandedness” does not primarily refer to the attempt to open up a larger piece of empirical reality for institutional analysis, but to an attempt to include all the major alternative institutional perspectives under one analytical umbrella? Above, we examined two ambitious institutionalist programs that turned out to be partial. The “institution as environment” program of Meyer & Co., settled – after recovering from its deferential dualist start – in the extreme holist/idealist corner. From this point of view it has a lot to say about how macro-institutions can constitute and constrain individual perception and action. But since it allows virtually no room for individuals to act on their own, detached from institutional scripts, it has very little to say about the way human actors can constitute or change institutions. The “institution as organization” program of Zucker, Barley and Tolbert, on the other hand, has its strength in the analysis of the processes about which Meyer & Co. remain silent. This version of institutionalism could contribute important insights as to the way institutions arise from action. However, since it is very reluctant to grant such institutions autonomous existence, detached from individuals’ thinking and acting, this program has little to say us about how institutions work back on the level of action. As this implies, these two partial institutional perspectives, if they could be blended together, would seem to be able to constitute a more complete institutionalist program. One way to interpret the expanded institutional program of Powell and DiMaggio, then, is that it attempts to

77

Holm

do exactly this. Such a program would span both the “institution as environment” perspective of Meyer & Co. and the “organization as institution” program of Zucker & Co. In this way, an expanded institutionalism should be able to capture both moments of the relationship between micro action and macro institutions, allowing at the same time that actions are shaped by institutions and that institutions arise from action. But in addition to these two perspectives, located quite far to the idealist left of our map, Powell and DiMaggio also wanted to include perspectives further towards the materialist right. In this way, the expanded institutionalism would be better equipped to handle classical rationalist issues of power, interests and rational action. In this interpretation, a main point of the expanded institutionalist program is to construct an integrative analytical framework capable of containing the different brands of institutionalist perspectives. This is of course exactly what Powell and DiMaggio’s 1991 edited volume does, since it brings together “classical” and new exemplars of neo-institutional analyses representing a range of different, perhaps even conflicting perspectives. The question that remains is whether Powell and DiMaggio manage to supply their integrative framework with a sound theoretical foundation, or whether it remains a hodge-podge of conflicting theories that can stay together only as long as the basis on which it rests is not examined too closely. Now, Powell and DiMaggio’s edited volume on institutional theory is supplied with an introduction (DiMaggio and Powell, 1991a) which discusses the theoretical foundation of institutional theory. Its main ambition is precisely to supply institutional theory with a sounder theoretical basis. After reviewing a range of alternative propositions, in particular with a view to finding a sound micro-foundation for institutional theory, DiMaggio and Powell (1991a: 25-26) suggest that Bourdieu’s theory of practice (Bourdieu, 1990) is the answer. While this may indeed be the case and has been developed in other work (DiMaggio, 1986; DiMaggio and Mohr, 1985; Mohr, 2000), this is not a theme that is developed in the rest of the 1991 edited volume. Instead of a systematic development of institutional theory on the basis of Bourdieu’s concepts, we get a medley of institutional analyses that depart from very different, often conflicting, theoretical foundations. It is perhaps unfair to Powell and DiMaggio to require them to solve adequately the problems of neo-institutional theory. As they themselves admit (DiMaggio and Powell, 1991a: 22), “a full discussion” of the micro-foundation of institutionalism would have required a full

78

The Invisible Revolution

volume of its own. It is clear that this is not what Powell and DiMaggio had intended for this volume. Instead, the main focus here seems to have been much more modest, simply bringing together different neo-institutional perspectives and seeing if they could bring something to one another. As such, I think it succeeded. Nevertheless, this also means that one should not see the expanded institutional program as a unified theoretical perspective. The expanded institutionalist program is better viewed as an assemblage of different theories than as a true attempt to construct one comprehensive and consistent perspective.

Institutional Theory and the Invisible Revolution There are many different institutionalisms. The group of institutionalist perspectives which have been in focus in this chapter, the neo-institutionalism within organizational sociology, is located in a relatively holist and relatively idealist part of the landscape of social theory. Its holism is apparent in its insistence, against the stronger forms of methodological individualism, that institutions are not reducible to aggregates of individual behavior. Its idealism is apparent in its insistence, against the naturalistic conception of the rational actor and the materialism of Durkheimian-type functionalism, that the reality of institutions in the end results from people’s belief in them. If there is a common starting point among the organizational institutionalist perspectives, it is the conceptualization of institutions as collective ideational patterns made solid by people acting as if they were real. This said, we must add that there are important varieties and tensions within organizational institutionalism. We have looked into three major varieties. In the first, Meyer, Scott and Co., resorting to an idealist version of Durkheimian structuralism, place the ultimate reality of institutions in a macro world of abstract, universal ideas. The problem here is to understand why institutions would change, and why some ideas have stronger impacts in certain situations and periods than in others. In the second version of organizational institutionalism, Zucker, Tolbert & Co. instead seek the essence of institutions in the micro world of individuals and the subjective will to make a difference. While this solves the problem of institutional change, it immediately opens up the associated problem of institutional stability. Why would there ever be stability and order in society if institutions hinge on people’s beliefs in them? In the third version of

79

Holm

organizational institutionalism, DiMaggio and Powell have proposed to solve the problems of the other two perspectives by combining them, and throw in a bit of rational choice for good measure. The expanded institutionalism still starts out in the holist/ideational square, insisting that institutions are collective social constructions, thus depending on people’s beliefs in them. However, by allowing individual (and collective) actors some measure of freedom from the processes in which they are created, expanded institutionalism solves both the problem of stability and change. Institutional change is possible because institutions come about as the result of political projects in which intentional, rational actors are involved. Institutional stability is possible because institutions, once they are set up, materialize and gain some independence from the processes in which they were created. While this perspective solves most of the explanatory impotence of the former two, it may be objected that it does so by skipping over, rather than solving, the fundamental ontological and epistemological problems of what institutions are and how we can study them. In this context, of course, we are not so much interested in the identities and problems of institutional theories in themselves, but whether and, if so, how neo-institutionalism can help us out in the context of the invisible revolution. As we saw in Chapter 1, the introduction of fisheries resource management brought about many important changes in the institutional structure of Norwegian fisheries. I wanted to see this as a revolution, in which one coalition managed to impose their interests, values, beliefs, and perspectives as dominant within the sector, thus marginalizing the interests, values, beliefs and perspectives of formerly dominant groups. The problem with this interpretation was that virtually no-one else seemed to acknowledge it. If there had been a revolution, it was invisible to almost everyone except myself. A possible solution to this problem, as outlined in Chapter 1, was that the revolution had been rendered invisible because of the success of its naturalization. The institutional system of fisheries resource management is portrayed as a rational response to nature, as certified by its reliance on science and scientific knowledge. The resource management system thus does not appear as a political system, at least not in the same way as the MSO system which it replaced. If we accept this account, the revolution disappears from view. Instead of being seen as revolutionary, the changes brought about by the introduction of fisheries resource management then can be understood as the substitution of rationality for irrationality, as progress from a system that lived on an illusion

80

The Invisible Revolution

(“the ocean and its resources are unlimited”) to a system that rests reassuringly on facts and rationality (“fisheries resources are actually vulnerable and limited; they must be managed”). Now, there can hardly be any doubt that neo-institutional theory in principle will side with the invisible revolution. From their idealist starting point, neo-institutionalists will have no problem with the notion that fisheries resource management should be seen as institutional, and that as such it is not fundamentally different from the MSO system it replaced.3 Accepting that both systems are institutional, the neo-institutionalist will focus on the difference in the way the two are accounted for. Hence, the MSO system drew on and instantiated dominant cultural perceptions of the late 19th and early 20th centuries, featuring collectivist ideologies connected to the labor movement, the cooperative movement and corporatism. During the 1980s and 1990s, this system collapsed and was replaced by institutions that relied on very different notions. Collectivism and corporatism could no longer command much legitimacy and support. Instead, much more individualistic and rationalistic ideas had gained ground, as exemplified by the “new right” movement, market liberalization and globalization. In the Norwegian fisheries, such ideas gained access and were institutionalized, in particular through the implementation of fisheries resource management. An important part of this is a shift in the taken-for-granted perception of the fundamental problem situation of the sector. Under the MSO system, the basic situation of the fisheries had typically been defined as one conforming to the logic of class struggle, in which the fisher was seen as exploited and in need of state protection from oppressive capitalists. Under the new regime, the fish was defined as the weak, vulnerable party in need of state protection from rapacious fishers. With this shift, the explicitly political metaphors of the MSO system were replaced with rational, scientific metaphors of management and control. Through the institution of fisheries management, the dominant cultural configuration of modernity, rationalized science, finally entered the domain of the fisheries (Papers 1, 2 and 3). In this interpretation, the invisible revolution is an instantiation of the great drama of modernization. Within the MSO system, the fisheries had gained temporary shelter from the modern myths of rationalization because of the establishment of a strong integrated institutional

3

This may not be the case for the early dualist institutionalism of Meyer and Scott, in which there is a possibility that the MSO system will be classified as ‘institutional,’ while the resource management system will be classified as ‘technicalk.’

81

Holm

system built on collectivist, anti-capitalist foundations during the early 20th century. The major analytical problem then becomes explaining the mechanism and timing of the collapse of the MSO system. Here, the different neo-institutionalist perspectives will tend towards different interpretations. From the perspective of an idealist structuralism of Meyer and Scott, one would want to explain the shift in the dominant institutional pattern of Norwegian fisheries as a reflection of a congruent shift at some higher level. The obvious candidate is the establishment and institutionalization of the new international/global oceans regime during the 1970s and 1980s. With this, a new set of ideas and institutionalized forms had been authoritatively established at the international/global level, and could gradually work their way into the Norwegian fishery sector and ultimately undermine the established institutional order. In contrast, the more process-oriented perspective of Zucker and Co. would focus on bottom-up rather than top-down mechanisms. In this view, the shift of dominant ideology on the global or international level will not be of much importance in itself. The crucial thing is whether people in concrete organizational settings choose to act according to established institutional patterns, or shift towards new ones. If the new oceans regime works its way into Norwegian fisheries, then, this cannot be explained only with reference to the characteristics of the ideas on which it rests. If new institutionalized patterns are going to take hold, this must be because the legitimacy and taken-for-grantedness of the established ideas and institutions have somehow been eroded and are out of touch with existing practices. From this perspective, one would be very interested in tensions between practical experience and institutional order, for instance the way the actual depletion of fish stocks contradicts the institutionalized notion of unlimited fishery resources or the way the very success of the collectivist system of market regulations has helped create economic groups who no longer see this system as in their own interests. When the established institutions are replaced by new ones, this is because they no longer command support from below, and not because they impose themselves in some top-down process. In the expanded institutional perspective, one would be able to include both top-down explanations à la Meyer and Co. and bottom-up explanation à la Zucker & Co. Environmental shifts, such as the new oceans regime, globalization and European integration are clearly relevant; but so are internal congestion problems and an increasing level of tension between practices and legal and organizational structures. In the nested systems perspective, the shift between the MSO system and the resource management system becomes a paradigmatic shift. Each of these

82

The Invisible Revolution

systems stabilizes itself through positive feedback. A shift from one to the other can occur when something – a crisis or some external shock – allows one stable self-reproducing configuration take over from the other. During the transition phase, the system will be open to and incorporate the cultural patterns which at that moment are floating around in the environment. But when a new equilibrium is reached, the sector again closes itself off in relation to the environment and becomes much less sensitive to environmental shifts (Papers 1, 2, 3). All three neo-institutional perspectives thus allow for the invisible revolution. Nevertheless, “revolution” sounds a bit too strong in the structuralist and the subjectivist versions. From both these perspectives, one would expect that the shift towards new institutional patterns would be more continuous and devoid of conflict than the concept of a revolution suggests. In the structuralist version, this is because it is difficult to understand where the resources to defend yesterday’s institutions would come from when the overarching ideational patterns from which they draw their legitimacy have become outdated. In the subjectivist version, it is difficult to understand why old institutions do not immediately drop out of sight when new patterns of practices have been established. In contrast, the expanded institutionalist perspective offers a view entirely compatible with the invisible revolution. Here, the transition from the MSO system to the resource management system is portrayed as a shift between two radically different self-reproducible systems. When such a system is in place, it tends to be stable and shielded from the environment. When a shift occurs, it tends to be dramatic and total, like a revolution (cf. Powell, 1991). Neo-institutional theory, in particular in the expanded version, is able to produce the invisible revolution. Nevertheless, there are important silences in the way neo-institutionalism speaks to this issue. Consider again the dominant accounts by which our two institutional orders – the MSO and resource management systems – are legitimized. The MSO system comes with a story of weak, vulnerable fishers in need of protection by the state from exploitative capitalists. The context featured in this story is one in which export markets are limited, and fish resources are abundant. The resource management story, in contrast, is one that turns most of these assumptions on their head. Here, the fish is the vulnerable party in need of protection. The fisher has been turned into the greedy villain who needs to be restrained. As we saw in the introductory chapter, it is difficult to see the switch from the MSO system to the resource management system

83

Holm

as a revolution if the first story is perceived as fictional while the second is seen as factual. If the conception of fish as a limited and vulnerable resource, as portrayed by fishery science, is true, and hence the image of fish resources as indestructible is false, it becomes impossible to portray the shift between the two systems as a revolution. It is only if both these accounts can be seen as institutional artefacts, as legitimizing accounts created and maintained within an institutional configuration, that it is possible to see the shift as a revolution. If such a view cannot be maintained, we are stuck with the rational basis of resource management, and hence must accept it as an improvement over the MSO system. The crucial test for neo-institutional theory, then, is whether it can extend an institutional perspective to include science and scientific knowledge. Is it able to see not only the legitimizing accounts of the MSO system as wholly institutional, but also to treat the scientific legitimation of the resource management system in the same way? Neo-institutional theory does not have a well-developed answer to such a question. On the contrary, institutionalists have surprisingly little to offer when it comes to the analysis of science and scientific knowledge as institutions. There are to my knowledge no examples of such analyses, and the theoretical discussion of these issues is absent. Institutionalists have been very intent on leaving behind the restrictions imposed on them by the dualism of the institutional versus the technical. This done, institutionalists could take on important stuff, like market phenomena, competition, and other issues of consequence. But even the most ambitious and able of institutionalists, it seems, have shied away from science. When science and scientific knowledge are featured within this literature, it is not as subjects for detailed empirical analyses, as institutions worth studying in their own right. Instead, the remarks on science are restricted to vague references to its role within the general culture of rationalization, as symbolizing the essence of the Western-style cultural account (Thomas et al, 1987; Scott and Meyer, 1994a). When one stays on such a general level, of course, science remains a cultural perspective, and one need never confront the validity of its claims.

84

The Invisible Revolution

So we leave institutionalism with some disappointment.4 While it could take us quite far, it failed the litmus test of science. It could produce the invisible revolution, but only because it systematically avoided confronting the most important and most difficult task at hand. If we had to rely on institutional theory in order to make our invisible revolution invisible again, it would be very hard to succeed. This would require us to reconstruct institutional theory so as to be able to take on science and scientific knowledge. Luckily, we do not have to take on such a formidable job. This is because there is another tradition within sociology in which this has already been accomplished. This is the Sociology of Scientific Knowledge, to which we now turn.

4

We note, of course, the ‘positivistic’ feel of much institutionalist writing. Institutionalists – in particular in the dominant US branch – don’t carry on an on, like other anti-rationalists, about qualitative methods, interpretations, local truths, and so on. They are doing science in an old-fashioned way, with a strongly positivistic feel, forwarding and testing hypotheses, collecting data, doing the numbers, trying as best they can to contribute to the great work of pushing the knowledge front forward, one step at a time. This would suggest that science, natural as well as social, is simply excluded from consideration. Even if science in principle would fall into the category of the institutional, since this is all there is, nothing follows from this in practice. Knowledge is sacred; removed from study. It counts as a resource, one that can be relied on to do the explaining, and can never be made a research topic, that which is to be explained. 85

Holm

86

The Invisible Revolution

3. THE SOCIOLOGY OF SCIENTIFIC KNOWLEDGE

Introduction Modern fisheries management builds on science. In order to understand fisheries management, we need to come to grips with science. When I first became interested in fisheries management, it was because it had started to invade the established political and institutional formation of Norwegian fisheries. My perspective was institutional, and I wanted to see fisheries management and fishery science as social institutions. While it is fairly easy to analyze fisheries management as a social institution, however, extending such a perspective to science and scientific knowledge represents a much tougher challenge. Yet this was exactly what I wanted to do, and I did so without much consideration of the risks involved. Consider, for instance, the opening paragraph of Paper 5, “Creating Alternative Natures”: Ostrom’s first requirement for long enduring institutions managing common pool resources (CPRs) is to have clearly defined boundaries: ‘Individuals or households who have rights to withdraw resource units from the CPR must be clearly defined, as must be boundaries of the CPR itself’ (Ostrom, 1990: 90). We do not disagree. If appropriation rights are ill-defined, or resource systems are fuzzy around the edges, management becomes difficult. While clearly defined boundaries are typical of successful management institutions, Ostrom’s ‘design rule’ will not necessarily be of much help for prospective institution builders, however. Unambiguous boundaries are consequences of long enduring CPR institutions as much as preconditions for their successful establishment. Nature is complex and open to alternative interpretations. When a single interpretation comes to dominate and a resource system as a result is managed as if it is clearly delimited, this is an institutional artefact, not a fact of nature. (Paper 5: 79)

As suggested here, and insisted upon in the paper’s title, there is not one single and easily knowable nature out there to which management institutions simply must adapt. Instead, “[n]ature is complex and open to different interpretations.” How a given section of nature comes to be defined is not something external to management institutions, but on the contrary something that is produced by the institutions themselves. The paper thus concludes: As demonstrated by this study, the identity of the cod and the boundaries of the resource system should not be taken as pre-defined and external to social institutions. Instead, the nature of the resources is

87

Holm

defined endogenously and is part of the management institutions themselves. The reason why the present Norwegian fisheries management [regime] survives, is not that it relies on a ‘true’ model of nature, but that it commands sufficient organisational and political power to uphold its definition of nature as the authoritative one, and push alternative definitions into oblivion. (Paper 5: 89)

True knowledge of nature is not available. Instead, nature is open to different interpretations. Which interpretation will be dominant is a question of power. Nature is a social construction. With this paper, we have arrived in that strange place where science is no longer allowed to report on reality, but instead takes part in its construction. Hence, neither of the two competing scientifically based interpretations of the structure of cod stocks, the confrontation of which is the main point of the paper, is credited with any greater access to the truth than the other. “Creating Alternative Natures” is not a modest paper. It does not merely suggest that social forces influence how nature is perceived. It claims to have demonstrated that nature is defined institutionally. If this rings true, we have come a great step closer to our goal of making the resource management revolution visible again. When science constructs nature instead of reporting on it, we no longer need to take seriously the claim that resource management is simply an answer to the call of nature. The resource management system then becomes as social as the MSO system. When the two clash, it is all about power and politics, conflicting values and world views. In short, a revolution. Where did the resources on which this paper relied come from? While the paper is not particularly explicit with regard to its theoretical basis, it bears all the signs of social constructivism (Eythórsson, 1999). The specific form of constructivist analysis it performs locates it in the neighborhood of the Sociology of Scientific Knowledge (SSK), a social constructivist perspective within the sociology of science that emerged in the UK during the 1970s. This chapter reviews and criticizes this perspective. Is SSK strong enough to take on science and scientific knowledge? Can it really carry out the work with which it is entrusted in Paper 5, transforming nature from objective reality into artefacts produced within social institutions? In other words, is SSK what we need to make the resource management revolution visible again?

88

The Invisible Revolution

The Orthodox Notion of Science We cannot present SSK without saying a few words about the position against which this perspective contrasts itself and which it wants to defeat. In much the same way as the institutionalists’ battle against rationalism, SSK goes to war and defines itself in contrast to another perspective. This is the orthodox view of science, found in a variety of rationalist and realist forms, holding that science at its best can produce true knowledge about reality. A key idea behind the orthodox notion of science is that of the autonomy of knowledge. This implies that objectively true knowledge, as established by science at its best, is not caused by anything and does not require any explanation. Truth carries its own cause. It is only false beliefs that need to be explained: We were led astray by faulty perceptions; we were deceived by lapses of reasoning; our judgements were clouded by wishful thinking or vested interests, etc. (Bloor, 1991: 8). Now, this position gives a quite restricted role to the sociology of science, as can be seen, for instance, from Lakatos’ “History of Science and its Rational Reconstruction” (Lakatos, 1978). In this paper, Lakatos sketches a methodology for the rational reconstruction of the growth of objective knowledge that also gives a place, although as we shall see a very small one, to the sociology of science, or as he labels it, the “external history” of science. The first step in the procedure is to pick, from the repertoir of modern philosophy of science, a methodology or “logic of discovery,” according to which established scientific theories can be appraised. Which methodology is chosen is not arbitrary. In fact, an important point of the paper is to argue that Lakatos’ own alternative, labeled the “methodology of scientific research programmes,” is superior to the three main alternatives. What characterizes the four rival methodologies, and what criteria they establish for judging scientific theories and scientific progress need not concern us here. What is important is that they set up normative principles for what Lakatos calls “rational reconstruction” of the “internal history” of science. What the historian of science, according to Lakatos, wants to do is not to reconstruct the development of scientific theories and practices as they have actually occurred. Rather, it is the growth of objective, disembodied knowledge that is at the center of attention: … whether an experiment is crucial or not, whether a hypothesis is highly probable in the light of the available evidence or not, whether a problemshift is progressive or not, is not dependent in the slightest

89

Holm

on the scientists’ beliefs, personalities or authority. The subjective factors are of no interest for any internal history. (Lakatos, 1978: 118)

In developing this point, Lakatos draws on the trichotomy between the ‘first’ world of matter, the ‘second’ world of feelings, beliefs, and consciousness, and the ‘third’ world of objective knowledge. Thus, the internal history of science will concern itself with the third world, that of objective knowledge. The historical facts concerning what scientists were thinking and feeling during the process of discovery belongs to the second world “which is only a caricature if its counterpart in the third world.” (Lakatos, 1978: 119). The rational reconstruction of science is thus a process of purification, of disposing of all those irrational and irrelevant factors that distract us from the true discoveries of science: One way to indicate discrepancies between history and its rational reconstruction is to relate the internal history in the text, and indicate in the footnotes how actual history ‘misbehaved’ in the light of its rational reconstruction. (Lakatos, 1978: 120)

While the internal history of science is primary and should be in the center of attention, Lakatos accepts that it must be supplemented by “external history,” which takes care of the irrational residue. The rational aspect of scientific growth is fully accounted for by internal history. Nevertheless, since even the greatest of scientists make false moves and fail in their judgement, any rational reconstruction of science will be “submerged in an ocean of anomalies” (Lakatos, 1978: 134), which is the subject area of external history, and, presumably, the sociology of science. As Bloor (1991:10) points out, the best philosophy of science, according to Lakatos, is one that minimizes the role of external history or sociology of science; progress in the philosophy of science is measured by the amount of actual history that can be exhibited as rational. However, Lakatos is only too happy to hand over to historians or sociologists all instances of science gone bad. He thus suggests that the Stalinist intervention in science, as exemplified by the Lysenko affair, must be left to external history (Lakatos, 1978: 114). Likewise, if the astronomers were suddenly gripped by a feeling of crisis and converted to astrology, this would be a “horrifying problem” that could only be accounted for by some externalist explanation (Lakatos, 1978: 135136).

90

The Invisible Revolution

One might think, perhaps, that such a position, which reserves the core of scientific activity and results for the philosophy of science and only leaves anomalies and irrationality to the sociology of science, would not be taken up by sociology itself. Nevertheless, this idea of science is so well entrenched in the philosophy of science and in the self-understanding of natural science, that it has also made the leap into contemporary sociology. More specifically, its influence can be seen in the Merton School, which remains one of the main perspectives in the social studies of science program today (McMullin, 1992: 17). While this perspective, as we shall see, certainly grants a larger residue to sociology than does Lakatos’, it remains a sociology of the organization of science, of its form, rather than of its content. In his treatment of “Science and the Social Order”, Merton (1996a: 277) starts with Max Weber’s observation that “the belief in the value of scientific truth is not derived from nature but is a product of definite cultures.” He goes on to add that this belief can easily be transmuted into doubt and disbelief. To Merton, then, the pursuit of science is precarious, and support for science is only forthcoming under specific cultural conditions and if science is properly organized. At the center of Merton’s attention is the preservation of the “purity of science:” One sentiment that is assimilated by scientists from the very outset of their training pertains to the purity of science. Science must not suffer itself to become the handmaiden of theology or economy or state. The function of this sentiment is to preserve the autonomy of science. For if such extrascientific criteria of the value of science as presumable consonance with religious doctrines or economic utility or political appropriateness are adopted, science becomes acceptable only insofar as it meets these criteria. In other words, as the pure science sentiment is eliminated, science becomes subject to the direct control of other institutional agencies and its place in society becomes increasingly uncertain. The persistent repudiation by scientists of the application of utilitarian norms to their work has as its chief function the avoidance of this danger, which is particularly marked at the present time. A tacit recognition of this function may be the source of that possibly apocryphal toast at a dinner for scientists in Cambridge: To pure mathematics, and may it never be of any use to anybody! (Merton 1996a: 282)

In a later paper, Merton goes on to outline “The Ethos of Science” (1996b), i.e. a complex of norms that are held to be binding for scientists. This ethos comprises four sets of institutional imperatives, which are presented under the headings of Universalism, “Communism,” Disinterestedness, and Organized Skepticism. While Merton acknowledges that these norms and values have a methodological and technical rationale for systematic and valid prediction, they are

91

Holm

at the same time moral prescriptions, thought of as right and good, and function to define and protect the autonomy of science in relation to the wider society. Merton did not doubt that scientists, if the ethos of science was observed, would be able to produce objectively valid knowledge. Hence, he was not preoccupied with the methods and substantive findings of science (Merton, 1996b: 267). Instead, the purpose of his program for the sociology of science was to understand under which social conditions science could reach the “fullest measure of development” (Merton, 1996b: 268), and under which conditions it would be threatened. While Merton was deeply concerned with the pressures against the “purity of science,” however, it would not be correct to say that Merton’s sociology of science, in line with Lakatos’ perspective, was a sociology of scientific failure. His research program, involving the empirical study of science as a social institution, led him to focus on organizational factors that would affect the production, selection, and distribution of scientific knowledge. Thus, Merton himself has done important work on priority disputes, on the reward system in science, on disproportionate distribution of acknowledgment in cases of collaboration or independent multiple discoveries, and on the effect of aging within the scientific community. However, and much in line with Lakatos’ perspective, the fundamental faith is that, at least as long as the “ethos of science” is observed, the content of science will not be much influenced by the organization of scientists (cf. Sztompka, 1996). Merton’s perspective, then, presupposes a fundamental divide between the content of science, which remains out of reach for sociology, and the context of scientific discovery, which is within the reach of sociology. In the words of another early contributor to the sociology of science, Joseph Ben-David, “the possibility for either an interactional or institutional sociology of the conceptual and theoretical contents of science are extremely limited” (BenDavid, 1971: 13-14). The characterization above is, of course, not meant as a comprehensive review of Mertonian sociology of science. Incomplete as it is, it nonetheless suffices here. The Mertonian perspective on science, implying a naïve realist position not unlike that which many practicing scientists will hold, has been very important as a counterpoint against which the social constructivist position was developed. Closely reminiscent of the critique of rationalist models as an obligatory starting point for new institutionalist work, as we saw in the previous chapter, the social constructivist position was to a large extent defined in opposition to realism and to

92

The Invisible Revolution

Mertonian sociology of science. In order to come to grips with the constructivist position, then, you need to start with the position it was set up to defeat. Having done this, we can now move to the main issue of this chapter, the Sociology of Scientific Knowledge (SSK).

SSK: A "Strong" Sociology of Science As we have seen, the Mertonian perspective started out with a sharp distinction between the external context of discovery, covered by “external history,” and the internal context of justification, the subject matter of a rationally reconstructed “internal history.” According to such a view, a lot of what actually happens as scientific discoveries are made, is irrelevant. Drawing on examples from the history of statistics, Barnes thus observes that within such a perspective, the practical circumstances surrounding scientific discoveries to a large extent remain irrelevant: So, although tales of Adolphe Quetelet’s travels to Britain in the first part of the 19th century (Hilts, 1978), Francis Galton’s private wealth (Pearson, 1914-30), or the enmity between Karl Pearson and Ronald Fisher (Fisher Box, 1978) are interesting stories, they remain just that, stories that have nothing to do with the nature of mathematical and statistical inventions associated with them. (Barnes, 1998: 204)

When true knowledge has been established, it does not matter much how it came about. Internal history covers all the quality stuff; external history is left with blunders, ineptitude and fraud. In contrast, Barnes suggests, a constructivist approach presupposes that the context of discovery is inseparable from the context of justification. Which knowledge claims come to be accepted as true is a question of the circumstances of their production. Going back to the metaphor applied by Lakatos (1978), referred to above, we might say that the constructivists turn on its head the idea of a rational reconstruction of science, according to which the “third world” of objective knowledge is extracted from the “second world” of actual historical events. In contrast, the constructivist position insists that the “second world” is primary. To the extent that the third world of knowledge exists, it is not one of objective knowledge representing the first world of matter, but a construct emerging out of the second world of scientific practice and culture. In the realist picture, then, reality – the first world of matter – acts as a referee with regard to propositions about it, sorting true knowledge claims from false ones. The accepted claims accumulate in the 93

Holm

third world of knowledge, which thus comes to correspond to the first world. The realist position holds to a correspondence theory of truth. In the social constructivist position, in contrast, the first world of matter is denied an active role. Reality will be indifferent to science’s claims about it. When a scientific knowledge claim is accepted as true, this must be seen as the result of social processes of negotiation and agreement within a community holding the authority to make such decisions. In other words, scientifically certified knowledge is a social institution, the production of which is completely open to sociological analysis. This way of looking at science and scientific knowledge emerged during the early 1970s under the label of the sociology of scientific knowledge, or SSK. Sometimes it is also referred to as the "strong program" within the sociology of knowledge (Bloor, 1991). From the outset, the centers of this perspective were Edinburgh (Barnes, 1974; 1977; 1982; Bloor, 1991; 1983; Shapin, 1979, 1982) and Bath (Collins, 1992). While there certainly are differences in emphasis and focus between scholars engaged in SSK, it is not unreasonable to treat it as one unified perspective (cf. Pickering, 1992a; 1992b). As already suggested, this unity broke during the 1980s, when the constructivist perspective branched out into a number of divergent schools. I will return to this in due course. Let us first try to characterize SSK as it emerged during the 1970s. As suggested above, SSK is sometimes identified with the "strong program" in the sociology of knowledge. This phrase was coined by David Bloor in Edinburgh, who summarized the program as consisting of the four tenets of causality, impartiality, symmetry, and reflexivity: 1.

It [i.e. a sociology of knowledge according to the strong program] would be causal, that is, concerned with the conditions which bring about belief or states of knowledge. Naturally there will be other types of causes apart from social ones which will cooperate in bringing about belief.

2.

It would be impartial with respect to truth and falsity, rationality or irrationality, success or failure. Both sides of these dichotomies will require explanation.

3.

It would be symmetrical in its style of explanation. The same types of cause would explain, say, true and false beliefs.

4.

It would be reflexive. In principle its pattern of explanation would have to be applicable to sociology itself. Like the requirement of symmetry this is a response to the need to seek for general explanations. It is an obvious requirement of principle because otherwise sociology would be a standing refutation of its own theories. (Bloor, 1991: 7)

94

The Invisible Revolution

Let us start here with the second and third points, impartiality and symmetry, both of which have turned out to be of great importance in constructivist sociology of science. As can easily be seen, these two tenets are formulated in opposition to the realist perspectives on science, as formulated by Lakatos and Merton. These would indeed not be impartial, treating false beliefs (the subject matter of “external history”) as radically different from true ones (belonging to “internal history”). Nor would these perspectives be symmetrical, as they would insist that false beliefs would require causal explanations, while true beliefs are true simply because they correspond to reality. In contrast, SSK wants to treat all knowledge claims, whether true or false, on equal terms, as equally open to sociological analysis. This is what makes SSK “strong” in comparison to Mertonian sociology of science, or “external history,” as suggested by Lakatos. It is not only stronger in the sense that it is more inclusive, aiming to explain science even when it is regarded as successful and true. Perhaps even more importantly, it is stronger, or at least needs to be stronger, because it does not defer to science itself, to the internal history of science, when it comes to accounting for the knowledge claims that are accepted as true. In insisting that these accounts are not acceptable and must be replaced by sociological ones, it challenges deeply held convictions in science and society. The tenets of impartiality and symmetry also help explain what the first requirement, that of causality, means. In a later comment on the strong program, Barnes, Bloor and Henry (1996: 120-121) say: ... it is not knowledge which is treated as caused, but the change in knowledge brought about by an action… Goals and interests help to explain the change as the consequences of goal-oriented or interested action. (...) Moreover, every action is to be regarded as caused action: it is not conceivable that an action could have some alternative ‘uncaused’ form, the form which ‘rationality itself’ would give it if scientists were ‘disinterested’, and research in no way goal-oriented. The aim is precisely to account for the rational action of scientists, action which ‘of itself’ is underdetermined: the claim is that such action is to be understood as purposive and goal-oriented, and that goal-orientation causes the action to be of one form rather than another… .

The argument for a causal account of knowledge here is formulated in opposition to the orthodox notion of true knowledge claims as realizations of “rationality itself,” obtainable only if scientists remain “disinterested” and unaffected by goals and interests. In contrast, the strong program

95

Holm

generalizes the type of explanation the likes of Lakatos and Merton had reserved for false beliefs. Both true and false belief must be causally explained. Now, for Bloor, as for the Edinburgh School more generally, such causal analysis draws on interests and goal-orientation as key explanatory resources. Scientific knowledge, true as well as false, is social because it is influenced by goals and interests. The fourth and final point on Bloor’s list is reflexivity. Suffice it to say at this stage that SSK scholars routinely acknowledged that the explanatory principles SSK advances with regard to scientific knowledge also must be valid for SSK. Otherwise, as Bloor puts it, SSK “would be a standing refutation of its own theories.” Having made this acknowledgement, however, reflexivity is usually treated as rather uninteresting (Pickering, 1992a: 19; cf. Ashmore, 1989). I shall return to the discussion of reflexivity below. Before that, however, we must try to get a better sense of the dominant mode of SSK’s explanatory practice. We can start here by noting that SSK, much like the philosophical perspectives on science it so strongly rejects, is primarily preoccupied with science as a body of knowledge. SSK certainly opened up for a perspective of science as practice, as opposed to a logical structure of theoretical propositions. But such ideas do not come to fruition in the sociology of science until the next decade. Instead, SSK in the 1970s still perceived of science in knowledge terms, as a conceptual network in which concepts at various levels of abstraction are linked to one another and to the natural world. Now, we already know that SSK rejects the idea that such links can be explained simply by posing some correspondence between the concepts and an objective natural world. What, then, is SSK’s perspective on this relationship? At this point we have to introduce two very important figures in our story, namely Thomas Kuhn and Ludwig Wittgenstein. As should be well known, the key term in Kuhn's (1970) essay on scientific revolutions is that of a “paradigm.” A paradigm is defined as a “universally recognized scientific achievement that for a time provides model problems and solutions to a community of practitioners” (Kuhn, 1970: viii). It is only when such a paradigm has been established that a scientific discipline will engage in the puzzle-solving activities that characterize mature or “normal” sciences. Such normal sciences can, however, break down in crisis. This will happen with the accumulation of anomalies, that is, problems that resist solutions according to the paradigmatic problem/solution exemplars. A scientific revolution then occurs: the old paradigm is replaced by a

96

The Invisible Revolution

new one; the anomalies are resolved and a new era of normal science can begin (Kuhn, 1970; Golinsky, 1998: Chapter1). It has been pointed out that Kuhn's usage of the term “paradigm” is ambiguous and open to a number of different interpretations (Masterman, 1970). Kuhn himself has distinguished between two senses of the term. The first is “the entire constellation of beliefs, values, techniques and so on shared by members of a given community.” The second is that “concrete puzzlesolutions which, employed as models or examples, can replace explicit rules as a basis for the solution of the remaining puzzles of normal science” (Kuhn, 1970: 175). In Barnes’ (1982) and Bloor’s (1991: 59) readings, both of these meanings were fused together. In the first sense, the concept of the paradigm allowed for a treatment of scientific communities like any other communities. Scientific knowledge could be seen as a culture, constituted by shared beliefs, values and perspectives. This opened science up to sociological analysis. In Barnes' (1985: 89) words: “paradigms, the core of the culture of science, are transmitted and sustained just as is culture generally: scientists accept them and become committed to them as a result of training and socialization, and the commitment is maintained by a developed system of social control.” In the second sense, a paradigm features concrete exemplars that guide further scientific activity. This gave the sociologists a pragmatic alternative to the dominant philosophical view of science as a theoretical structure: It was Kuhn's crucial insight that the fundamental units of scientific knowledge are not theories, nor even theories and associated observations, but solved problems. Problem solutions are the irreducible units of resource in scientific research: they are the models on the basis of which further problems are solved. Some of these problem solutions come to be recognized as holding special promise for future research, and become accepted as authoritative bases for its practice in specific disciplines or specialities [sic] in science. These are exemplary achievements or paradigms. In them, theoretical discourse, practice and instrumentation are linked together and grasped in operation: they are understood in use in a way that they could not be understood by abstract consideration. And this understanding is the understanding of the collective for which the paradigm is authoritative, and thus constitutes both that which inspires research and that which informs its evaluation. To carry out research and to judge its results one must be able to do things with exemplars, not to make deductive inferences from theories. Hence, the difficulty of identifying the ‘real meaning’ of a theory, or even what precisely the theory is, has no adverse consequences for actual practice of research. (Barnes, Bloor and Henry, 1996: 101-2)

97

Holm

In developing this point, we need to go from Kuhn to the late Wittgenstein (1968), or rather, to SSK's interpretation of Wittgenstein (Bloor, 1991; 1983; 1996; 1997; Barnes, 1982; 1985; Collins, 1992; Barnes, Bloor and Henry, 1996). The crux of the argument, known under the label “finitism,” is that classification and meaning are always open-ended and created in a step-by step fashion: “Meaning follows usage, and so cannot exist ahead of it to guide and act as an independent reality against which to measure usage” (Bloor, 1996: 850; cf. Barnes, Bloor, and Henry, 1996). The paradigmatic exemplar here is Wittgenstein’s “language game” with a teacher trying to instruct a pupil how to “go on in the same way” from one set of instances of the series created by the rule (n+2) to the next instance of that rule. We are asked to assume that the pupil masters this rule for numbers less than 1000: Now we get the pupil to continue a series (say + 2) beyond 1,000 – and he writes 1,000, 1,004, 1,008, 1012. We say to him: ‘Look what you’ve done!’ – He doesn’t understand. We say: ‘You were meant to add two: look how you began the series!’ He answers: ‘Yes, isn’t it right? I thought that was how I was meant to do it.’ (Wittgenstein, 1968: §185; quoted from Lynch, 1992: 221)

Since the pupil has not been given examples of the application of the rule for numbers larger than 1000, his proposal is not illogical, but consistent with an interpretation of the rule saying “add 2 up to 1000, 4 up to 2000, 6 up to 3000.” (Lynch, 1992: 221-2). In Collins’ rendering, asking the student to “go on in the same way” from the series “2, 4, 6, 8” may bring innumerable responses apart from the “correct” one “10, 12, 14, 16”: “For example it allows ‘2, 4, 6, 8, 10, 2, 4, 6, 8, 10, 12, 2, 4, 6, 8, 10, 12, 14’ or ‘2, 4, 6, 8, 2, 4, 6, 8, 2, 4, 6, 8’ or ‘8, 6, 4, 2, 2, 4, 6, 8, 8, 6, 4, 2’ or any number of other sequences” (Collins, 1992: 13). An attempt to specify the rule so as to avoid such “misunderstanding” is to no avail. If the rule were amended to “add a 2 and then a 2 and then another 2 and so forth,” the pupil might write: “82, 822, 822, [sic] 8222,” or “28, 282, 2282, 22822” or “8 2, etc” (ibid: 13). Wittgenstein (1968: §201) concludes: “This was our paradox: no course of action could be determined by a rule, because every course of action can be made out to accord with the rule. The answer was: if everything can be made out to accord with the rule, then it can also be made to conflict with it. And so there would be neither accord nor conflict here.”

98

The Invisible Revolution

In SSK, this argument on rule interpretation is extended not only to classification, but also to beliefs, theories and knowledge (Bloor, Barnes, and Henry, 1996; Chapter 3). Knowledge, like classification, is non-determined and open-ended. Yet SSK does not deny that in practice, scientists are perfectly able to deduce “correctly” from given rules or theories: The relationship between a theory and its exemplary applications is like that between a class and its existing instances. How a theory is to be applied is not fixed by rules or definitions, or by deductions therefrom; it is decided by those using the theory, who take its existing applications as precedents and proceed from them on the basis of analogy. A theory is its exemplary applications at a given point of time, and that these are the resources for deciding what other applications will be made, the precedents for further problem-solving, the basis for the further case-to-case development of ‘the theory’. (Barnes, Bloor, and Henry, 1996: 105)

Here, of course, we are back with Kuhn, and the importance of the concrete “exemplars” as the key stabilizing features of a paradigm. So, the finitist account does not totally reject the view that theoretical structures have a guiding function, that there in fact will be some “correct” correspondence between a rule and the practices that accompany it. What it denies is that rules, concepts or theories by themselves can determine practices. Instead, something else is required for people to use open-ended rules and theories as guides to action. In the place of logical necessity and deduction, as in the rationalist account, SSK poses social convention and interests as the mechanism that allows scientists “to go on in the same way,” that is, to interpret rules or theories correctly. Social order replaces logical order. But all this is of course rather abstract and can be taken to mean anything, at least if we take SSK’s version of finitism seriously. Let us therefore leave theoretical discourse and take a closer look at the paradigmatic SSK exemplars, that is, the concrete examples of research on which the SSK “form of life” hinges.

99

Holm

Exemplar 1: Mathematics As a demonstration of the “Strong Program,” David Bloor chose mathematics. From a “finitist” standpoint this of course was an ideal case since mathematics is usually perceived in terms of abstract rules and logical deduction. In the case of mathematical knowledge, then, Wittgenstein's arguments could be applied without any need for transposition. If it can be demonstrated that even mathematics and logical thinking – “the most stubborn of all obstacles to the sociology of knowledge” (Bloor, 1991: 83) – can be understood as socially constituted, this would be a great victory for SSK. Unsurprisingly, this is also how Bloor’s analysis concludes: “The compelling force of mathematics derives from being accepted and used by a group of people. The procedures are not accepted because they are correct or correspond to an ideal; they are correct because they are accepted” (Bloor, 1983: 92; cf. Bloor, 1991). Mathematics is a social convention; it is true because it is held to be true by a group of people. But how does Bloor (1991) argue his way towards this conclusion? The main argument here is divided between three chapters. The first of these (“Chapter Five: A Naturalistic Approach to Mathematics”), focuses on the foundation of mathematics, i.e. how numbers can be conceived of, and how to understand algebra and calculus. The target here is the notion of mathematical knowledge as an eternally true and compelling conceptual structure. Against this, Bloor mobilizes J.S. Mills’ psychological theory of mathematics, which looks at mathematics as a body of practical skills, beliefs and thought processes into which individuals must be initiated. The concept of, say, numbers and simple arithmetic, grows out of direct experience with objects (like fingers or pebbles) that can be grouped and sorted in different ways. While Bloor writes with approval of this perspective, since it, in contrast to the standard, rationalist notion of mathematics, is “naturalistic,” that is, it ties mathematical knowledge to practical skills and experiences rather than some eternal conceptual universe. But he still rejects it, since mathematics in Mills’ version becomes too closely aligned to the properties of physical objects (Bloor, 1991: 99). Instead of having the properties of physical objects determine mathematical knowledge, Bloor wants social mechanisms as a crucial component: “The central idea that the behaviour of objects provides a model for our thinking still survives. The difference is that now not any such behaviour functions as a model but only certain socially fixated or ritualised patterns” (Bloor, 1991: 100).

100

The Invisible Revolution

In the next chapter (“Chapter Six: Can There Be an Alternative Mathematics?”), Bloor takes issue with the notion, central to the rationalist account, that mathematics constitutes one single body of theorems (Bloor, 1991: 85-86). Aiming to refute this, Bloor tries to demonstrate that there are significant variations in mathematical thought, and that there indeed exist alternative forms of mathematics. So the chapter illustrates that there are variations in broad cognitive styles of mathematics; that there are variations in frameworks of associations, relationships, uses, and analogies, and the metaphysical implications attributed to mathematics; that there are variations in the meaning attached to computation and symbolic manipulations; and that there is variation in rigor and the type of reasoning that is held to prove a conclusion. In the following chapter (“Chapter Seven: Negotiation and Logical Thought in Mathematics”), Bloor demonstrates that formal principles of inference are not based on self-evident logical truth. Instead they are negotiated agreements within the community of mathematicians or logicians. Bloor’s examples here are the negotiations over the meaning of the logical syllogism: The discussion over whether the Azande have a different logic to that accepted by Western logicians, and the negotiation of a proof in mathematics. I will not run through all this material here. Suffice it to give one example, i.e. the great difference in the perception of numbers between Greek mathematics and contemporary mathematics (Chapter 6: 110-18). For the Greeks, a number was a number of units, always thought of in discrete terms. Nothing could exist between the whole numbers. The contemporary notion of numbers, which took hold in the 16th century, rests on an analogy between numbers and a line. So (whole) numbers are ticks on a continuous line, an analogy that also gives meaning to “0” and negative numbers. Why did this new way of thinking take hold? Bloor answers by showing how practical issues and social interests are at stake: The major mathematical practitioners of the time all had preoccupations which were technological or applied (cf. Strong (1966)). Their practical bias led them to use number not merely to count but also to measure. It was probably practical concerns which broke down the boundaries between geometry and arithmetic. Numbers came to perform a new function by indicating the properties of moving, active processes of change. For example number and measurement became central to an intellectual grasp of ballistics, navigation and the use of machinery. (Bloor, 1991: 117-8)

101

Holm

To Bloor, then, alternatives to the form of mathematics accepted today not only are perceivable, but actually exist. Mathematical proof does not rest on eternal, abstract logic, but on a negotiated agreement within the community of mathematicians of what should count as proof (see also Barnes, 1998: 204-6). Mathematics, like all types of knowledge, including scientific knowledge, is a social institution. Social institutions exist because people believe in them; “[t]hey can be thought of as realities created by reference to them” (Bloor, 1996: 842; Barnes, 1983).

Exemplar 2: Replication Bloor argues as if mathematics and logic are the most difficult of cases for SSK. It could be objected, however, that the negotiated nature of mathematics is obvious since it is (or is perceived to be) so radically removed from the empirical world. The truth of mathematics does not rest in confrontation with data, as in the empirical sciences, but in axioms and abstract procedures of proof. In this way, then, mathematics is not the hardest, but the simplest of cases (cf. McMullin,1992: 19). Can SSK-type insights also be argued convincingly in the case of the empirical sciences? Now, the dominant perception of science, whether in realist or rationalist terms, postulates that there is only one real world, and hence that there can be only one truth about that world. Scientific knowledge is universal. This idea is embedded in and arguably confirmed by the celebrated scientific institution of replication. A scientific discovery will not be acknowledged unless it can be replicated by other scientists. This would seem to hinge on the universality of scientific theories. Unless the “laws of nature,” as discovered by science at its best, were valid everywhere and at all times, replication would be patchy. Replicability, and by extension, the fact that science not only works, but is a great success in practice, would seem to be a standing refutation of relativist doctrines like that of SSK. SSK of course does not accept this. Relying on the finitist argument, SSK scholars insist that a specific application of a theory is contingent on the establishment of a shared paradigm (Kuhn), or a specific “form of life” (Wittgenstein). When a discovery can be replicated, this cannot be taken as proof of the universal validity of some theory. Instead it may be replicated

102

The Invisible Revolution

because the beliefs, values, methods, and techniques on which it hinges have come to be shared within the community of scientists. It is not the truth of a discovery that forces agreement with it. On the contrary, it is agreement that leads to the acceptance of some discoveries as true. This can be exemplified by Collins’ (1992) work on the Transversely Exited Atmospheric (TEA) laser. Now, the successful building of the first TEA laser, which uses carbon dioxide as a lasing medium, was first announced in 1970 by a Canadian defense research laboratory. It turned out, as is often the case, that it was not easy for other groups of scientists to replicate this work. Only those scientists who had spent some time in the laboratory where the success had been achieved proved capable of successfully building their own version of the laser. More generally, the reproduction of an experiment requires close interaction between scientists and experimental arrangements. In order to transmit a discovery to a new setting, the entire culture that allows it must also be transmitted, including a certain know-how and specific ways of seeing and interpreting (Collins, 1992: Chapter 3; cf. Callon, 1994: 42-43). Collins refers to this way of seeing scientific learning and transmission of scientific knowledge as the enculturation model. Here, knowledge, including scientific knowledge, is like, or based on, a set of social skills: It is what the child or the stranger must know before they understand what it means to go on in the ‘same way’, whether the same way is what is required at a cocktail party or required as an audio-typist or required as a member of the community of physicists, mathematicians, parapsychologists or laser builders. (Collins, 1992: 57)

Exemplar 3: Controversy and disagreement in science Collins’ TEA laser example is perhaps not totally convincing. One may of course concede that science is often complicated and that it requires a number of different practical skills that are best learned by way of hands-on experience. But this does not necessarily imply that replication only concerns these types of skills. In the case of the TEA laser, replication did succeed in the end, even if it was not easy to do it. The example is thus not inconsistent with a notion that it was based on some universal principles, even if they were perhaps poorly understood. Instead of focusing on replication that succeeds, then, one may perhaps look at cases when an announced discovery is not confirmed. This brings us to a cherished genre within SSK, pioneered by Collins: 103

Holm

The controversy study. Controversies are great entry points for those who want to study the making of scientific knowledge because they present a context in which underlying positions and usually taken-for-granted assumptions are brought out into the open. What ordinarily lies hidden is now exposed and opened up for study (see also Latour, 1987). Let us take as our exemplar Collins’ study of the controversy over the discovery of gravitational waves (Collins, 1992; Collins and Pinch, 1994). It is generally agreed that Einstein’s relativity theory predicts that moving masses will produce gravity waves. The problem is that these are very weak. However, certain type of violent events in the universe – exploding supernovae, black holes, binary stars – should produce gravitational waves in the form of slight oscillations in the value of “G,” the gravitational pull. Now, a technology for detecting gravitational radiation was developed by Weber in the late 1960s. The idea was to measure the strain on a massive aluminum bar caused by the changes in gravitational attraction between its parts. Such strains are too small to be measured directly, but by making the length of the bar exactly right, the frequency of the gravitational radiation will set off vibrations in the bar. These vibrations can then be measured. The instrument Weber constructed will measure any vibration in the bar, however, and cannot distinguish between those caused by gravitational waves and those produced by other forces. It thus has to be insulated from all other potential sources of vibrations, be they electrical, magnetic, thermal, acoustic, or seismic forces. Despite all these precautions, there will always be some vibrations in the bar, in part because there will be some movement in the atoms of the bar itself as long as the temperature is above absolute zero. So the normal output of the instrument is random noise, typically registered on graph paper as an irregular, wavy line. A gravitational wave will then be recorded, it is postulated, as a particularly high peak that rises above the “normal” noise level. Relying on this instrument, Weber in 1969 announced the detection of gravitational waves (Collins and Pinch, 1994: 91-94). This – which would be a great achievement if it could be confirmed – led to a number of attempts by other teams of physicists to build similar instruments in order to replicate Weber’s results. This did not go well for Weber, and his claims of having “discovered” gravitational waves were rejected (Collins and Pinch, 1994: 94-95). The key point for Collins concerns how agreement on claims such as those of Weber is reached within a scientific community. The starting point of Collins’ argument is that neither the

104

The Invisible Revolution

gravitational waves themselves, nor the machine Weber built to measure them, can supply the answer. If scientists knew how much gravitational radiation were around, they would also know whether the instruments built to measure them were working. If scientists knew that the instrument were working properly, they could use them to measure the amount of gravitational waves. But neither of these things are known: Thus, what the correct outcome is depends on whether there are, or are not, gravity waves hitting the earth in detectable fluxes. To find this out we must build a good gravity wave detector and have a look. But we won’t know if we have a good detector until we have tried it and obtained the correct outcome. But we don’t know what the correct outcome is until ... and so on ad infinitum. (Collins and Pinch, 1994: 98)

This is what Collins calls the “experimenter’s regress.” Such a regress can only be broken with reference to already accepted beliefs. If scientists already believe in a phenomenon such as gravitational waves, they tend to recognize the competence of the experiment that detects it; if they disbelieve it, they tend to reject it: “It is not the regularity of the world that imposes itself on our senses but the regularity of our institutionalized beliefs that imposes itself on the world” (Collins, 1992: 148).

Exemplar 4: Scientific knowledge and social interests For Collins, scientific knowledge rests on institutionalized belief. Such belief is integral to scientists’ participation in a given “form of life,” that is, the taken-for-granted ideas, rules, methods, techniques and practices shared by the participants in a scientific community. Controversies can then be understood as confrontations between different forms of life. Collins' focus is on the negotiations that occur when claims are made that for some reason or other do not fit easily with most or the dominant participants in a given form of life. His “empirical programme of relativism” (EPOR) is thus one of micro-sociology (Collins, 1992:149). This, however, should not be taken as a denial that there are macro variables at work when scientific controversies are settled. On the contrary, Collins suggests that microscopic explanations always refer, explicitly or implicitly, to the influence of a cultural/structural context. For the analysis of such influences, Collins refers to the work of the Edinburgh School of SSK and the interest model (cf. Pickering, 1992: 5). 105

Holm

This model can be broadly characterized as follows: Scientific knowledge cannot be seen as some abstract body of theories, but rather consists of expressions of specific forms of life, embodied as paradigmatic exemplars of scientific achievements serving as models for further activities. But how do we know when some new event is consistent with a given model? What is a correct match of model and what is modelled? Here, the interest model introduces concrete goals and purposes to give the answer. A successful model is a pragmatic accomplishment, something which those who evaluate it take to serve their purposes. Success occurs at that point in the process of adjustment and adaptation when model (exemplar) and modelled (problem) are brought into maximally effective relation for practical purposes. And this in turn, if it is correct, means that to understand modelling and the use of exemplars in science it is necessary to consider practical purposes of the users. (Barnes, Bloor, and Henry, 1996: 109)

Scientific “modes of life” or paradigms are typically goal-oriented: “Their organized activities must generally be related to purposes, aims, goals, objectives, interests, or some such categories, if they are to be made intelligible” (Barnes and McKenzie, 1979: 52). Following Habermas, Barnes and McKenzie here suggest that one should think of paradigms as resources or instruments. Paradigms can be interpreted as calculative procedures scientists can use in order to do things. When scientists evaluate one paradigm against another, they will look at their perceived and potential achievements rather than their formal characteristics: “Where scientists disagree in their choice of paradigms one should hence always check whether they do not differ also in the instrumental interests which pre-structure their evaluations” (Barnes and McKenzie, 1979: 52). But such interests are not to be understood in individual terms: The situated patterns of instrumental interests in terms of which scientists construct and evaluate knowledge-claims, and from time to time, compare paradigms, are generally in turn related to the fact that their activity is socially structured, and related to a set of social interests. (Barnes and McKenzie, 1979: 53)

Let us now turn to a case-study where this model is put to use: McKenzie's analysis of the controversy between Karl Pearson and George Udny Yule over how best to measure statistical association (McKenzie, 1978; see also Barnes and McKenzie, 1979; Woolgar, 1981). In this controversy, Pearson argued that in the case of nominal data, association was best measured by his “tetrachoric coefficient of correlation,” rT , assuming that the observations under scrutiny

106

The Invisible Revolution

correspond to more basic interval-type variables that follow a bivariate normal distribution. Yule protested, arguing that such an assumption was quite often not fulfilled, making Pearson’s coefficient invalid. In its place, he proposed another measure of association, Yule's Q, which treated the data as if the variables in question were categories rather than of the interval type. The controversy was intense and protracted, lasting from 1900 to 1914. Importantly, it was not settled by agreement that one side was right and the other wrong: … although doubtless there were ‘reasonable men’ on both sides of the controversy, abstract argument could not hope to provide sufficient reason for choosing one side or other side. Thus, Pearson regarded it as a defect of Yule’s coefficients that they could indicate complete association between sets of data generated from incompletely associated underlying variables; for Pearson this indicated a failure of measurement. Yule, however did not accept the notion of association which was used to define the failure of his coefficient here. He continued to define perfect association in his own way (…), which vindicated his own practice and in turn cast doubt on aspects of Pearson’s work. (Barnes and McKenzie, 1979: 57-8)

Since reference to some abstract notion of truth cannot explain the controversy, McKenzie and Barnes refer us to social interests. Pearson was committed to the practical utility of statistical techniques in prediction and control. The overarching motive for his work was his interest in eugenics, a social Darwinist ideology with a considerable following at the time. In order to put such ideas into practical politics, statistical instruments that were able to show the heredity of given mental and physical characteristics were needed. His rT allowed for that. Yule, on the other hand, was not particularly preoccupied with eugenics. His methods were more influenced by an interest in associations between nominal data in their own right. But here McKenzie cautions us about individualist explanation, and goes on to explain how the two opponents can be seen as representatives of broader social interests. Hence, Pearson was a typical member of the rising professional class who developed and adhered to ideologies that “on the one hand, emphasized its difference from, and superiority to, the manual working class, and on the other, pointed to the social value of professional knowledge and skills as against the mere ownership of capital or land” (McKenzie, 1978: 67). Eugenics was just such an ideology. Yule’s background, in contrast, was in an old-established family of army officers, Indian civil servants and orientalists. He was a member of a downwardly mobile traditional élite with generally conservative tastes and a dislike of eugenics (ibid: 70).

107

Holm

So statistical techniques as correlation and regression, usually thought of as value-free, neutral techniques, turn out to carry with them the assumptions, conditions, and politics of the times in which they arose. In the controversy between Pearson and Yule, the underlying interests were those connected with eugenics. But arguments of similar type have a wide applications: … if 19th century professors at Edinburgh University believed in phrenology it was because of their class location (Shapin, 1975); if 20th century medical researchers believed that sperm were intrepid, forceful, and resolute it was because of the power of masculinist thought (Harding, 1986); and if Walter Isard believed in the practical efficacy of regional science then it was because he shared a peculiarly white masculinist ‘can-do’ American mindset (Barnes 1996: 130-136). In each case, the world is constructed not as it is, but according to the dictates of the social interests of those putting forward such representations. (Barnes, 1998: 205)

Taking their lead from Kuhn and Wittgenstein, sociologists like Barnes, Bloor and Collins rejected rationalist and realist understandings of science. Against the rationalist notion of scientific knowledge as bound up in reason and logical thinking, SSK objects that there is no determinate relationship between a rule (concept, theory) and its next application. Against the realist notion of correspondence between theory and real objects, SSK objects that such relationships cannot be tested without assuming that which is in question. Any attempt at founding scientific knowledge in reason and truth will thus fail. SSK's alternative is to ground scientific knowledge neither in abstract logic nor objective reality, but in society and culture. Scientific knowledge is a social institution, reflecting not objective reality as it is, but as it is engaged through societal goals, interests, power and politics.

Science as Knowledge Let us return for a moment to Lakatos (1978) and his division between the first world of matter, the second world of feelings, beliefs and consciousness, and the third world of objective knowledge. In this picture, the third world of knowledge will achieve some correspondence with the first world of matter, while the second world, which contains all those trivial details of how science is actually carried out, is mostly a distraction which need not be of any concern unless

108

The Invisible Revolution

false knowledge claims happen to be mistaken for the real thing. As we have seen, SSK offers a different take on this picture. Here, the first world of matter is pushed far out into the distance and made virtually unreachable. Then, the order of priority between the second and third world is reversed, so that the world of knowledge – no longer to be taken as objective but as socially constituted – rests firmly on the second world of concrete scientific practices, understood as belief, values, goals, traditions, interpretive frameworks, methodologies, script, etc., in short, as culture. The important thing to note here is that, although SSK reverses the priorities among Lakatos’ three worlds and to some extent redefines their respective contents, the fundamental trichotomy is kept intact. As Hacking (1983: 3) points out, when people have strong disagreements about something, they will often agree about fundamentals. In the quarrel between the philosophy of science and SSK, there is a common understanding of science as knowledge, as ideas, beliefs, theories, meanings and concepts. Science thus perceived is fundamentally about representations, about things that stand for or are taken to represent something else. The interpretation of SSK as remaining in a representational mode (Pickering, 1992; 1995) requires some justification. As we have seen, SSK employs a number of arguments that are apparently at odds with the idea of science as primarily conceptual and discursive in nature. SSK was influenced by Kuhn’s notion of a paradigm, rejecting conventional models of science as essentially theoretical structures. Instead, scientific paradigms are built around sets of exemplars, concrete scientific achievements offering (open-ended) models of what can be seen as relevant types of problems and adequate types of solutions. Employing Wittgenstein’s notion of science as a “form of life,” SSK scholars also seem to be distancing themselves from science as representational, focusing instead on the way scientists are engaged in practical attempts to master the material world. The Edinburgh version of SSK even wants causal analyses of science. “Interests,” says Bloor (1991: 173), “don't have to work by our reflecting on them, or interpreting them. Some of them, some of the time, just cause us to think and act in certain ways.” Isn't this suggesting that SSK moves away from the notion of science-as-knowledge and the representational perspective? While the seeds for a break with representationalism certainly were there, it nevertheless seems fair to say that SSK did not follow this through. SSK, was, as the K insists, the sociology of scientific Knowledge. This can be seen in SSK's use of Wittgenstein and the notion that there

109

Holm

can be no unique, logically compelling way to infer from a rule (or concept, theory, belief, or meaning) to the next application of that rule. This would seem to imply that the whole notion of science as knowledge must be abandoned since the link between signifier and referent breaks down. This is, as I understand it, the ethnomethodological interpretation of Wittgenstein, as summed up in the notion of “indexicality” (Lynch, 1992; Bloor, 1992). But this is not the way SSK uses Wittgenstein. Instead, SSK introduces the social in order to save the coherence of science as knowledge. While inferring from a theory to its next application will be impossible if we rely on logic or other rational procedures alone, it can happen anyway because scientists are embedded in a context of goals, purposes and social interests. This is, as we have seen, Collins’ sociological solution to the problem of induction. Science is still perceived of as knowledge, as relatively consistent conceptual networks of propositions. The coherence and inner life of this network derives basically from the status of concepts as social institutions: We perceive regularity and order because any perception of irregularity in an institutionalized rule is translated by ourselves and others as a fault in the perceiver or in some other part of the train of perception. (…) Our regular judgements of similarity and difference are made with no more difficulty than attends the categorization of experimental activities once the proper outcome has been defined. The open system quality of these judgements disappears when the appropriate outcome is known. After this there is only one proper continuation of the series. (Collins, 1992: 147-8)

SSK remains in the representational mode. Instead of discarding Lakatos' trichotomy, SSK only substitutes the second world, society, for the first world, nature, as the fundamental reference point of knowledge. While this allows SSK to talk about scientific exemplars, practices and even causal analysis, it cannot be taken to mean that science in any way has access to the material world directly. Nature is safely stowed away behind the impenetrable barrier of interpretive flexibility, and can only enter as “reflected upon” by human agents (Pickering, 1992; Pickering, 1995: Chapter 1; Schaffer, 1991). In this way, practice for SSK cannot really leave the realm of the social. It stays securely with human actors involved in meaning production and interpretation. Science is perhaps practice, but only to the extent that practice is culture and culture is knowledge.

110

The Invisible Revolution

Is SSK Reflexive? As we have already noted, “symmetry” was a slogan coined by the strong program in order to defeat the Mertonian perspective, according to which the analyst would shift explanatory repertoire when s/he moved from true to false knowledge claims. True knowledge claims don’t need explanations, since they simply correspond to facts. False or discredited knowledge claims, on the other hand, need to be explained since they do not correspond to facts. For such claims only Mertonians will be interested in social explanations, like political pressure or biased perception caused by prior commitments, etc. For SSK, this is wrong because it assumes what should be explained, namely how “true” and “false” knowledge claims – fact and fiction – can be distinguished in the first place. The accusation of asymmetry, however, can in effect be turned around against the strong program itself, albeit in a slightly different way. Now, the asymmetry lies in the treatment of knowledge claims that concern the state of nature and those that concern the state of society. For SSK, knowledge claims concerning natural phenomena cannot be taken at face value, since observation and experience of nature always lend themselves to different interpretations – an idea inherent in Collins’ notion of the “experimenter’s regress” and generally referred to as the principle of “interpretive flexibility” (cf. Pinch, 1986: Chapter 7). In sharp contrast, no such problem is acknowledged when it comes to knowledge claims that concern how societal forces – social interests, cultures or forms of life – come to determine scientific knowledge. While SSK sociologists are relativists when it comes to knowledge about natural facts, they are realists when it comes to their own knowledge about social facts. Whereas nature can never be tricked into legislating among different knowledge claims about it, society will not hesitate to do so when asked by sociologists. Nature is flexible; society is definitive. So SSK is asymmetrical in its treatment of the way nature and society influence knowledge. This, again, allows SSK to assume for itself epistemological privileges that it denies natural scientists. Since nature is surrounded by an impenetrable barrier of interpretive flexibility, knowledge claims about it can never be objectively true. To SSK, no such problems exist when society and social causes are concerned. In practical terms, then, SSK’s asymmetrical position with regard to natural and social causes of knowledge, respectively, makes its commitment to

111

Holm

reflexivity somewhat empty. That is, the assumptions it routinely makes for the knowledge claims it subjects to analysis becomes highly problematic when extended to its own knowledge claims. As we have seen already (p. 94-96), SSK acknowledged reflexivity as the fourth tenet of the strong program. This tenet remained a paper tiger, however, and was never really allowed to have any effect in SSK’s empirical analyses. Michael Mulkay, who adhered to the strong program in the early part of his career, but has later pursued science studies under the labels of “discourse analysis,” has commented on his own early position, which he takes as typical of the SSK perspective within SSK. During the early 1970s, he was engaged in attempts to account for the social production of scientific knowledge, while putting the relevance of SSK for sociology itself in parentheses. Instead of a reflexive approach – in which the sociologist’s explanations would not be privileged in relation to the natural scientists’ explanations – the SSK accounts conformed to the scientists’ idealized models of natural phenomena. The approach was scientific in its mode of analysis, in its use of linguistic forms, and in its refusal to acknowledge its own social location (Mulkay, 1991: xv). SSK took over from the natural scientists and assumed for itself precisely the assumption it denied the scientists they studied. Instead of symmetry and reflexivity, as claimed by the strong program, we see asymmetry and non-reflexivity here. In accounting for his own change of position, Mulkay refers to the problem caused by “the proliferation of accounts:” … the more closely we studied our data in order to establish the nature of participants’ actions, the more confusing and internally irreconcilable those data became, due to participants’ ability to reformulate and interpret their own and others’ actions in an apparently open-ended fashion. This caused us difficulty because the language of social science and its conventions of reportage are unitary in character; that is, they presuppose the existence of one real social world which is available for coherent representation in the world of social science. But the harder we tried to capture the one real social world of science, the greater was the multiplicity of divergent accounts of that supposed world that was revealed and/or generated by our attempts at understanding. (Mulkay, 1991: xvi)

It gradually became apparent that these problems, when taken seriously, would have important consequences for the sociology of science. If social phenomena are as open to different interpretations as natural phenomena, how can it be assumed, as SSK does, that the former can explain the latter? If social scientists no more than natural scientists can claim independent access to reality, be it social or natural, how can SSK expect anyone to take its explanations of

112

The Invisible Revolution

knowledge seriously? Why should we prefer SSK accounts over the natural scientists’ own accounts if both are socially constructed? Actually, the problem of reflexivity can be solved, as revealed by the “reflexive turn” within science studies. In one version, pursued by Mulkay and others under the label of “discourse analysis,” conventional SSK analyses themselves were turned into objects of analysis, revealing the constructional work involved in making social science out of interview material, in much the same way as natural scientists make knowledge out of raw data (Mulkay, Potter and Yearly, 1983). In a later version, the program labeled “new literary forms” (Mulkay, 1991) challenged the conventional linguistic form of science. Instead of single authorship, the conventional form of constructing certainty in scientific texts, texts produced under this program applied multi-vocality to escape this type of “scientism” (see Ashmore, 1989; for an ethnomethodological approach to the problem of reflexivity, see Lynch, 1992). While the reflexive turn thus solves the problem of asymmetry, saving science studies by refusing to assume for itself epistemological privileges it denies others, this transforms science studies into a somewhat different project from what they used to be. There can no longer be much hope of explaining knowledge claims, least of all in the form of causal explanations that SSK wanted. Instead of explanation, the most one can hope for are modest descriptions of how scientists go about negotiating the meaning and adequacy of different accounts. What rhetorical and literary techniques do scientists employ in order to infuse authority into some accounts and detract authority from other accounts? Here, self-referentiality is no longer a problem. Whether reflexivist do or do not employ the same rhetorical techniques as conventional scientists typically do, they do so because they want to convince others, not necessarily because they believe they have found the truth. In this way, the self-referentiality problem by no means defeats all versions of science studies. While the impression of self-refutation is one of the functions reflexivity is regularly made out to perform, a seasoned discourse analyst will usually have no problem showing how it can be put to effective use as a resource for science studies (Ashmore, 1989). I think it is safe to say that SSK has not joined the reflexive turn. For authors like Collins, Yearly and Pinch (Collins and Yearly, 1992; Collins and Pinch, 1994), the reflexive argument is a bad thing, used by enemies of the strong program to make trouble for it (cf. Trigg, 1978; Laudan, 1982; Ashmore, 1989: 42). To Collins and Yearly (1992: 308), reflexivity ends in a self-

113

Holm

referential quagmire, in an endless relativist regress: “[F]ollowing the lead of the relativists, each new fashion in SSK has been more epistemologically daring, the reflexivists coming closest to self-destruction.” Instead of moving closer to “epistemological bedrock” by the route of reflexivity, however, this can only lead to unproductive navel-gazing. Therefore, Collins and Yearly no longer want to be relativists. Since there are no safe routes to epistemological firm ground, particularly not via reflexivity, they instead retreat to social realism: Natural scientists, working at the bench, should be naive realists - that is what will get the work done. Sociologists, historians, scientists away from the bench, and the rest of the general public should be social realists. Social realists must experience the social world in a naive way, as the day-to-day foundation of reality (as natural scientists naively experience the natural world). That is the way to understand the relationship between science and the rest of our cultural activities. Close description of the human activity makes science look like any other kind of practical work. (Collins and Yearly, 1992: 308-309)

Rejecting reflexivity, which can only undermine SSK's key explanatory resources and weaken the humanist struggle against technocracy, they jump off the relativist band-wagon.5

Does SSK Believe in Reality? SSK rejects the notion that scientific knowledge at its best is simply true, that is, that it accurately portrays objective reality. Instead, SSK argues that truth is a social category, and that science and scientific knowledge are social institutions and thus based on conventions. It is, says SSK, the social world that orders our perceptions of the material world. Phrased like this, the SSK position sounds like a radically idealist and relativist one. If the material world does not act as a referee with regard to contradicting claims about it, all knowledge claims will be equally valid. Astronomy cannot be distinguished from astrology on the basis of the way their claims correspond to objectively real entities and causal mechanisms. If there are differences in these bodies of

5

Although there certainly are differences between the positions of Collins and Yearly on the one hand and the Edinburgh School on the other (cf. Barnes, Bloor and Henry, 1996: 76), it would seem that they come fairly close in their reaction against the reflexivist challenge. Hence, Barnes, Bloor and Henry (1996:87-88), in an extended presentation of the strong program which miraculously manages to avoid discussing reflexivity (p.xxii), recommend a naïve realist position that sounds much like that of Collins and Yearly (Barnes, Bloor and Henry, 1996: 87-88).

114

The Invisible Revolution

knowledge, it is because they are differently situated in the dominant networks of interests and power. In this interpretation, then, the SSK position not only embraces the notion that the ideas, concepts and theories science employs to get at objective reality are socially constructed – not a very risky claim – but also the notion that the entities and processes that are referred to by such concepts and theories are in some way socially constituted. Following Sisimondo (1996: 8), we can label this a neo-Kantian position. Perhaps the best way to understand this position is to start with the notion of social institution in SSK. As we have already seen, SSK defines social institutions in terms of bootstrapped induction (Barnes, 1983): Social institutions exist because people believe in them; they can be thought of as realities created by reference to them (Bloor, 1996: 842; Barnes, 1983). This is of course not especially problematic when it comes to entities like money or marriages, or even truth. In order for them to be real, they must be treated as real by the people bound by them. They have no objective existence apart from people’s beliefs and actions; they must be enacted to exist. There is a causal arrow from what people think and do with regard to such entities to the structure of the real world. Now, the neo-Kantian position, in Sisimodo’s rendering, implies that such causal mechanisms will also exist in the case of what we would normally think of as non-social things, that is, purely material entities and processes: … I use ‘neo-Kantian’ as a label for the family of positions asserting that there is some special causal or semantic connection from what scientists say (or do) to the structure of the material world. Neo-Kantian constructivists assert that scientific consensus has some direct material import, that representations create their own objects. (Sisimondo, 1996: 8)

In this way, then, a neo-Kantian constructivist would assert not only that the idea of, say, quarks, is socially constructed, but that also quarks as material entities are constituted by the agreement among physicists of their existence (Pickering, 1984; see Hacking, 1999: 29-31). The question is: Doesn’t SSK believe in a material world? Or rather, is SSK denying the existence of an independent, material reality, as in the neo-Kantian position outlined above? In contrast to Sisimondo (1996) and Hacking (1999), who do not think that SSK is neo-Kantian, I believe that early SSK came quite close to such a position. In its programmatic statements, SSK did not deny that there is some objective reality out there. Bloor (1991: 33) said: “No consistent sociology could ever present knowledge as a fantasy unconnected to our experiences of the material world around us. We cannot live in a dream world.” The real world does exist. And the 115

Holm

first tenet of the strong program even allowed for non-social factors – perhaps even material reality? – to influence belief: “Naturally there will be other types of causes apart from social ones which will cooperate in bringing about belief” (Bloor, 1991: 7). Collins (1992) does not tell his readers that the material world does not exist. Instead he asks them to “treat descriptive language as though it were about imaginary objects” (1992: 16). Collins’ position is not a neo-Kantian one, but one of methodological relativism (Collins, 1992: 185; see also Bloor, 1991: 158), which is not incompatible with a mild form of realism (Sisimondo, 1996: 114). Notwithstanding SSK’s acceptance of the real world in principle, however, this reality is not allowed to make much of a difference. For Bloor, correspondence between some independent reality and knowledge claims about it cannot be used as a criterion for truth: “We never have the independent access to reality that would be necessary if it were to be matched up against our theories” (Bloor, 1991: 40). In a later discussion, Barnes (1992: 134) explicitly rejects a radically relativist or idealist position and declares that the Edinburgh variety of SSK is: … perfectly happy to accept that different cultures inhabit one single, shared, external reality. On this view it is not the lack of an external world which leads to the equivalence of different bodies of knowledge, but the silence of that world in the face of alternative accounts of it, its evenhanded indifference, its tolerance. There is one world, but there are many alternative equally reasonable accounts of its nature and of how it behaves. This is relativism with a realist flavour: it could even be described as a form of realism. (Barnes, 1992: 134)

While Barnes thus confesses to a mild form of realism, he still denies that we can obtain any true knowledge about the real world. The acceptance that an objective world actually exists, then, does not make much difference in the analysis. It is written into the plot, as it were, but does not have any bearing on the action. This is even more striking in the paradigmatic case-studies of concrete scientific controversies, as exemplified above. Here, the main thrust of the analytical effort is precisely to show how societal factors can be mobilized to explain the settling of scientific controversies. The most important alternative explanation, which is seen as defeating the SSK account and against which most of the analytical work is directed, is that the controversies in question were settled by rational agreement, or by some evidence linking the claims of one side to the real order of things in some definitive way. While it may be that SSK, as suggested by Barnes and Bloor, in principle allows that there may be other things than social interests that will influence belief, the thrust of the exemplary case studies is to deny this. The case studies, then, much more

116

The Invisible Revolution

than the theoretical propositions, conform to radical relativism and approaches a neo-Kantian position. In the picture SSK offers us, material reality and nature are still present beside society and culture. But almost all the action is on the social side; very little happens on the side of nature. SSK may believe in a material world. This world, however, is one that is made to be completely silent.

SSK and the Invisible Revolution SSK and neo-institutionalist perspectives are of course parts of the same theoretical movement. They are both best identified by what they are opposed to, realist and rationalist perspectives. They are both social constructivist, insisting that phenomena and objects that look natural and objectively given – and are reified as such by realist and rationalist theories – in reality are socially constituted. But there are of course also crucial differences between SSK and neo-institutionalist perspectives. The most important is that SSK has made its key topic precisely that which neoinstitutionalists systematically avoid: Science and scientific knowledge. As was argued in the previous chapter, institutionalists have stayed safely away from this topic. If it features in their analyses, it is in a superficial way, as symbolizing the rationalistic obsession of the Western cultural configuration. In sharp contrast, SSK wants to study how science works in practice; how it is organized; how the work is conducted, the interaction between scientists; what instruments and resources they utilize; what values, beliefs and perspectives they hold; how they come to agree that some claims are true while others are false. It is in SSK’s analyses of such issues that we recognize it as a version of institutionalism. SSK’s basic faith with respect to knowledge claims parallels exactly the fundamental institutionalist perspective on social institutions: Institutions cannot – as realist and rationalist perspectives will have it – be taken at face value as objectively given and external to human affairs. For SSK, scientific knowledge is a social institution. As such, knowledge claims are valid to the extent that they are believed in and accepted as valid. In the previous chapter, we saw that while institutional perspectives are more than willing to produce the invisible revolution, they are not very capable, since they are not equipped to handle science and scientific knowledge as institutions. This would seem to imply that SSK,

117

Holm

understood as an institutionalist perspective specializing in science, is perfect for the job at hand. Let us now take a closer look at the way SSK can be used to make the invisible revolution visible again. As we recall from Chapter 1, the issue at hand concerns the transition between two sector systems in Norwegian fisheries, from the MSO system, which dominated the sector until the 1970s, to the resource management system, which replaced the MSO system during the 1980s and early 1990s. I wanted to see this as a revolution, indicating that it was a fundamental transformation, involving not only changes in practices, organizational forms, property rights, basic goals and values, the definitions of the sector’s fundamental problem situation, and so on, but also important shifts in the political structure of the sector, the establishment of a new dominant sectorpolitical coalition, and important changes with respect to the authorities’ regulation of the sector. The problem with this interpretation was that virtually no-one else seemed to recognize the extent and importance of these changes. If they were noticed, they were certainly never described in terms of a revolution. My theory was that there had been a revolution, but that it somehow had remained invisible: The invisible revolution. One way to explain the invisibility of the revolution, inspired by institutional perspectives, is to focus on the mechanisms by which institutions are accounted for and naturalized. An institution must, in order to be stable, be accepted and taken for granted. Such taken-forgrantedness is essential for institutions to function; they exist to the extent that they are believed in and enacted. It follows from this that when a new institutional configuration takes over from an old one, the accounts and legitimizing stories that once allowed it to be taken for granted are no longer accepted. This will mean, in general, that a transition between two institutional configurations, once completed, will not be perceived as a great change by those involved. This is because the past will be interpreted through the perspectives, values and stories of the present institutional configurations. Seeing the changes between the two systems as a revolution requires that they are understood as two autonomous and equally valid belief and value systems. But maintaining such a view is extremely difficult, particularly for those who are committed to and bound by the dominant institutional configuration. In short, the institutional perspective will explain the invisible revolution by pointing out the mechanisms by which a new institutional configuration can rewrite the history and rationality of the institutional configuration which it replaced. The two

118

The Invisible Revolution

systems can no longer be seen as equally possible, internally consistent and valid forms of life; hence the revolution that occurred when one replaced the other cannot be seen. Such a viewpoint, argued above from an institutionalist perspective, is perfectly compatible with an SSK analysis. The easiest way to see this, perhaps, is to utilize the concepts of paradigm and paradigm shifts of Kuhn. Just as in a Kuhnian paradigm shift, the forms of life under MSOs and resource management will be incompatible. Different exemplars will be guiding practice; different world views and values will be held; different methodological prescriptions and standards of verification will be valid. In the analysis of Norwegian fisheries, we might want to replace some of these concepts, designed to get at the essence of scientific practices, with concepts better able to grasp the practices of fishing. But such replacements would not change the basic thrust of the argument. As is the case with a paradigm shift in science, the revolutionary character of such a shift within the fisheries will become invisible because the winning side gets to rewrite history. In Kuhn’s words: When it repudiates a past paradigm, a scientific community simultaneously renounces, as a fit subject for professional scrutiny, most of the books and articles in which that paradigm had been embodied. Scientific education makes use of no equivalent for the art museum or the library of classics, and the result is sometimes drastic distortion of the scientist’s perception of his discipline’s past. More than practitioners of other creative fields, he comes to see it as leading in a straight line to the discipline’s present vantage. In short, he comes to see it as progress. No alternative is available to him while he remains in the field. Inevitably those remarks will suggest that the member of a mature scientific community is, like the typical character of Orwell’s 1984, the victim of a history rewritten by the powers that be. Furthermore, that suggestion is not altogether inappropriate. There are losses as well as gains in scientific revolutions, and scientists tend to be peculiarly blind to the former. (Kuhn, 1970: 167)

Relying on SSK as a resource, we will have no problem transporting Kuhn’s scientific revolution into Norwegian fisheries. We can now retell the story of the introduction of resource management as a paradigm shift. The goal is clearly in sight; the revolution is beginning to become visible again. Can we thus declare that our mission has been accomplished? Unfortunately, there are some problems with the way SSK achieves its goal. As we have seen in this chapter, SSK secures its explanatory control over scientific knowledge by abolishing objective reality, at least as an active participant in social affairs. SSK can explain scientific knowledge socially only because

119

Holm

it assumes that material objects will be indifferent and refuses to distinguish true from false claims about their identity. SSK’s account of knowledge claims (“they reflect social interests”) can beat the scientists’ own accounts (“they reflect objective reality”), because nature itself is rendered virtually silent. This is exactly the same move that allows Kuhn to substitute his notion of scientific revolutions for the scientist’s notion of scientific progress. We get revolution instead of progress because there cannot, for Kuhn and for SSK, be true knowledge. The visibility of the revolution thus comes with the price-tag of relativism. Instead of one truth, we get many different but equally valid interpretations. If one interpretation wins over another, that does not prove its truthfulness, but must be explained by the power resources that were mobilized in its defense. The problem here is that SKK refuses to allow material objects to make a difference in social affairs. Let us take a closer look at this problem in the context of the fisheries, returning first to the paper with which this chapter started, namely “Constructing Alternative Natures” (Paper 5). As we remember, that paper focuses on two contrasting interpretations of the structure of cod resources. The first, institutionalized in the Norwegian (and Russian) resource management regime, insists that the resource is a unitary one, but allows for the existence of coastal cod for distributional purposes. The second interpretation, promoted by a North Norwegian research institution (Fiskeriforskning) and a loose coalition of North Norwegian municipalities, social scientists and activists, proposed that the coastal cod was an autonomous stock (or stock system), independent of the Barents Sea cod and in need of explicit management. Now, when these two images of the structure of the resource confronted each other, the former survived while the latter was defeated. In Eythórsson’s explicitly social constructivist reading of the paper: The authors describe empirically how the Ministry has used its superior power position in order to strike down both the political initiative taken to have local management of local fish stocks tested out and Fiskeriforskning’s own research priorities. They concluded that the reason for the survival of the national management regime for northeast Arctic cod is not that it is based on a ‘true’ model of nature. It survives, they say, because it is supported by sufficient organizational and political power to maintain the validity of their own definition of nature and render alternative definitions harmless (Holm, Rånes and Hersoug, 1989: 89). (Eythórsson, 1999: 8-9; my translation)

This does not mean, however, that power suppresses the “truth,” here in an objective sense, and that the fisheries establishment’s categorization of cod is false while the alternative image

120

The Invisible Revolution

promoted by the North Norwegian coalition is true. Nothing is really true or false in a constructivist perspective; it is all a question of power: In the constructivist spirit, the authors also express skepticism towards the alternative knowledge. They bring us the question of whether the existence of local stocks is also a question of social constructions or institutional artefacts. The constitution of local stocks requiring specific management procedures would serve Fiskeriforskning as an institution, give it legitimacy, leeway for expansion, and power, would it not? We get the impression that from their [the authors’] point of view it will not be possible to test alternative hypotheses about which of the categorizations of the cod stock is the most valid, from biological criteria alone. The criteria for validation, it seems, will have to be defined according to power relations, not only within a community of researchers, but within the broader resource management discourse. The authors’ conclusion seems to be that as long as one does not succeed in mobilizing sufficient political power behind knowledge of local stocks, such knowledge will remain invalid. (Eythórsson, 1999: 9; my translation)

Truth, as established and contained within institutions, is nothing but a social phenomenon. It relies singularly on power, taken here as political power constituted by social institutions. Invoking the contrast between the community of researchers on the one hand, and the “broader resource management discourse” on the other, the paper, as read by Eythórsson, implies that the final validation of knowledge claims ultimately comes from outside the scientific community, from society and politics. There are strong causal arrows from Society towards Nature, but weak or non-existent causal arrows from Nature to Society. Consider now how we could explain, in the SSK style adopted by Paper 5, that it has been scientifically established, by way of genetic mapping techniques, that the coastal cod constitutes an independent stock system (Fevolden and Pogson, 1997; Fevolden and Christiansen, 1999), and thus that there has been some movement towards recognition that the coastal cod must be managed separately. Starting from within SSK, we are barred from explaining this with reference to objective realities. We thus would have to assume some sort of independent power shift, preferably outside science. There are several candidates here, of course. One source could be a shift in the relationship between Norwegian and Russian authorities, allowing the former leeway to move towards a definition of the cod resource that would allocate more of it to Norway. Another source could be a new stage in modernity itself, be it late- or post-, so that the shift from a unitary 121

Holm

to a fragmented model of the cod “reflects” another shift in which linear, optimistic models of yesterday give way to more fragmented and chaotic ones of today. A third possible source could be the political and cultural mobilization of the Sami, who clearly would stand to gain from the invention of coastal cod as independent of Barents Sea cod. Does this sound plausible? Perhaps. Such things surely must be looked into. Nevertheless, we note that SSK is firmly committed to disregarding the simplest explanation of all, namely that the cod itself, allied with the scientists, can make a difference; that the genetic pattern as established by science is itself strong enough to tip the balance. That a few biologists with the right training and equipment is all it takes to beat the apparently invincible power triangle between Oslo, Bergen and Trondheim. For a related example, consider Finlayson’s (1994) Fishing for Truth. Here, Finlayson, armed with SSK (primarily referring to Pickering, 1984), takes on the scientific crisis triggered by the collapse of the northern cod stock odd the east coast of Canada in the late 1980s and early 1990s. Starting with the premise of interpretive flexibility where nature is concerned, he argues that the failure of fishery scientists to perceive the decline of the northern cod stock during the late 1980s can be explained by the scientists’ social location. As employed by the Canadian state, fishery science was deeply committed to the Canadian authorities’ claim that the management of the northern cod stock was a “success story.” We can here note how Finlayson applies the basic assumption of strong program SSK, as discussed above: While Nature is soft (i.e. scientists enjoy full interpretive flexibility where stock fluctuations are concerned), Society is hard (i.e. the scientists’ interpretations are fully determined by their interests as constituted by their relationship to the State). Now, consider the context in which the crisis of Canadian fishery science occurred, and which Finlayson’s book portrays: The dramatic decline of the northern cod stock during the 1980s, leading to a moratorium from 1993. Finlayson’s main task is to explain why DFO scientists during the 1980s continued to believe that the northern cod was recovering fast, while it actually was declining rapidly. Since this has to be done according to a social constructivist framework, assuming generous interpretive flexibility where natural phenomena are concerned, the only explanatory variables that are allowed in the analysis are societal and political ones. Finlayson’s account here draws on the scientists’ dependence on the state, and, in turn, the state’s need for northern cod management to be a success for legitimation purposes.

122

The Invisible Revolution

One problem here concerns how most people, including the scientists, the authorities, and Finlayson himself, apparently, became convinced in the 1990s that the cod stock in the northwest Atlantic had been depleted and hence that the Canadian fishery scientists’ knowledge of a growing cod stock had been wrong. From an SSK standpoint, which Finlayson consistently sticks to, this cannot be accounted for with reference to the behaviour of the cod itself – that it virtually disappeared – since such behaviour will always be open to interpretation. Instead, the shift of interpretation, from one insisting that “the cod stock is healthy and growing” to one insisting that “the cod stock is almost extinct” came about as the result of “a complex set of negotiations between competing interests with differing kinds and amounts of authority at their disposal and with differing, often conflicting objectives” (Finlayson, 1994: 17-18). We do not get to know what changes in interests, authority or objectives that could have brought about this spectacular shift, allowing the weakest party of all, a disorganized group of inshore fishers, to have their interpretation replace that of the most powerful of all, the coalition between the state, big capital and the scientific establishment (Apostle et al., 1998). In contrast to the case of the coastal cod, it is here very hard to concoct a plausible account for this if we are only allowed to use social ingredients. These two examples serve to show the fundamental thrust of SSK-type explanations. When a shift occurs in knowledge, so that one interpretation replaces another, SSK wants to reject the idea that this has to do with non-human or material factors. Instead, SSK will seek to relate such changes to parallel changes in societal factors, be it in overarching cultural configurations or more local interest constellations. In the case of the substitution of the resource management system for the MSO system, an SSK analysis would feature explanatory variables like overall shift in cultural perceptions (from collectivist and corporatist ideas until WWII, towards more rationalist, market-oriented, scientistic notions in the post-war period); the introduction of the new oceans regime negotiated during the 1970s and 1980s; the tendencies towards internationalization and globalization, which in Norway and for Norwegian fisheries in particular happened via the process of adaptation to the EU; as well as more sector-specific events, like the decimation of the Atlanto-Scandian herring stock during the 1960s. The SSK analysis of the resource management reform in Norwegian fisheries would thus parallel a neoinstitutional account. The only substantial difference is that SSK would not shy away from involving the science of fisheries assessment and management in the analysis. Where an 123

Holm

institutional analysis will not go further than noting that the breakthrough of fisheries management opens the fishery sector to this archetypal modern institution, SSK will be much more detailed and concrete in its analysis of the role science has within fisheries resource management. While SSK confronts the science issue head on, it does so with blinders of an institutionalist type, however. What does SSK find when it looks at fishery science? Will it, like the scientists themselves, see the fish, fish stocks and ecosystems? If it does, it does so only to the extent that these things lead firmly back to the social. When SSK explains fishery science, as in Paper 4, its basic assumption is that fishery science, whatever it does, cannot tell the truth about the fish. Instead, its explanatory project is to show how the scientific model reflects larger cultural patterns or institutional orders, and in particular how – despite its inability to get at the truth – it is a great instrument for making people accept its claims as true, thus helping to stabilize and extend the dominant power constellations. What can we conclude? Despite its problems, SSK of course is a great improvement compared to institutional theory. Where the institutionalists could not offer any practical guidance whatsoever in accounting for science, SSK has a lot to offer, both in the form of theoretical discourse and practical exemplars. With its help, it is fairly easy to take on the analysis of science as an institution in the context of fisheries resource management, thus rendering the resource management revolution visible again. If we had to chose between SSK and institutional theory, then, there could thus be no doubt. Nevertheless, we leave SSK with serious doubt. Isn’t the non-reflexive privileging of its own knowledge claims basically self-defeating? And what is left of society when all material objects have been exorcised? Can it really stand on its own? Is not systematic engagement of material objects and forces exactly what natural science does? Isn’t science’s success in this respect one of the most important sources of power for the Western world. How can we account for this if we at the outset have decided that the material world doesn’t make a difference? Let us see if we can get better answers to such questions when we in the next chapter turn to Actor-Network Theory.

124

The Invisible Revolution

4. ACTOR-NETWORK THEORY

Introduction Paper 7, “Resource Management as a Heterogeneous Network,” analyzes fisheries resource management. As the title suggests, it regards resource management institutions and practices as a heterogeneous network, implying that they bind together and give identity to entities of different kinds, like fish and vessels, fishers and exporters, state agencies and local communities, legal texts and newspapers, fishery scientists and politicians. The paper’s main task is to compare two of the most important competing forms of fisheries resource management. These are, on the one hand, the one currently is in operation in most western fishery nations, and which is often characterized as “hierarchical,” “top-down,” “biology-based,” “state-centered,” “linear,” or “numerical.” On the other hand, it is the co-management model promoted by social scientists (excluding most economists), a model which is usually characterized by some combination of terms like “decentralized,” “bottom-up,” “community-based,” “flexible,” “non-linear,” or “parametric.” The paper emphasizes that both of these alternatives are heterogeneous networks, and as such basically similar and comparable. There are nevertheless important differences between them. One of these, of course, is that the first model is a great success, since it has been put to use, while the second, which largely remains a paper tiger, alive and dangerous primarily within texts, is deemed something of a failure. An important question for the paper is precisely this: Why does the former succeed and the latter fail? In answering this question, the paper looks at how the two models propose different roles for and relationships among the major entities involved. Here, the ANT perspective leads to a somewhat different emphasis than what a conventional social science analysis would do. In particular, the paper has a different view of the widely held notion, traceable to natural as well as social scientists, that the established “biological” management regime is overwhelmingly preoccupied with the fish itself, and ignores social and political issues. The paper turns this on its head and insists that this regime’s modelling of the fish is simplistic, bordering on failure. Its success, then, cannot be explained by the way this regime deals with nature, but what it has to offer in the social and political arena. The fishery biologists are, contrary to what the social scientists claim, bad biologists, but compensate by being good sociologists. Pursuing this theme, 125

Holm

the paper then suggests that the co-management model’s social analysis, which is promoted by its inventors as its main strength, is actually its weakness and an important reason for its failure. Why would important actors like the fishers and the representatives of the state choose the hard work, moral strictures and powerlessness of co-management, when they can have the easy and prosperous life offered by the established regime? It is also apparent that the author of this paper would very much like to pursue the paradoxical line of reasoning to its symmetrical extreme, portraying the sociologists as the better biologists, since the co-management model would allow for a better model of the fish than the simplistic caricature of the established regime. Actually, this was done in an earlier version of the paper (Holm, 1999). But in this version, such an analysis is reluctantly rejected. Instead, the paper blames the sociologists for not exploring the potential of their model with respect to the fish. In contrast to the natural scientists, who happily turn themselves into social scientists when they need to do so in order to stabilize their network proposals, the social scientists truly believe that reality is built by social relations alone. They therefore refuse to engage in negotiations with the fish directly, and instead proceed as if the fish is some social construct in the minds of local fishers: That is a pity. Because, as shown by the analysis of the biologically based resource management model, the strong alliance between the state and the fishery biologists has meant that the fish itself has been ignored. Whoever can construct a model that is more to the fish’s liking, will have a good chance of putting it into effect. It also seems as if the social scientists’ proposal of a harmonious triangle between the fish, the fisher and the community has a lot of potential. But it is difficult to see that this model can be realized as long as the social scientists refuse to negotiate directly with the fish, and instead treat it as a mythological creature that will submit to anything the fishers at any moment will claim it to be. It therefore seems more likely that the social scientists’ model, if it is going to be realized at all, will have to be built by biologists. (Paper 8: 102; my translation)

I suppose that it can readily be agreed that Paper 7 takes positions and draws conclusions that are not very easy to come to terms with, particularly if they are viewed from a conventional social science perspective. One problem, of course, is the paper’s main argument, that the biologists are good sociologists while the social scientists are bad ones. The pragmatic criterion for success that are employed here and lead to this conclusion would seem to need some further justification. What, moreover, can we make of this paper’s repeated line that non-human actors can behave like humans, with will, intentions and interests of their own? When this is done

126

The Invisible Revolution

with respect to entities like ANT, it might be overlooked as an unfortunate linguistic shorthand, substituting the perspective itself for the scholars employing it. But this feeble excuse cannot work when, particularly towards the end of the paper, this figure of speech is extended to the fish itself and the paper talks about what the fish would “like” and proposes that there is a possibility, albeit an unexploited one, for social scientists to “negotiate” with the fish directly. This seems farfetched. Indeed, the commentators on this paper, including the book’s excellent editor and even the person cleaning up the language, found it very hard to accept. When these formulations remained in there, then, it was certainly not by oversight, but because its author (or should I say: “the paper itself”?) insisted. As the reader has now realized, the explanation for all this is that the paper tried very hard to be an exemplar of Actor-Network Theory (ANT). It explicitly identified ANT as its theoretical basis; it briefly outlined some features of this perspective, and then it set out to demonstrate ANT in an analysis of resource management. To a large extent this is a didactic piece, then, written more to demonstrate the particulars of ANT than to use it as a tool in the pursuit of another goal. But of course, the mere 20 pages or so that were available for that exercise were not sufficient, and it would not come as a surprise if most readers leave it – like the editor and the proofer – with more questions than answers. Hopefully, a careful reading of this chapter – in which a more extensive presentation of ANT is attempted – will correct that. I will contend that some of the strangeness of ANT results from its origin in and break with SSK. ANT is best seen, I suggest, as an ambitious attempt at solving the puzzles and dilemmas that became evident, as we saw in the last chapter, in SSK’s attempt to explain scientific knowledge sociologically. How can we account for science and scientific knowledge as social institutions without claiming for ourselves privileged access to reality and truth, an access we want to deny natural scientists? How can we solve such dilemmas, moreover, without ending up in an endless proliferation of accounts, as the reflexivists do? How can we retain the simple insight that the real world is there and makes a difference without importing the realist mistake of using nature to explain our knowledge about it? ANT promises to resolve these problems and dilemmas comprehensively and without contradiction. We will see about this as we go along. But even if ANT can keep good on its promises in this respect, that will, of course, not be enough in this context. We also want to know

127

Holm

whether ANT gives us the resources we need to make the fisheries resource management revolution visible again. The notion of heterogeneous networks indeed looks promising in this respect. If both the MSO system and the resource management system can be seen as heterogeneous networks, it will, it seems, no longer be particularly problematic to see the process in which the second replaces the first as a revolution. But what about the re-introduction of the material objects, as ANT insists upon? Will this not bring us back to the realist starting point, in which resource management pushes out the MSO system simply because this is the rational thing to do? How can ANT retain the objects without returning to the rational world of Lakatos and Merton? Isn’t ANT, when all its strange neologisms are peeled off, simply a return to oldfashioned realism? Let us find out.

The Sociology of Translation ANT can be seen as emerging out of SSK, but breaking with this perspective’s preoccupation with science as knowledge. Where SSK remains in a representational mode, ANT is part of a broader movement of perceiving science in terms of practice, intervention or performance. It can perhaps be said that SSK was interested in science as practice and intervention but was prevented from pursuing such notions because of its commitment to understanding science as knowledge. While its conceptualization of scientific practice remained thin (Pickering, 1992: 5), it inspired work – the so-called laboratory studies – that would fill the concept of practice with considerable detail and dimensions. Now, the laboratory studies approach featured long-term (participant) observation in a laboratory setting, modelled on the ethnographic field-work in a foreign culture. This approach was pioneered by Bruno Latour and published, with Steve Woolgar as co-author, in a volume entitled: Laboratory Life: The Social Construction of Scientific Facts (Latour and Woolgar, 1986). The publication of this book led to many followup laboratory studies (e.g. Knorr, 1981; Lynch, 1985; Traweek, 1988; see Knorr-Cetina, 1994). The result of all this was a new wealth of detailed information on the work that scientists were engaged in; not only dealing with scientific concepts and ideas (i.e. how findings were interpreted and presented for publication) but also the practical efforts and skills required in trying to elicit the identity and cooperation of material entities themselves. This growing interest in the

128

The Invisible Revolution

materialist side of scientific practice is apparent in the titles of key volumes, like Hacking’s Representing and Intervening (1983); Latour’s Science in Action (1987), and Pickering’s The Mangle of Practice (1995). Instead of science as knowledge, these scholars want to explore science as more directly engaging and being engaged by material entities. One can start from the idea that the world is filled not, in the first instance, with facts and observations, but with agency. The world, I want to say, is continually doing things, things that bear upon us not as observation statements upon disembodied intellects but as forces upon material beings. Think of the weather. Winds, storms, droughts, floods, heat and cold - all of these engage with our bodies as well as our minds, often in life-threatening ways. … Much of everyday life, I would say, has this character of coping with material agency, agency that comes at us from outside the human realm and that cannot be reduced to anything within that realm. My suggestion is that we should see science (and, of course, technology) as a continuation and extension of this business of coping with material agency. (Pickering, 1995: 6-7)

The material world is here cast as active instead of passive, as doing work and making a difference instead of being impartial and open to many different interpretations. In contrast to SSK’s emphasis on interpretive flexibility, with many degrees of freedom between scientific practices and theoretical interpretation, this perspective emphasizes the resistance offered by the natural entities conjured up in the laboratory. The miracle is not that “anything goes,” but that scientists from time to time, if they are skillful and lucky, will be successful in their attempts to construct stable constellations of instrumental and conceptual procedures, natural phenomena, and theoretical understandings (see Pickering, 1992: 8-14; Hacking, 1992; cf. Also Ashmore, Wooffitt, and Harding, 1994; Preda, 1999; Breslau, 2000; Pickering, 2000). The materialist turn in science studies comes with an emphasis on the heterogeneity and fragmented nature of scientific practices. This stands in sharp contrast to the reductionist trait in SSK, in which practice is reduced to knowledge, and knowledge to interests. Instead, the turn towards science as intervention and performance comes with an insistence that scientific practice is multidemensional. Hacking’s (1992) provisional list of elements put to use in a laboratory context includes 15 different items, which can be roughly sorted, but by no means reduced to, three broad categories: ideas, things and marks. Referring to Hacking with approval, Pickering likewise emphasizes that all the dimensions of science, be they conceptual, social, or material,

129

Holm

must be seen as “fragmented, disunified, scrappy” (Pickering 1995: 3). On the same note, ANT sees science as attempts to construct and stabilize networks binding together the entities of many different kinds, be they material, mechanical, textual, symbolic, human, legal, organizational, or whatever. Reality is thus composed of heterogeneous networks, hybrids or quasi-objects. But at this point we need to be careful. Despite its emphasis on networks built from heterogeneous materials, there are sides to ANT that would seem to suggest that it does not break with the representational mode, and that it retains a lot of SSK’s reductionist tendency. In support of such viewpoints, one can point towards ANT’s name for itself as “the sociology of translation.” The work that goes on in the laboratory, used broadly to denote places where science is conducted, is presented as understandable as (part of) a writing exercise. Science can be understood as rhetorical; it all boils down to persuasion (Latour and Woolgar, 1986; Latour, 1987; see below). The impression that ANT wants to reduce the heterogeneous reality of the laboratory into just one dimension – signs, writing, language, discourse – is further strengthened by its reliance on semiotic concepts and analyses (Latour, 1987; 1988a; 1992a). In this way, it would seem that ANT must be understood not as part of a materialist turn, but quite the contrary, as staying within the representational mode and understanding the world as text (Shapin, 1988; cf. Pickering, 1995: 12-13). What is the best interpretation? Does ANT breake with the representationalism and reductionism of SSK or does it not? I think it does. But to argue my case, I will have to go through a more detailed presentation of ANT’s notions of translation and representation.

The materiality of text ANT wants to study “Science in Action,” that is, the work, practices and procedures that actually go on in scientific institutions, research centers and laboratories. For such studies to be successful, timing is essential. Following SSK’s lead, ANT distinguishes between “ready-made science,” and “science-in-the making” (Latour, 1987: 2). Ready-made science, which is science as presented in text-books, when facts have been stabilized, and machines made to work, when claims of truth are no longer contested, is like a black box. It has been closed, and there is no way in. Sciencein-the-making, which denotes science at a stage when controversies are still open, when scientists

130

The Invisible Revolution

still debate what the facts are, how machines can be made to work, what the truth might be, is not like that at all. Here, the black box is still open, indeed, the question is whether there is a box there to be closed at all. What ANT wants to study, then, is how science-in-the-making is turned into ready-made science; how controversies are closed, agreement reached, boxes closed, statements turned from soft hypotheses to hard facts. ANT’s first step in doing this is to turn everything that happens in the laboratory into writing, the production of texts and literature. It is useful here to start with Latour’s notion of inscription. In Laboratory Life, Latour and Woolgar (1986) have their “estranged observer” note the importance of “inscription devices, “ that is, machines that can transform a material substance onto a figure or diagram. Not all the machines in the laboratory are of this kind. Many, for instance, evaporators, centrifuges, and grinders, are designed to transform matter from one state to another. However, the purpose of such transformations always is to prepare it for inscription devices: At this point, the observer felt that the laboratory was by no means quite as confusing as he had first thought. It seemed that there might be an essential similarity between the inscription capabilities of apparatus, the manic passion for marking, coding, and filing, and the literary skills of writing, persuasion and discussion. Thus, the observer could even make sense of such obscure activities as a technician grinding the brains of rats, by realising that the eventual end product of such activity might be a highly valued diagram. Even the most complicated jumble of figures might eventually end up as part of some argument between ‘doctors.’ For the observer, then, the laboratory began to take on the appearance of a system of literary inscription. (Latour and Woolgar, 1986: 51-52)

Inscriptions include such diverse things as graphic displays, laboratory notebooks, tables of data, reports, articles and books. And it is not only people who do the writing. The whole point of the many different apparatuses in the laboratory is to allow other entities – peptides, hormones, or whatever – to make themselves known by their inscriptions (Latour, 1987: 72). At this point, we can note that “translation” denotes the transformations between different types of inscriptions: “From marks to diagrams, from table to graph, from graph to statement, and from statement to statement - each is a translation” (Callon, 1994: 51). Such translation chains link the statements in a published paper via the stacks of diagrams and tables in the author’s office, through the laboratory notebooks and the graphs made by a diversity of inscription

131

Holm

devices, all the way back to the extracts and the entities who are thus allowed to make themselves known. Now, this still sounds a lot like a conventional picture of representation, with the knowing subject on the one hand and the still unknown objects on the other. But this is not what ANT is about. Using another example, the development of cartography from the 15th to the 18th century (Latour, 1987: Chapter 6), we notice the emphasis ANT puts on the practical skills, the instruments, the amount of work that goes into setting up translation chains. In this example, we are no longer confined to a laboratory in a literal sense. Nevertheless, cartography, much like the laboratory sciences, involves the construction of reliable and stable translation chains. So, the ships sent out from the cities of Europe were equipped with the best of contemporary scientific instruments and skills: Clocks to keep the time accurately so as to be able to measure longitude, compasses to measure latitude, and astronomers to tend the instruments and make the measurements. The first ships to go did not know where they were going and, if they came back, did not know where they had been. Since they brought back carefully reported measurements of longitudes and latitudes, drawings of coastlines, passages, etc, the next generation of ships had a little more to go on. After several such cycles, enough reports had been accumulated for the captain to know, before he had even left, where he was going. In the end, by collecting and comparing such reports, it was possible, in the Geographical Society in London, to simplify and concentrate this information in the form of an Atlas. The world can now be dominated, it can literally be held in one hand. So map making is one example of how science, by creating and stabilizing chains of translation, can gain leverage. Representation is not so much about knowing as about acting. Through reliable chains of translation, says Latour, science can act at a distance. How is this done? Answer: by somehow bringing home these events, places and people. How can this be achieved, since they are distant? By inventing means that (a) render them mobile so that they can be brought back; (b) keep them stable so that they can be moved back and forth without additional distortion, corruption or decay, and (c) are combinable so that whatever stuff they are made of, they can be cumulated, aggregated, or shuffled like a pack of cards. If those conditions are met, then a small provincial town, or an obscure laboratory, or a puny little company in a garage, that were first as weak as any other place will become centers dominating at a distance many other places. (Latour, 1987: 223)

132

The Invisible Revolution

Things (events, places, societies) can be brought home if they are translated into immutable mobiles (Latour, 1990). Here we are back at the point of departure and the importance of inscriptions. Instead of Shakalin Island, which is impossible to move, you bring back the trace of its coastline carefully recorded in a log-book; instead of the exploding star, you bring back its recording on graph paper. The translation of the object into inscriptions makes it at once more mobile and less vulnerable to corruption and change, and it can now be moved through time and space. At the same time, it means that the scale of the inscriptions can be modified at will, that they can easily be reproduced, combined with other inscriptions, and incorporated in a written text (Latour, 1990: 44-47). The centers of these translation networks, the places from which the delegates are sent and to which they return, Latour calls centers of calculation. Why calculation? Well, a key problem in such centers is the logistics of it all. The sheer number of reports and inscriptions that can be assembled in this way quickly threatens to fill up the center and turn it into an unmanageable heap of paper. To stay on top of the situation, a diversity of techniques – for instance tallies, totals, graphs, tables, lists, statistics – must be applied so as to allow the efficient mobilization of a large number of inscriptions and their fusion (Latour, 1987: 237-243). In particular, statistics, coefficients, formulas, formal theories, and the like, are simply techniques that allow “the acceleration of displacement without transformation” (Latour, 1990: 47). To ANT, representation in science concerns the work of building delegates, people or machines that can be sent out somewhere, collect samples or information, and then reliably return back home. The problems of representation are very often concerned with practical sides of constructing working metrologies, systems of measurement and reference, that will allow the centers to rely on what the delegates bring back. We note that this perspective is not compatible with one that sees representation as conceptual, as primarily linked with ideas, language, and discourse. On the contrary, when ANT talks about scientific essentials like theory, the emphasis is on how these things, together with statistics and mathematics, do the practical work of transporting, sorting and storing. For ANT a theory is in principle not much different from a filing cabinet (Latour, 1999a: 35) or a pedocompactor, a device for sorting and comparing soil samples (Latour, 1999a: 48).

133

Holm

If ANT is reductionist, it would seem more appropriate to charge it with the sin of reducing concepts, ideas, discourse, theory, and language to work, action, and materiality than vice versa. But this viewpoint is also problematic, since it starts from the premise of a fundamental divide between the world of things and the world of ideas. To ANT, there is no such divide. On the contrary, ANT must to a large extent be seen as a sustained effort to argue against this notion. We shall come back to this in a more systematic way below. Here, we can note that ANT’s chains of representation or translation are not designed to link together the two sides in the great Kantian divide between Nature, the world of objects, on the one hand, and Society, the world of knowing subjects, on the other hand. For ANT, rather, reality is in the middle, in the work of translation. This is not the middle as defined by the two extremes of objects and subjects. There are, says ANT, no final, underlying realities to which representations are linked. The chains of representation are not anchored; they have no limits at either end. Nevertheless, the outcome of the stabilization of such chains is the creation of a material world on the one hand and a formal world of language and meaning on the other. These, however, are not transcendental realities that were there from the beginning, waiting to be discovered, but the outcome of the work of translation. As opposed to reference as a bridge between objects and concepts, Latour emphasizes the circularity of reference. As in the cartography example mentioned above, the stabilization of translation chains, as summed up in the final map, allow delegates (ships) to go out and return home again. The chains of translation transport both ways between matter and form, between materiality and text, between locality and universality, between multiplicity and standardization: Phenomena, …, are not found in the meeting point between things and the forms of human mind; phenomena are what circulates all along the reversible chains of transformations, at each step losing some properties to gain others that render them compatible with already-established centers of calculations. Instead of growing from two fixed extremities toward a stable meeting point in the middle, the unstable reference grows from the middle towards the ends, which are continually pushed further away. (Latour, 1999a: 71-72)

Let us return to the question with which we started: Does ANT break with the representational mode? Yes, it does. While ANT suggests that science is a writing enterprise, this is only because for ANT writing is not primarily about concept, ideas, language and discourse. To the extent that

134

The Invisible Revolution

ANT sees science as text and discourse, it is because it insists on the materiality of scientific text and discourse (Schaanning, 1996). But more fundamentally, ANT refuses the dichotomy between the word and the world which representationalism presupposes. There simply is no gap for representations to bridge between objects and subjects.

The principle of extended symmetry As we recall from Chapter 3, SSK – rejecting Mertonian sociology of science – wanted to explain true and false knowledge claims in the same way. This is known as the principle of symmetry. But while SSK is perhaps symmetrical in some ways, it is heavily asymmetrical in other respects. Since it wants to substitute society for nature as the final arbiter of truth, it grants to social science what it denies natural science. Natural scientists are (and should be!) naïve realists, but they are wrong. Sociologists are social realists, and they are right. For ANT, SSK’s position is unacceptable, since it presupposes the fundamental Kantian dichotomy between objects and subjects, between Nature and Society. SSK substitutes the objectivism of the natural sciences with its mirror-image humanism. Together, then, the natural sciences and SSK form the bonded pair in a great Kantian pas-de-deux, forever circling each other. Neither side is correct. Instead of choosing one of them for the truth, says ANT, we should see them both as co-constituents of reality. That is, the distinctions between object and subject, things and society, world and word, and so forth, are the outcome when translation chains are stabilized. This means that science, which is fundamentally about creating and stabilizing such chains, cannot be explained by reference to these dichotomies. That would be to turn subject (what is to be explained) into resource (what is relied on to do the explaining). One cannot presuppose that which is to be explained. This leads to the principle of generalized symmetry, abolishing not only different repertoires in the explanation of true and false beliefs, but also such differences when analyzing the behavior of entities that seem to belong to Nature and Society respectively. In Callon’s (1986) version, the tenet of generalized symmetry comprises three related methodological principles. The first is agnosticism: When studying science, typically in the context of some controversy, the observer should not only, as SSK recommends, be impartial with regards to the scientific and

135

Holm

technical viewpoints of the different sides. In addition, he must also abstain “from censoring the actors when they speak about themselves or the social environment. (…) No point of view is privileged and no interpretation is censored. The observer does not fix the identity of the implicated actors if this identity is still being negotiated ” (Callon, 1986: 200). The second methodological principle, which Callon calls generalized symmetry, goes like this: “… the rule which we must respect is not to change [explanatory] registers when we move from the technical to the social aspects of the problem studied” (ibid). The third of Callon’s methodological principles is free association: “The observer must abandon all a priori distinctions between natural and social events. He must reject the hypothesis of a definite boundary which separates the two” (ibid: 200-201). We see that the three principles overlap and reinforce each other and that they are all related to rejection of an a priori distinction between Nature and Society, whether this is applied to structure the respective identities of the actors being studied and those doing the studying (first principle), the explanatory repertoire deemed relevant (second principle), or basic ontological distinctions (third principle). This perhaps sounds, if not reasonable, at least like a logical extension of the rejection of the Kantian divide. Nevertheless, following the principle of generalized symmetry quickly leads into highly counter-intuitive and controversial positions. One of the most controversial ANT viewpoint is its post-humanism: That the distinction between human and non-human actors must be abandoned; that things and machines are equally important and should be treated as no different from human actors. We have already seen an example of how this position may express itself in practice. In the analysis of fisheries resource management in Paper 7, as summarized at the beginning of this chapter, the fish was systematically brought into the text in the same way as the fishers. The paper even talked about the possibility of negotiating with the fish directly. In much the same way, Callon (1986) treats scallops and fishers symmetrically, and Latour writes about the sociology of a door (1992b) and a hotel key (1991). I shall return to the various objections and criticisms that have been leveled against ANT on this point in a little while. First, however, let us have a closer look at the way ANT proposes to conduct symmetrical analyses. In order to get a handle on entities regardless of the identities that are ascribed to them within stabilized networks, ANT introduces the concept of “actant.” Imported from semiotics, this concept denotes any character, be it human, machine, natural or artefactual, or symbolic, that has

136

The Invisible Revolution

a role to play in a story. So, for instance, in the folk-tale “The princess who always had to have the last word” (Asbjørnsen and Moe, 1992), Per, Pål and Espen Askeladd are actants, but so are the strange objects Espen Askeladd picks up along the road as they are travelling to the king’s manor. The principle of free association means that we cannot know the identity of these actants until the story has unfolded. What are the dead magpie, the old willow hank, the broken saucer, the pair of crooked ram’s horns, a worn-out shoe sole? “Trash,” say Per and Pål, “throw them away.” But their true identity, like the identity of Espen Askeladd himself, is not revealed until they turn out to be exactly what is needed to make the princess hold her tongue. So, rather than predetermining the answer, like Per and Pål – who get thrown out and branded on both ears with the king’s big branding iron – we should follow Espen Askeladd and allow them to reveal their identities as the story unfolds. How do actants reveal their identities? By what they can do, says ANT. In the folk tale referred to above, the identities of Per, Pål and Espen Askeladd only became visible in the confrontation with the Princess. This is exactly how actants in science and technology stories are given identities: “An actant is a list of answers to trials – a list which, once stabilized, is hooked to a name of a thing and to a substance. This substance acts as a subject to all the predicates – in other words, it is made the origin of actions…” (Latour, 1991: 122). But of course, in science and technology, the trials are of a slightly different kind than those we meet in the folk tale, although the kingdom, if not the princess might be at stake here too: We do not know beforehand what an agent is doing. We must try it out. This one corrupts veal stock; that one transforms sugar into alcohol; the other one survives in gelatin but is interrupted in urine. How are we to define a shape? Like all the others: they are the edge of trials of strength that others subject them to. If we boil water five degrees more, a new species is then defined, whose ‘edge’ is to resist the temperature of 100°. If we deprive it of oxygen, then others are defined that do not need air. (Latour, 1988b: 79)

Within semiotics, actants are acting entities defined by their relations to other entities (Åslund, 1994: 141). This is also the case within ANT. The identities of the actants are not given in the order of things, but are relational effects (Callon and Law, 1997: 171). In other words, it is only when actants find their place within networks of relations - networks that are established and tested by trials of strength – that they take on specific shapes and properties.

137

Holm

This brings us from the actants that make up networks to the networks themselves. While ANT pays a great deal of attention to actants, there should be no confusion about the fact that the actor-network is at the center of the theory. We can here note that ANT stands for actornetwork theory, not actant-network theory. Actants are, so to speak, the (unknowable) raw material from which networks are constructed. An actor is always a network (Callon and Law, 1995: 490). This is often difficult to keep in mind, however, because, when networks are stabilized, they tend to fold in on themselves and present themselves as single entities or black boxes. The entities or black boxes come to represent the networks they stand for in a simple and coherent form. In this way, although actors are networks, they are also points (Callon, 1991; Callon and Law, 1995; 1997). To illustrate the point, Callon and Law (1995) give the example of “Andrew,” the powerful managing director of a large laboratory. He is an “agent,” a strategist and decisionmaker: Now let’s try a thought experiment. Just imagine what would happen if they took away Andrew’s telephone and his fax machine. If they blocked the flow of papers and reports. Imagine what would happen if they shut down the railway line to London and stopped him from using his car, so he couldn’t get to the smoke-filled rooms. Then imagine, also, that his secretary were to disappear. And his room, with its conference table, its PC and electronic mail, were to vanish. Imagine what would happen if other people in the laboratory started to ignore him. Or to treat him as if he were the messenger bringing ’round their letters and paychecks. Call this the baboon experiment. An experiment where, as in a Paul Auster novel, the materials of sociality are progressively withdrawn until the person is sleeping rough in Central Park. Under such a regimen of deprivation, would Andrew still be an agent? (Callon and Law, 1995: 484)

Even if Callon and Law immediately say that this is an unanswerable question, they maintain that Andrew’s being an agent or actor is an emergent effect, a particular arrangement of bits and pieces. Take these bits and pieces – “the materials of sociality” – away, and Andrew stops being an actor, at least the kind of actor he used to be. Accepting as a fact the impression created by the punctualization or black-boxing of the network that makes up Andrew-the-network makes invisible all the work that allows Andrewthe-point to be an effective agent. As feminists have pointed out, this perspective tends to privilege male executives, while it discredits all the work done by female laborers, wives, and

138

The Invisible Revolution

secretaries (Star, 1991). ANT extends this point to cover all the actants that make up networks, including things, machines, and texts (cf. Latour, 1992b). In contrast to feminist theory, however, where this is primarily a moral and political issue, it is a methodological issue within ANT. ANT certainly does not want, at the outset, to extend the same type of moral status to things and machines as are routinely granted to (some) human actors. Such statuses remain emergent network effects. The principle of generalized symmetry urges the analyst to abstain from prejudging the entities involved.

Hylozoism An oft-repeated reason to reject ANT is that it illegitimately ascribes capacities that only humans can have to non-humans. This has been labelled the “heresy of hylozoism,” defined as “an attribution of purpose, will and life to inanimate matter, and of human interests to the non-human” (Schaffer, 1991: 182).6 This objection to ANT has been repeated by several of the key representatives of SSK (Collins and Yearly, 1992; Gingras, 1995; Bloor, 1999a; 1999b) and others (Amsterdamska, 1990). It has also been raised in much the same form against others who have introduced non-material agency (eg. Pickering, 1995; see Gingras, 1997). Why is hylozoism, in this view, a heresy, and what has ANT had to say in its defence (Callon and Latour, 1992; Latour, 1992c; 1999b; Pickering, 1999)? The question under dispute here concerns the correct way to permit non-human entities into the analysis. As the charge of hylozoism suggests, there is considerable difference between SSK and ANT in this respect. Where SSK’s focus remains overwhelmingly on the human side, ANT thinks that also non-human actors will make a difference. Phrased in these terms, this may perhaps look like a question of emphasis, and not much to get exited about. Yet, as we shall see, the annoyance provoked by it is considerable. In order to understand this, we must place the question of hylozoism in its proper context, as a focal point in a more general quarrel between SSK and ANT. While these two varieties of science studies have many things in common, they

6

This is Encyclopedia Britannica’s entry on hylozoism: “(from Greek hyle, ‘matter’; zoe, ‘life’), in philosophy, any system that views all matter as alive, either in itself or by participation in the operation of a world soul or some similar principle. Hylozoism is logically distinct both from early forms of animism, which personify nature, and from panpsychism, which attributes some form of consciousness or sensation to all matter.” 139

Holm

are also sharply opposed in some respects. Depending on the issue at stake, the emphasis will vary between commonality and conflict. Exchanges on symmetry and material agency tend to bring out the conflicts. Let us see why. For ANT, SSK is the in-house representative of a dualist type of sociology that it wants to replace. ANT thus tends to read SSK as imposing an extreme asymmetry between objects (Nature) and subjects (Society). It resides at the Society pole of the Kantian divide in which “objects count for nothing; they are just there to be used as the white screen on to which society projects its cinema” (Latour, 1993: 53). SSK, according to ANT, wants to explain nature in terms of society (see Callon and Latour, 1992; Latour, 1999a; see also Bloor’s [1999a: 87-87], rendering of ANT’s criticism of SSK, which is a good summary). SSK’s representatives do not accept this diagnosis. SSK does not want to explain nature with reference to society; the object is not nature itself, but beliefs about nature. It is as if Latour and ANT have difficulty telling apart nature and accounts of nature, a distinction that is vital to uphold (Bloor, 1999a: 87). SSK wants to be naturalistic and pursues causal explanations. Far from reducing one pole to the other, SSK strives to take account of natural as well as social causes of belief. Both non-social nature, the world of things, and society, the world of people, will be implicated in the formation of belief. ANT’s charge that SSK pictures science as responding only to society, and not at all to (non-social) nature, is wrong: “The working assumption (…) is that scientists are always responding to nature, but doing so collectively through their shared conventions and institutionalized concepts” (Bloor, 1999a: 90). Indeed, the complexity and richness of sensory inputs are important points for SSK: Nature will always have to be filtered, simplified, selectively sampled, and cleverly interpreted to bring it within our grasp. It is because complexity must always be reduced to relative simplicity that different ways of representing nature are always possible. How we simplify it, how we choose to make approximations and selections, is not dictated by (non-social) nature itself. These processes, which are collective achievements, must ultimately be referred to properties of the knowing subject. (Blorr, 1999: 90)

The object, says Bloor, is there and has a role to play, but, since different interpretations of it are always possible, the object itself “drops out of the story because it is a common factor behind … different responses, and it is the cause of the difference that interests us” (ibid: 93). Latour’s

140

The Invisible Revolution

picture of SSK, says Bloor in conclusion, is a false stereotype, portraying SSK as idealist rather than what is really is, a species of materialism (ibid: 110). The SSK reading of ANT (Shapin, 1988; Schaffer, 1991; Sturdy, 1991, Collins and Yearly, 1992; Gingras, 1995; Bloor, 1999a: 95) ends with an equally devastating criticism as ANT’s reading of SSK. In some of the attacks on ANT (in particular Collins and Yearly, 1992; Amsterdamska, 1990; Kenshur, 1996), there is not much appreciation of a key feature of ANT, namely its rejection as explanatory resources of the dichotomies between accounts of objects and the objects themselves. Since ANT wants to explore such dichotomies as the result of scientific settlements, they cannot be relied on as the cause of these settlements. Without this firmly in place, ANT indeed appears as a strange exercise, a bizarre joke (Amsterdamska, 1990). In contrast to these critics, Bloor (1999a) clearly acknowledges ANT’s own starting point (ibid: 87, 97). Despite this, however, Bloor quickly comes to the conclusion that ANT fails miserably: We are told to encourage the new perspective by deliberately inverting our usual conceptual conventions, using a purposive vocabulary for things which don’t have purposes, and a mechanistic vocabulary for things that do. We must try to think of Pasteur as if he was a microbe, and microbes as if they were like Pasteur, or treat Millikan as if he were an electron and electrons as if they were like Millikan. But unless we are very confident indeed that the exercise is necessary and justified, this looks like a formula for imposing confusion on ourselves: it is obscurantism raised to the level of a general methodological principle. (Bloor, 1999a: 97)

The vocabulary that ANT proposes, featuring strange things like monads, quasi-objects, actants, etc, is at a metaphysical level and it remains unclear how it connects to everyday reality. In Bloor’s reading, Latour’s attempts at explaining his terminology and methodology do not succeed; most of it remains incomprehensible and baffling. To the extent that there are things that can be grasped amid all the confusion, it would seem to presuppose and refer back to the dichotomies it was designed to overcome: “Latour’s attempt to get to metaphysical bedrock doesn’t work: he can’t get away from a pre-existing nature and a pre-existing society” (ibid: 98). While SSK representatives sometimes admit that ANT has learnt a lot from SSK and also has a lot to bring back to SSK (Schaffer, 1991; Shapin, 1988), the central features of ANT, such as the radical symmetry principle, the granting of agency to non-humans, and the co-production of nature and society, remain deeply obscure and deficient (Bloor, 1999a: 93). 141

Holm

What has ANT to say in its defence? Well, ANT has in some way an easy task, since SSK does not claim – as ANT does with respect to SSK – to have understood and diagnosed it successfully. Even Bloor, who is the one who goes furthest in an attempt to understand ANT on its own premises, admits bafflement and incomprehension. All ANT has to do is to try once more to explain its position. Since such explanation is the overall purpose of this chapter, I will not go into a summary of these attempts; that would only reiterate points that have already been made or will be made later. Instead, let us leave the metaphysical level, and explore what all this means in terms of practical analysis. What difference does it make whether one relies on ANT or SSK in concrete analyses of science in practice? As already indicated in the above rendering of the SSK-ANT debate, there are important differences between the two sides in terms of what type of factors they would look for in accounting for scientific settlements. For SSK, objects are never enough; they drop out of the picture because they can always be interpreted differently. For ANT, in contrast, the objects “themselves” are allowed to make a difference. This clearly comes out when SSK and ANT disagree in their interpretation of specific cases of scientific practice. Our illustration here is brought up by Schaffer (1991) in his review essay about Latour’s book on Louis Pasteur (Latour, 1988b). The issue at stake is how to account for Pasteur’s work on Bacillus anthracis, and the relationship between his work and that of his competitors, including Koch and Pouchet. To Schaffer, the main problem of Latour’s account is that it loads way too much responsibility on the anthrax bacillus itself: … Latour brilliantly gestures at the skills which were needed to make the candidate Bacillus anthracis count as the cause of the disease anthrax, that is, to constitute the microbe. It was necessary to show that the material which lab skills could master was indeed connected aetiologically with the problems farmers faced. But then he stops. Instead of an analysis of the controversies through which these skills were forged, he baldly states that the microbes ‘enthusiastically’ behaved the way Pasteur wanted. Hylozoism stifles an account of laboratory life. (Schaffer, 1991: 187)

For Schaffer, this is the heresy of hylozoism. Instead of talk about the anthrax bacillus itself, he wants to talk about different interpretations of the agent that was causing anthrax, how it could be purified in a lab setting, the development of skills and techniques to isolate it, etc: “The microbes by no means ‘enthusiastically’ obeyed the Pastorians in preference to Koch. Indeed, we

142

The Invisible Revolution

have no basis for speaking of their wishes if we are to obey the dictates of symmetry: their reported behaviours varied between the two places” (ibid). So we cannot rely on the objects in question to decide the outcome of a dispute since it is precisely the identity of these objects that is under dispute: “Hylozoism directs our attention towards the items whose action is in dispute. It directs our attention away from the forces which help close that dispute” (ibid: 189) When Pasteur won over his competitors, the microbes did not have anything to do with it: “… he won, and it was, as ever, his human colleagues who were crucial” (ibid). In response, Latour accuses Schaffer of ignoring the microbes and only focusing on the humans: The critiques made by Schaffer (1991) of my position are exactly the same as those of Bloor, and exactly as ill-conceived. Schaffer wants to share unequally the sources of uncertainties, so that all the uncertainties reside with humans, while the sensory inputs remain utterly neutral. (Latour, 1999a: 117 n. 6)

More illustrations like this could have been given. We could have taken a look at the way SSK and ANT differ in the interpretation of the Lysenko affair (Bloor, 1999a: 88-90). We could have explored how the two sides view the relationship between scallops, fishers, and scientists in St. Brieuc Bay (Callon, 1986; Callon and Latour, 1992; Collins and Yearly, 1992), or their quarrel about the correct way to bring doors, seatbelts and sleeping policemen into social analysis (Latour, 1992b; Collins and Yearly, 1992). We could have looked into the differences between Bloor’s and Pickering’s analyses of the mathematics of Sir William Rowan Hamilton (Bloor, 1981; Pikering, 1995: 147-156). Or we could have returned to our discussion in the previous chapter of how to account for Norwegian coastal cod (Paper 5) or the northern cod crisis in Canada (Finlayson, 1994). The conclusion would be the same: Where SSK focuses on what goes on between the humans involved, ANT also wants to allow objects to make a difference. This contrast, however, should not be exaggerated. SSK does not insist that objects do not play any role whatsoever. And ANT of course would not say that objects have all the say and that controversy between scientists means nothing. There is nevertheless a clear difference in emphasis here: SSK looks towards people while ANT also looks towards things. SSK’s symmetry principle insists that all humans, whether they are right or wrong, be treated equally. ANT’s symmetry principle insists that all actants, human and non-human, be treated equally.

143

Holm

How facts escape from laboratories According to ANT, scientific facts are created in laboratories. They do not simply, as the realists will have it, exist beforehand, just waiting to be discovered. Instead, facts are points on stabilized networks or translation-chains, which are constructed within the laboratory. But how can ANT then explain how facts and artifacts escape from laboratories? How can they survive and multiply in the outside world? These are important questions for ANT to answer, since the leap out into the real world seems to vindicate the realist claim that facts simply correspond to eternal nature. According to this interpretation, objects and relationships, as “discovered” in the micro-cosmos of the laboratory, reflect universal natural laws. Once they are identified correctly, they can therefore be put to use in the macro-cosmos of the outside world: “Universal gravitation appears everywhere; we are convinced of it. Boyle’s laws, Mariotte’s laws, Planck’s constants legislate everywhere and are constant everywhere” (Latour, 1993: 118). Now, SSK gets around this problem by focusing on the accounts of objects and relationships instead of the objects and the relationships themselves. Because of the principles of interpretive flexibility, i.e. that phenomena are always open to many different interpretations, it is no real mystery that science works. One can always imagine that scientific discoveries could have taken a different path, leading to different accounts being accepted as facts, and they would have worked equally well. In this perspective, then, the question of how facts escape from laboratories becomes a question of why a specific interpretation of a phenomenon, as established within small scientific communities, comes to be accepted as true in the society outside. In the interest model of the Edinburgh school, this is no mystery, since it wants to explain scientific knowledge with reference to the scientists’ location in society, i.e. by their interests. The causal arrow goes from macro-society into the laboratory. The knowledge established within the laboratory will then have no problem in being accepted on the outside since this is where it came from in the first place. Now, ANT’s answer to the question of how facts escape from laboratories differs from both realist and SSK accounts. In contrast to the realists, ANT does not believe in universal laws and eternal facts since such things are wholly created in laboratories. For Latour (1987; Chapter 2) this is a key point, as underlined by his third “rule of method:” Since the settlement of a controversy is the cause of Nature’s representation not the consequence, we can never use the outcome – Nature – to explain how and why a controversy has been settled. (Latour, 1987: 99)

144

The Invisible Revolution

In opposition to conventional realist and rationalist accounts, ANT denies the possibility of referring to objects-in-themselves in order to account for scientific settlements. In contrast to the social constructivists, ANT refuses the a priori distinctions between outside and inside, macro and micro, context and content, social and technical, so that the former (outside, macro, context, social) can be cast as explanatory resources and the latter (inside, micro, content, technical) as the topics to be explained. This type of explanation cannot be accepted, since it assumes what should be explained. The differentials between outside and inside, macro and micro, nature and society, etc, are for ANT the outcome of scientific settlements, and can therefore not be used in order to explain such settlements. For ANT, the starting point here is to challenge the dichotomies that lie at the bottom of both realist and social constructivist perspectives. Laboratories do not have outsides, and facts never leave them. Instead, when scientific facts are made to work in other places than the laboratory in which they originated, it is because these places also have been turned into laboratories. In other words, the network in which the fact forms a point has been extended so that it can be made to survive elsewhere: “… we might compare scientific facts to frozen fish: the cold chain that keeps them fresh must not be interrupted, however briefly” (Latour, 1993: 119). As an illustration of how translation chains are extended and stabilized, take Latour’s analysis of Pasteur’s development of a vaccine against anthrax, which was prevalent among cattle at the end of the 19th century (see Latour, 1988b; Latour, 1983: 143-153; also Koch, 1995: 324-5). Latour describes this process as divided into three related moves (Latour, 1988b: 7593). The first move was to set up a field laboratory in the infected areas, and thus learn from the disease. Here, Pasteur and his colleagues translated what the farmers and veterinarians knew about the disease, its idiosyncrasies and variations, into their own laboratory language. In this way, events in the field could be translated into events in the laboratory. This prepared the way for the second move, in which the field laboratory and the anthrax bacillus extracted from the sick cattle could be moved back to Paris and the main laboratory. While the disease in the field, invisible and unpredictable, was stronger than the cattle, the farmers, the veterinarians, and of course, also Pasteur himself, Pasteur became stronger than the bacillus in the laboratory. Here he could grow it in such large quantities that it became visible. He could infect animals of his choosing

145

Holm

with the microbes, and thus learn to control the outbreak of the disease. He could eventually also, by first injecting strands of weakened microbes, make cattle resistant against strands of intact microbes. In the third move, Pasteur was then able to transfer his vaccine from the confines of the laboratory back to French farmland. A crucial event for this to happen was a field trial, conducted at Pouilly le Fort, in which sheep that had received Pasteur’s vaccine remained healthy, while those in the control groups all died of the disease. This succeeds only to the extent that the laboratory practices – disinfection, cleanliness, timing and recording, in short, the procedures that are essential for the vaccine to work – are also observed when the vaccine is used on the farms: … the vaccine works at Pouilly le Fort and then in other places only if in all these places the same laboratory conditions are extended there beforehand. Scientific facts are like trains, they do not work off their rails. You can extend the rails and connect them but you cannot drive a locomotive through a field. The best proof of this is that every time the method of extension of the anthrax vaccine was modified, the vaccine did not work and Pasteur got bogged down in bitter controversy... His answer was always to check and see if everything was done according to the prescriptions of his lab. (Latour, 1983: 155).

Scientific facts cannot escape the laboratory, but this does not matter when the world can be turned into a laboratory. Explained like this, ANT’s answer to the question of how facts escape laboratories sounds acceptable, perhaps even reasonable. It is nevertheless important to acknowledge how easy it is to allow the distinction between objects and accounts of objects to slip back into place. Once it does, ANT’s position becomes problematic. One way to bring this out is to ask not how facts can move through space, as has been the focus of our discussion until now, but how they can move through time. In other words, do things have histories? Were there microbes before Pasteur? It follows from the third rule of method that ANT’s answers must be: “Yes, things do have histories,” and “No, there were no microbes before Pasteur ‘discovered‘ them in his laboratory” (Latour, 1999a: Chapter 5). The same type of answer is given by Latour when he discusses the claim made by French scientists that the Pharaoh Ramses II died (around 1213 BC) of tuberculosis. “How could he pass away due to a bacillus discovered by Robert Koch in 1882?” (Latour 1998, cited from Sokal and Brichmont, 1998: 96-97, n. 123). According to Latour, the claim that Ramses II died of tuberculosis is just as anachronistic as saying that he was

146

The Invisible Revolution

killed by machine gun fire or died from the stress of a stock-market crash. Like these things, the tuberculosis bacillus came into existence long after Ramses II had passed away: “Before Koch, the bacillus has no real existence” (ibid: 97). Sokal and Brichmont, sticking to a realist position, insist that when science has got it right, the facts are universally valid outside time and space. While the discovery of a fact has a history, the fact itself does not. If Koch were right that the tuberculosis bacillus is the cause of tuberculosis – and of course we know that without a shred of doubt – it is not in principle problematic to hypothesize that this bacillus killed Ramses II. To Sokal and Brichmont – and it is easy to see their point – Latour’s claims are preposterous: … unless Latour is putting forward the truly radical claim that nothing we discover ever existed prior to its ‘discovery’ – in particular, that no murderer is a murderer, in the sense that he committed a crime before the police ’discovered’ him to be a murderer – he needs to explain what is special about bacilli, and this he has utterly failed to do. (Sokal and Brichmont, 1998: 97)

While Sokal and Brichmont cannot bring themselves to believe that Latour is putting forward the “truly radical” claim that objects postulated by science do not exist before their discovery, this is exactly what he does. He is therefore not committed to explaining what is special about bacilli. But this puts Latour’s position on a par with that indicated by the murder example, which is, I suppose, meant to bring the debate to an abrupt and absurd end. However, the murder case is one that is very easy to reply to. An act of killing someone is not a “murder” until it has been so classified in a court of law. There are plenty of other possible ways of classifying an act of killing someone; some of them do not even constitute crimes, for instance, if done in self defence or as part of a war. It is also quite clear that what constitutes murder and its various sub-species varies over time and between societies. Murder is a social institution; social institutions have histories. The example thus does not serve Sokal and Brichmont’s purpose well: While it is clearly counterintuitive to accept that things have histories, the idea that murders and murderers have histories is quite obvious. We must not make this too easy for ourselves, however. The murder example would seem to suggest that the reason why ANT insists that things have histories is that we are talking about classification of things instead of the things themselves. When we can agree that “murders” have histories, then, this may be because their historicity refers to the classification of certain 147

Holm

types of acts rather than to the acts themselves. And of course, classification belongs to language and society; to the half of the Kantian divide that does have history. But this is a position that belongs to SSK, not ANT. Let us therefore take a closer look at how SSK will answer the question of the historicity of things. In order to answer this, we must remember that SSK starts by postulating a divide between the knower and the known, between accounts of objects and the objects themselves. What SSK wants is to explore scientific knowledge. While the reality of the objects that are postulated by this knowledge is not disputed, this reality does not make much of a difference. It is there, but does not act as a referee, deciding between different knowledge claims about it. There is one real world, but it lends itself to all sorts of classifications. This would imply that SSK answers both “yes” and “no” to our question. On the one hand, the things themselves have no history; they are forever the same. But this does not really matter since they interfere so little in human affairs. On the other hand, knowledge of things, how they are classified and thought of, has a history. In this sense, SSK would agree with Latour that Ramses II did not die of tuberculosis. In this interpretation, however, the point would not be that the tuberculosis bacillus itself did not exist at the time of the pharaohs. SSK has no opinion on such claims. Koch’s interpretation, in which the existence of such a creature is postulated, is but one of an endless number of possible interpretations. Perhaps Koch got it right, but there is no way to tell. Instead, the salient point is that the illness we call tuberculosis – which is today held to be caused by a bacillus the discovery of which is attributed to Koch – was not recognized as such by Ramses II’s contemporaries. Instead they would perhaps believe – but I must admit that my Egyptology is a bit shaky here – that Ramses II had simply decided that his period as a god embodied in human form had come to an end. SSK would here be relativist, reporting, but not ranking on a truth scale, the respective cultural inventories of causes of death in two different societies. To the question “Do things have histories?” then, ANT answers differently from both realists like Sokal and Brichmont and social constructivists like Bloor and Collins. In contrast to the former, who say that objects, once they have been discovered, have always existed, ANT insists that such objects have a history. In contrast to the latter, who will say that scientific knowledge claims have a history, but plead ignorance and downgrade the significance of objects

148

The Invisible Revolution

postulated by such claims, ANT refuses to accept the distinction between objects and knowledge of objects and insists on the historicity of the whole object/knowledge package. As this suggests, the solution to the mystery of how things have histories follows from ANT’s refusal to distinguish between ontological and epistemological questions (Latour, 1999a: 145-146). Instead of the Kantian starting point of a gulf between the knower and the known to be bridged by representations, ANT wants to start in the middle, in the potentially endless chain or network of translations to which an entity is linked: To define an entity, one will not look for an essence, or for a correspondence with a state of affairs, but for the list of all the syntagms or associations into which one element enters. This nonessentialist definition will allow for a considerable range of variations, just as a word is defined by the list of its usages: ‘air’ will be different when associated with ‘Rouen’ and ‘spontaneous generation’ than when associated with ‘rue d’Ulm,’ ‘swan-neck experiment,’ and ‘germs;’ it will mean ‘transport of life-forces’ in one case and ‘transport of oxygen and transport of dust-carrying germs’ in the other; but the Emperor will also be different when associated by Pouchet with ‘ideological support of spontaneous generation to maintain God’s creative power’ and by Pasteur with ‘monetary support of laboratories without any implication about the subject matters of science.’ What is the essence of air? All of these associations. Who is the Emperor? All of these associations. (Latour, 1999a: 161)

Starting from Latour’s third rule of method, accepting that entities, or actants, are defined by the translation chains to which they are connected, it follows that things have histories: They cannot take on existence independent of the networks within which they are defined. When Ramses II is claimed to have died of tuberculosis, this simply means that another actant (Ramses II) has been tied into the network in which the tuberculoses bacillus lives, redefining both the pharaoh and the microbe in the process. Ramses II becomes “he who can be killed by tuberculosis,” making his claims to divinity and sacred powers even less convincing. The tuberculosis bacillus gains another edge, adding “he who can kill pharaohs” to an already extensive Curriculum Vitae.

Sociologies ANT rejects all a priori distinctions in order to account for scientific settlements. Just as there is no transcendental Nature that can be the cause of such settlements, there is no transcendental 149

Holm

Society that can do so either. This is summarized by Latour in his fourth “Rule of method,” which is the same as the third, but focuses on Society instead of Nature: “Since the settlement of a controversy is the cause of a Society’s stability, we cannot use Society to explain how and why a controversy has been settled. We should consider symmetrically the efforts to enrol human and non-human resources” (Latour, 1987: 258). So, the question posed above, i.e. “How do facts escape from laboratories?” is all wrong. Laboratories have no outside, and nothing escapes from them. This is reflected in the title of Latour’s (1988b) book on Pasteur: The Pasteurization of France. Pasteur’s work in the laboratory not only produced new scientific facts, but also, in the same moment, changed French society. France was a significantly different place after Pasteur had introduced into it a crowd of invisible, dangerous creatures that only he was able to control: Everywhere Pasteurian laboratories were established as the only agency able to kill the dangerous actors that were until then perverting efforts to make beer, vinegar, to perform surgery, to give birth, to milk a cow, to keep a regiment healthy, and so on. It would be a weak conception of sociology if the reader were only to say that microbiology ‘has an influence’ or ‘is influenced by the nineteenth-century social context’. Microbiology laboratories are one of the few places where the very composition of the social context has been metamorphosed. It is no small endeavor to transform society so as to include microbes and microbe-watchers in its very fabric. (Latour, 1983: 158)

So ANT turns around the causal arrow posed by SSK. The direction is not from society out there to the laboratory in here, but the other way around, from the laboratory to society. Or rather, to escape the a priori distinction between outside and inside that still resides in this formulation, we should perhaps say that society, just like nature, is manufactured within laboratories. It is perhaps not perfectly clear what the consequences are of such a viewpoint. Let us therefore look into a concrete example, namely Callon’s (1987) analysis of the attempt by a group of French engineers to launch a major innovation, the introduction of an electric car in the early 1970s. In addition to the scientific and technical foundation of such a car, the accumulators and fuel cells that would be required, its precise technical design, etc, the engineers offered a social analysis, indicating the structure of the social universe in which the electric car would fit. Here, they announced the coming of post-industrial society, in which the conventional motorcar, based on the internal combustion engine was about to end up in a highly vulnerable position. In

150

The Invisible Revolution

addition to being responsible for noise and air pollution in the big cities, it constituted a primordial element of status, and was thus tightly linked to the consumer society. The electric car would solve the noise and pollution problem. Since it would also be slower and have a more restricted operational range, it would reduce the car to what it should be, a simple, useful object: The electric car could lead to a new era in public transport in the hands of new social groups that were struggling to improve conditions in the city by means of science and technology. The goal would be to put science and technology at the service of the user and to do away with social categories that attempted to distinguish themselves by their styles of consumption. (Callon, 1987: 85)

Here Callon points to the similarity of the sociological analysis proposed by the engineers and that proposed by the respected French sociologist Alain Touraine. At the heart of his sociology is the proposition of a change of the dominant class conflict in modern society, from that between the working class and the bourgeoisie, towards one where the large corporations stand one side and consumers on the other. The growing importance of this new class conflict is what Touraine identifies with post-industrial society. Now, this is exactly the analysis offered by the engineers promoting the electric car. The conventional motorcar is an integral part of industrial society; it stands as a symbol and a cornerstone of that system. The electric car, on the other hand, represents post-industrial society, in which technocrats and consumers form an alliance in direct confrontation with the large corporations. While the engineers, working for EDF (Electricité de France), initially were virtually unopposed, they soon encountered resistance. The toughest challenge came from a group of engineers at Renault, one of the large corporations whose future was highly problematical in the social universe predicted by the EDF engineers. Instead of the sociology of Alain Touraine, they invoked the sociology of Pierre Bourdieu, in which the conventional automobile is so socially embedded that it cannot simply be replaced with a radically different means of transport. The only realistic strategy is gradual improvement and change. This goes together with a rejection of the simple class analysis of Touraine. Instead of the picture of one major class conflict (working classes versus bourgeoisie, large corporations versus consumers), confrontation is fragmented between a number of different fields. The Renault engineers, along with Bourdieu, thus rejected both the coming of post-industrial society and the coming of the electric car.

151

Holm

The question now becomes: Who were right? Or, who were the better sociologists – the EDF engineers, or the Renault engineers? In answering this question, Callon points out that the sociologist-engineers did not have to defend their analysis in the academic arena, like Touraine and Bourdieu. So the brilliance or originality of the analysis was unimportant. What mattered was to be right, that is, for the EDF engineers to prove that French society was evolving in the way they claimed it was, by bringing the electric car from the drawing board and making it a success in the real world. “In short, if an engineer-sociologist is to be proved right he or she has to create a new market; success is measured by the amount of profit gained. This, in all its simplicity and toughness, is the test of truth” (Callon, 1987: 90). As it turned out, it was the sociology of the Renault engineers, and hence that of Bourdieu, that withstood this test of truth. An important reason for this, according to Callon, was that the new cheap catalysts that formed an essential component of the new car turned out to have the unfortunate tendency of quickly becoming contaminated, rendering the fuel cells unusable. The fuel cells were unable to enact their role, and the whole network on which the EDF engineers had relied became unstable. In the end, then, the Renault engineers, in alliance with the contaminating catalysts, managed to rehabilitate the traditional motorcar, although the motorcar underwent some subtle changes in the process – it polluted less, used less petrol, and cost less to manufacture (Callon, 1987: 91). While Bourdieu’s sociology turned out to be successful in this case and Touraine’s sociology failed, Callon is quick to add, that this is “pure luck” (Callon, 1987: 97). This is because Bourdieu, like Touraine, systematically excludes from his analysis a large number of the entities – electrons, catalysts, fuel-cells – that make up the network of which society consists. The engineer-sociologists, in contrast, do much better since they include in their sociology all the relevant elements, be they human, political, symbolic, technical or natural. Callon’s analysis of the engineer-sociologists seems curiously asymmetrical. If we follow the principle of generalized symmetry, and allow natural scientists and engineers to be sociologists, do we not at the same have to allow social scientists to be engineers and naturalists (Callon, 1986)? ANT’s treatment of natural and social science is often biased, taking an admiring tone towards natural scientists, particularly when they talk about social and political stuff, but not mincing words as to the incompetence of social scientists. Callon’s comments on Touraine and

152

The Invisible Revolution

Bourdieu are a case in point. Their theories are derided as hypothetical and speculative. In contrast to the engineer-sociologists, the professional sociologists remain in the relatively undemanding world of academia (Callon, 1987: 88). My own treatment of fisheries resource management is another case in point (Paper 7). While the biologists are hailed as fully-fledged network constructors, the sociologists are criticized for only wanting to build with soft materials, i.e. social relations and discourse. Is this really fair, or is there an asymmetrical pre-judgement here, an a priory censoring of social scientists on ANT’s part? I think the answer is that ANT vacillates between two different perspectives on the social sciences. On the one hand, sociology is cast as a competing perspective against which ANT defines itself and which it wants to replace: “Cast aside your sociology books and follow me instead,” says ANT to its readers. On the other hand, social science, like natural science, is part of ANT’s topic, what it wants to analyze and understand without interference or censorship. ANT’s maxim “Follow the actor!” also applies when it comes to social scientists. But as we have seen, ANT rarely manages to do this without pointing out their faults. Now, I shall have more to say about ANT’s attempt to corner the social sciences below. For now, let us concentrate on what ANT has to say when it treats social science as a topic of analysis. An important point for ANT is that all knowledge is of the same kind, it is only the length and stability of the networks that vary. If the natural sciences are different from the social sciences, it is only because they build with different kinds of materials, like electrons, fuel cells, and fish stocks. With Latour and Woolgar (1986:257) we might say: "The only difference [between natural scientists and us social scientists] is that they have a laboratory." But this formulation of the difference is difficult to sustain, since a laboratory for ANT is not a very confined place. If Pasteur managed to turn France into a laboratory it would seem to imply that also sociologists have access to this strategy. Callon’s (1998) exploration of the relationship between economic models and market phenomena demonstrates the point. Here, the starting premise is that economic theory performs, shapes and formats the economy rather than observing and analyzing how it functions (ibid: 2). More specifically, economic theory works because the market models, with all the preconditions and qualifications we find in economics textbooks, have been successfully transported and brought to life in the real world by means of legislation that creates property rights, sets the conditions for transactions, regulates competition,

153

Holm

bans monopolies, and the like. Just as Pasteur turned French society into his laboratory, economists have tremendous success in making the world perform their theories. Perhaps one could say that the economic model is not so much a map of the economy, as the working plan from which the economy has been constructed. In an earlier paper, Callon and Latour (1981) have argued that the same applies not only to economics, but also to sociology. In this paper, they uses Hobbes’ Leviathan as a general metaphor for the state, or any macro-actor for that matter. Instead of Hobbes’ social contract they suggest the general framework of translation to explain how such macro-actors are created and can act “like one man.” The point is that also sociologists, in their attempt to analyze such creatures, at the same time performs them: What else do sociologists do? Like everyone else, they never stop working to define who acts and who speaks. They tape the recollections of a workman, a prostitute or an old Mexican; they interview; they hand out open and closed questionnaires on every subject under the sun; they unceasingly sound out the opinions of the masses. Each time they interpret their surveys they inform the Leviathan, transforming and performing it. Each time they construct a unity, define a group, attribute an identity, a will or a project; each time they explain what is happening, the sociologist, sovereign or author - as Hobbes used the term add to the struggling Leviathans new identities, definitions and wills which enable other authors to grow or shrink, hide away or reveal themselves, expand or contract. (Callon and Latour, 1981: 298).

As long as ANT casts sociology as a topic for analysis, rather than a competitor in the category of “Best Social Theory,” it does not see much difference between the social and natural sciences. While natural scientists are mostly preoccupied with non-humans and sociologists are most interested in humans, they basically do the same things: They try to make their favourite type of entities perform theories. In both cases, success is decided by the market test: What counts is not whether a theory is true or not, but that it works. In this way, “sociologies” are for ANT the outcomes of scientific settlements. They result when actor-networks are stabilized, and can therefore not be used as analytical resources. While social scientists are major producers of such sociologies, they do not have any monopoly position in this market. Far from it; natural scientists and engineers are usually much better at delivering sociologies, since they prefer to build them instead of simply talking about them

154

The Invisible Revolution

The Modern Constitution ANT’s rejection of the divide between the world and the word, between objects and accounts of objects, between ontology and epistemology, makes it different in important respects to both the social constructivism of SSK and the naïve realism of the sciences. It is ironic, perhaps, that this aspect of ANT’s position becomes equally puzzling from both these perspectives. Hence, SSK believes that ANT, when the incomprehensible vocabulary has been peeled off, reverts to simpleminded realism (Collins and Yearly, 1992; Bloor, 1999a);7 while the science warriors see ANT as another version of postmodern, relativist anti-science, indistinguishable from SSK (Gross and Levitt, 1994; Koertge, 1998; Sokal and Brichmont, 1998; cf. Callon, 1999). For ANT, however, both the naïve realism typical of the natural sciences and the social constructivism typical of the social sciences are two sides of the same coin. They simply reside on opposite poles of the Kantian divide, one focusing on the realm of objects, the other on the realm of subjects and society. While they are of course opposites, and will often deny the relevance of their counterparts, they are closely related in their common acceptance of the fundamental divide between Nature and Society. Indeed, they are central in creating and policing this divide. But not only that. This dualism is again postulated as the centerpiece of modernity itself. In this way, ANT in a few neat transformations, has redefined itself from a esoteric sub-discipline within the sociology of science into a theory of modernity. ANT is not only on top of the sciences and SSK; it is also claiming to have mastered modernity itself.

Purification versus translation According to ANT, the separation and contrast between Nature and Society is of great significance within modernity. But why? What is its relevance? What is the function of this divide; what type of work does it perform? These question bring us to the contrast between “purification” and “translation.” Modernity is characterized by two distinct sets of practices, which must be kept strictly apart (Latour, 1993: 10). The first of these, which Latour calls the work of purification, is

7

Bloor (1999a: 82) accuses ANT of retreating to Mertonian sociology of science, and Collins and Yearly (1992: 322) think that ANT takes a backward step ‘to the scientists’ conventional and prosaic accounts of the world from which we learned to escape in the early 1970s.’ 155

Holm

intimately tied to the invention of Nature and Society as opposed and completely separate from each other. The work of purification ascribes all existents to two entirely distinct ontological zones, as non-humans or humans, objects or subjects. Within modernity, says Latour, existents are only acknowledged as real to the extent that they can be referred back to their pure form, as either belonging to Nature or Society (see Figure 5).

First dichotomy: Nature vs. Society

Nature Pole

Society Pole

Second Dichotomy: Work of Purification vs. Work of

Hybrids, Networks, Monsters Figure 5: Purification and translation (Adapted from Latour, 1993: Figure 1.1., p. 11)

However, the practice of purification presupposes another practice, namely that of translation, which mixes nature and culture and creates “hybrids,” “monsters,” or “heterogeneous networks,” in short, it blends together what the work of purification pries apart. Examples of such hybrids are the ozone hole, computer chips, high definition television, frozen embryos, in short, all the strange entities that populates the modern world, mixing politics and chemistry, humans and non-humans, culture and science (Latour, 1993: 1). These entities are, at one and the same time,

156

The Invisible Revolution

natural, cultural, and discursive. But this is not only the case for the new, science-intensive creatures that are constantly being invented. It is also the case for all existents; all creatures are like this. The problem is that this insight is almost impossible to retain. Within modernity, the hybrids, in order to be recognized, must be picked apart, analyzed, and referred back to their pure form, as emerging from and belonging to either the Social or the Natural order (Latour, 1993: Chapter 2). The work of purification thus renders the process of translation invisible, unthinkable, and unrepresentable. But it does not in any way slow down the work of translation, the production of new hybrids. On the contrary, “the modern Constitution allows the expanded proliferation of the hybrids whose existence, whose very possibility, it denies” (Latour, 1993: 34). To explain this, Latour draws a parallel with the premoderns. While the moderns ignore and deny the process of translation, the premoderns are endlessly preoccupied with it. For them, the meeting of nature and society is problematic, for it is impossible to change the one without changing the other. Every new hybrid is a possible threat to the social order, and its entry must be treated explicitly and with great caution. In contrast, the moderns, by denying the existence of such hybrids, allow them to proliferate unchecked (Latour, 1993: 41-2). The result is an unending stream of new hybrids: We are invaded by frozen embryos, expert systems, digital machines, sensor-equipped robots, hybrid corn, data banks, psychotropic drugs, whales fitted with radar sounding devices, gene synthesizers, audience analyzers, and so on. Precisely because this process goes unacknowledged, there are no restrictions on which allies the moderns can recruit, be they human, physical, biological, or symbolic. They can therefore create networks of unprecedented length and stability, allowing them to act at a distance as no-one before them. Instead of nature versus society and natural science versus social science, ANT postulates another dichotomy as the fundamental one, that between purification and translation. Pretending to represent totally different, incompatible perspectives, both natural and social scientists are in reality engaged in a collective effort to uphold an illusion of reality as contained between the subject pole of Society and the object pole of Nature. This covers up the real work going on, namely the practical task of producing hybrids through translation processes. The purificationtranslation dichotomy also allows ANT to do more than just “follow the actor.” While the principles of extended symmetry, self-imposed agnosticism and non-censorship indicated that

157

Holm

ANT could not have its own social theory, and could therefore not explain anything but only sit back and describe what others were doing, the distinction between purification and translation as offered here puts ANT back into the game. As we noted in the previous section, ANT shifts between two perspectives on the social sciences. On the one hand – as it routinely does with respect to natural scientists – it wants simply to follow sociologists without interruptions of any kind, finding out what they are up to. On the other hand, ANT wants to beat the shit out of them for not understanding the first thing about how society is really constituted. The purification-translation dichotomy allows ANT to do both. Following the sociologists around, ANT notices that they do very much the same as natural scientists. They are involved in the work of translation, but cover it all up in the process of purification. This means that the social theories they present become incapable of accounting for society, because everything non-human and non-social is systematically either cleansed out, or if that is impossible, magically turned into social relations. So, in contrast to social theory, which is modern and involved in the work of purification, ANT is non-modern; it is in the business of revealing the work of translation. Since ANT has penetrated the illusion of modernity, it can offer a much better social theory than the sociologists. Most importantly, ANT can reintroduce the “missing masses” into sociological theory. The missing masses are, of course, all the actants that make up networks, but mysteriously disappear as networks are punctualized or black-boxed. Sociological theory, engaged as it is in the work of purification, routinely filters out and throws back to nature all materiality before “human being and human doing” (Giddens, 1984: xx) are granted reality as the essence of society. In this way, all the “materials of sociality” are excluded, and sociology is therefore unable to explain the solidity of society (Callon and Latour, 1981; Latour, 1992b). According to ANT, this solidity resides in the networks, or translation-chains, and includes all sorts of materials. But since sociological theory only grants reality to a few points in these networks, their solidity becomes a mystery. ANT hence suggests that traditional sociology is more appropriate for baboons than for humans, since they (the baboons) have to build their society exclusively on unmediated interaction. They cannot include such things as texts, contracts, buildings, machines, and bombs to firm up ephemeral face-to-face interaction. As a result, their societies become as variable as their relationships; they remain in a state of nature, where no-one is strong enough to hold out against any coalition. Unlike humans, baboons are barred from constructing the Leviathan:

158

The Invisible Revolution

But if you transform the state of nature, replacing unsettled alliances as much as you can with walls and written contracts, the ranks with uniforms and tattoos and reversible friendship with names and signs, then you will obtain a Leviathan: ‘His scales are his pride, shut up together as with a close seal. One is so near to another that no air can come between them. They are joined one to another; they stick together that they cannot be sundered’ (Job 41: 15-17). (Callon and Latour, 1981: 284)

ANT’s analysis of how networks are extended and stabilized is not only a theory of science, but a general social theory, a theory about how society, the Leviathan itself, is constructed. Individuals, microbes, machines, and other micro-actors are constructed in the same type of processes and out of the same variety of materials as are macro-actors, be they states, solar systems, armies, or ozone holes. At the outset, all actors are isomorphic and of equal size. Different actors may not be equally successful in extending and stabilizing their networks, however. It is here lasting asymmetries are created, some actors ending up within longer and more stable networks than others. Some of the associations created in this process are ‘social’ (association of bodies), while some can be labeled ‘technical’ (association of materials). Regardless of the stuff from which networks are created, the only thing that matters is their length and stability. Macro-actors are micro-actors seated on top of many black boxes (Callon and Latour, 1981: 286).

The history of modernity How did modernity come about? Who wrote the “modern constitution”? ANT answers by sketching a historical analysis in which major figures in social and natural philosophy during the last four centuries – from Enlightenment philosophers all the way up to present-day postmodernists – are made to represent different stages in creating and widening the gap between Society and Nature (Latour, 1993: Chapter 3). Latour locates the starting point of this process in the confrontation between the physicist Robert Boyle and the philosopher Thomas Hobbes in the

159

Holm

17th century.8 What was their disagreement about? At the center of the debate was the nature of the vacuum, whether or not it could be artificially produced in Boyle’s laboratory, and in case it could, whether this vacuum could us tell anything whatsoever about the vacuum as a naturally occurring phenomenon. More generally, the debate concerned what is to count as knowledge, how truth is constituted. Boyle’s air pump was able, or so he claimed, to produce a vacuum inside an enclosed chamber, where he performed experiments, such as suffocating small animals and putting out candles. The details of these experiments are not what Latour is interested in. What matters is that Boyle’s work on the air pump led to the invention of the modern scientific method of experiment, and the most modern of literary genres, the scientific paper. The key to the experimental method was the idea of producing a micro-cosmos in which the experimenter was in full control and could keep everything constant except the target factors, and then record the effects of their variation on the independent variable. Equally important were the procedures Boyle applied to turn the results of the experiments into certified knowledge. Boyle’s laboratory was, in Latour’s words, a theater of truth. He would gather trustworthy witnesses at the scene of the action, so that they could attest the existence of the facts as established in the experiment. And even when witnesses could not be physically present, they were made present in spirit within Boyle’s reports of the events to the Royal Society. Here he would report the experimental set-up, the procedures undertaken and the results in obsessive detail, so that anyone who wanted could replicate the experiment and confirm the results for him/herself. If all this is well known to us from modern science, it was because Boyle invented it. And he did so against considerable resistance. One target for this resistance was the method of certifying truth: Testimonies from reliable witnesses in a quasi-judicial way. This ran contrary to what was accepted at the time, in which the received method for truth-making, as for instance practiced at the universities, lay in apodectic reasoning, in mathematics, logic and rhetoric. In other words, truth lay in the way a claim or observation would fit into a received body of knowledge as contained in a restricted set of authoritative texts (Shapin, 1996). Boyle rejected all

8

Latour’s main source in his discussion of Hobbes and Boyle is a book about them by two representatives of the Edinburgh School of SSK, Steven Shapin and Simon Shaffer, Leviathan and the Air-Pump. (Shapin and Schaffer, 1985). Both Latour’s use of Shapin and Schaffer’s work (Pels, 1995; Cohen 1997; Jacob, 1998; Preda, 1999), and this work itself (Pinnick, 1998; 1999; Shapin and Schaffer, 1999), have come under attack.

160

The Invisible Revolution

this and wanted, much like his contemporary Descartes, to start afresh with a tabula rasa. With his experimental method and literary style, Boyle wanted to dispense with received wisdom altogether, and instead consult Nature directly. To Hobbes, this was entirely unacceptable. Latour explains this by pointing out the context in which Hobbes found himself, and which troubled him deeply: Civil wars were raging; the Reformation was promoting free interpretation of the Bible on the part of the people and clerics. In other words, the established unity and order of medieval society was being undermined. How could God and King remain one when people could petition God directly, or even designate their own King? The solution for Hobbes, as reflected in his Leviathan, was the unification of the Body Politic. Hobbes portrays the social contract as the solution to the anarchy of the state of nature. The social contract means an escape from this war of all against all and the establishment of civilization because it allows the creation of a Sovereign, a Leviathan, as a representative of the multitude of individuals. While the Leviathan is the Actor of which the citizens are the Authors, this does not threaten his unity. "For it is the Unity of the Representer, not the Unity of the Represented, that maketh the Person One." (Hobbes, 1985, quoted in Latour, 1993: 18). For Hobbes, Power is Knowledge, meaning that there can exist only one Knowledge, and only one Power, the Leviathan himself. This is the source of Hobbes’ misgivings about Boyle’s style of fact-making, of letting Nature speak for itself through experiments. Such a thing would undermine the unity of the Leviathan. Civil war would rage, the state of nature would return, if the citizens feel that they could appeal directly to supernatural spirits, to God, or to inert matter or physical mechanism, to Nature, when they are prosecuted by the civil authority, by Leviathan (Latour, 1993: 19). What has the quarrel between Boyle and Hobbes to do with the modern constitution? As Latour points out, Hobbes’ Leviathan is made up of only citizens, calculations, agreements or disputes. Or rather, Latour adds, thanks to Hobbes and his successors, we are beginning to understand what is meant by social relations, powers, forces and societies. In Boyle’s laboratory, on the other hand, nature has been made to speak for itself. Of course it does not literally speak; the scientists speak for it. Yet, the whole point of detailed and accurate reporting of experimental set-up and procedures, the ideal of replicability, of controlling for all relevant factors, of testimonies of trustworthy witnesses, and so on, is to ensure that nature is reliably represented,

161

Holm

that, in effect, Nature can speak for itself. Or rather, as Latour again reminds us, with Boyle and his successors, we begin to perceive what natural forces, matter and Nature are. To Latour, then, Hobbes and Boyle are like a pair of Founding Fathers, acting in concert to promote one and the same innovation. Hobbes and Boyle quarrel in order to define the two resources that we continue to use unthinkingly: … on the one hand, social force and power; on the other natural force and mechanism. On the one hand, the subject of law; on the other hand, the object of science. The political spokespersons come to represent the quarrelsome and calculating multitude of citizens; the scientific spokespersons come to represent the mute and material multitude of objects. The former translate their principals, who cannot all speak at once; the latter translate their constituents, who are mute from birth. (Latour, 1993: 29)

So, there it is, the origin of the modern dualism between idealism and materialism, Society and Nature, Subject and Object. The rest of the story Latour tells here is one in which the distinction Hobbes and Boyle invented has gradually been pushed further and further apart, and is gradually transformed into the unbreachable gulf, that yawning gap with which we struggle today. Hence, according to Latour, Kant gave it its “truly canonical formulation,” contrasting things-inthemselves on the one hand and the transcendental subject on the other (Latour, 1993: 56), while Hegel raised it to the level of a contradiction and made it the driving force of history (p. 56). While pheneomenology continued to pose this separation in an attempt to overcome it (Latour, 1993: 57-8), the latest versions push the separation to the extreme by denying the relevance of Nature for Society. Hence, Latour quotes Habermas as saying: “I have already suggested that the paradigm of the knowledge of objects has to be replaced by the paradigm of mutual understanding between subjects capable of speech and action” (Habermas, 1987: 295-6; quoted in Latour, 1993: 60). This version of twentieth-century Kantianism, comments Latour, widens the abyss between the scientific and technical rationality on the one hand and speaking and thinking objects on the other. While this is pushing the gap to the extreme, posing a “hyperincommensurability” between Nature and Society, Latour finds an even more extreme version in post-modernism, an intellectual movement that he “has not found words ugly enough to describe” (Latour, 1993: 61). He illustrates his point with a quote from Lyotard, replying to a question from a scientist concerning the link between science and the human community: “I simply maintain that there is nothing human about scientific expansion” (Lyotard, 1988: xxxviii; quoted in Latour,

162

The Invisible Revolution

1993: 61). But, comments Latour, this means that Lyotard, by his insistence on the complete irrelevance of Nature for Society, accepts the very essence of modernity: "The postmoderns believe they are still modern because they accept the total division between the material and technological world on the one hand and the linguistic play of speaking subjects on the other ..." (Latour, 1993: 61). In this rendering of the history of social and natural philosophy, then, the invention of Society and Nature as two radically different entities and ontological spheres takes center stage. Even when philosophers have tried to overcome the distinction, like Hegel or the phenomenologists, they have only managed to recreate it and push the poles even further apart. And when the last generation of pre-postmodernists, like Habermas, or postmodernists, like Lyotard, try to escape from this legacy by declaring Nature irrelevant for Society, they in reality confirm it by indirectly acknowledging the total separation between the two.

Taking ANT to Trial Such is the drama of modernity that ANT wants us accept. There is a fundamental process of translation, of putting together and stabilizing hybrids, monsters or heterogeneous networks. But this goes on without recognition, or rather, it can only proceed when it is actively denied and hidden. This happens in the work of purification, also undertaken by the sciences. Reminiscent of Goffman’s sociology (Goffman, 1971), ANT divides between the back stage of translation, the kitchen-cum-laboratory, where different meals are put together from a number of different ingredients, and the front stage of purification, the restaurant in which all traces of the preparation process have been removed and the courses can appear as finished products. Both natural and social scientists are among the staff in the modern restaurant, taking part in the purification process as well as the translation process. But the purification process is so successful that the scientists themselves have become convinced of its truth. There is no mixing of ingredients going on behind the scene, they insist. All meals are either from the entrées menu of objects-in-nature or the dessert menu of subjects-in-society. Only ANT has gained access to the kitchen of

163

Holm

modernity, reporting back that all meals are hybrids, mixed together from pieces of salt nature and sweet society without any discrimination whatsoever. What can we make of such a fantastic story? Is it just a fairy-tale best left in limbo between the covers of Latour’s book? Or does it belong, on the contrary, to the genre of documentary, having gained access, somehow, to the hidden secret of modernity? What, in other words, is the identity we should ascribe to ANT – science or science fiction? Let us, in trying to answer this, do what ANT recommends as the proper scientific procedure for translation, namely put the entity in question on trial. So let us chain ANT to the chair, and force it to answer some really tough questions, not letting up until it has spilled all its secrets. What is at stake here is of course whether ANT really has come across the secret of modernity. The possibility before us, therefore, is that ANT is an agent for the non-moderns and the missing masses, and that a revolution is imminent. Will the modern constitution be overturned? What we need to know is whether ANT is operating alone, or is in alliance with the strong but hidden forces it claims to represent. Does it know the truth, or is it pulling a bluff? What are its resources? What instruments and weapons does it have at its command? Is ANT guilty of treason, willing and capable of putting all of modernity on the line? Or is it simply performing another version of that well-worn modern genre of denunciation (Latour, 1993: 43-46), pretending to be stepping out in the cold, while at the same time living comfortably within the protected home built for it by modernity? Perhaps ANT is not the dangerous non-modern in league with a silent army of things, as it claims, but a harmless post-modern writer, only in command of a world of its own invention? Let me, following the normal procedure in such cases, keep quiet about the electric shocks, the truth serum, the lie detector tests, the solitary confinements, etc, that were applied in getting ANT to come out with the truth, and just report the answers that were obtained.

Q: How did ANT happen to penetrate the secret of translation, if this is expressly forbidden by the modern constitution and cannot be seen by anyone else? As we know, ANT takes an heroic stance in most issues, recognizing few friends and no allies, least of all within the social sciences. Nevertheless, on this one issue, ANT in fact credits SSK. It

164

The Invisible Revolution

was SSK that broke the spell of the modern constitution. It was SSK that tried the forbidden crossover, applying the repertoire reserved for one end of the Kantian divide on the other: … this breach of the dualists’ game immediately bankrupted the whole enterprise. What had started as a ‘social’ study of science could not succeed, of course, and this is why it lasted only a split second – just long enough to reveal the terrible flaws of dualism. By treating the ‘harder’ parts of nature in the same way as the softer ones – that is, as arbitrary constructions determined by the interests and requirements of a sui generis society – the Edinburgh daredevils deprived the dualists – and indeed themselves, as they were soon to realize – of half of their resources. Society had to produce everything arbitrarily including the cosmic order, biology, chemistry, and the laws of physics! (Latour, 1993: 54-55)

While SSK quickly retreated in the face of this terrible revelation, as we saw in the last chapter, it allowed ANT to see through the whole dualist game and discover the true nature of modernity.

Q: Besides SSK, who quickly returned to modernity and therefore does not count, who are ANT’s accomplices and allies in this non-modern conspiracy? Again, ANT’s first impulse is to deny that anyone else is involved, and insist that all responsibility and credit belong to itself and no other. But this story distributes the odds so incredibly unevenly and is totally unbelievable. On the one side we have four hundred years of social and philosophical discourse, the whole of science, politics and capitalism; in short, modernity itself, the most successful and powerful thing in the known universe. On the other side, there is ANT; a handful of students of science and technology. So come on, we say to ANT, you could not have done this all alone. The least you can do is to give us some predecessors, some philosophical or theoretical tradition that helped you along the way? In the end, we manage to force some names from ANT. Machiavelli, Hobbes, and Nietzsche are among them, but really not as important as some would have it (cf. Barnes, Bloor and Henry, 1996: 115; Shapin, 1988). ANT has learned a great deal from William James, the American pragmatist (see in particular Latour, 1999a: Chapter 2); from symbolic interactionism (Law, 1994: 97-100), and from semiotic analysis (Latour, 1988a; 1988b; 1992b; 1993: 62-65). And in spite of ANT’s adamant condemnation of the post-moderns, both Derrida (Lee and Stenner, 1999) and Foucault (Schaanning, 1996a; 1996b) have clearly been important to ANT. But none of these, except, perhaps the elusive figures of Serres (1995) and Deleuze (Deleuze and

165

Holm

Guttari, 1988) are indicated as brothers-in-arms. If anything, ANT’s ancestry suggests that it could be classified as a post-modern species, but this is difficult to prove, since appropriate methods of genetic analysis are still lacking in this area.

Q: Whose side is ANT on? Who are its clients? If ANT is an agent, like 007, in which country will we find the MI6 that has issued its licence to kill? This is easy, and we scarcely need to hint at our instruments of truth to make ANT confess. ANT’s clients are the non-humans. If ANT is guilty of treason, the side it has betrayed is that of humanity. With Lee and Brown (1994), we can see ANT as the logical end point of social inquiry’s dealings with the Other. The Other here denotes all those entities and areas that become problematic via expansionist projects, be they political or theoretical. Hence, say Lee and Brown, English colonialism created a new definition of Englishness by rendering large parts of the rest of the world and their inhabitants strange – by creating Others. In much the same way, sociology has defined itself and society by marking some areas and entities as outside its disciplinary boundaries. The Other is, of course, not constant over time. If we compare contemporary social theory with those of the Victorian area, it is easy to see that the division of identity/Otherness has undergone drastic revisions, being steadily pushed outwards. The tendency within sociology and social theory has been strongly humanist, insisting that the boundary of the same must be extended to include all humans, regardless of social position, gender, nationality, ethnicity, religion, race, etc, reserving otherness for non-human entities. Within sociology, “… anything non-social or non-human, including the natural world and the world of technological artifacts, has been made Other” (Lee and Brown: 774). So ANT can bee seen as a response to, a protest against, this bracketing off of the non-social: To balance our accounts of society, we simply have to turn our exclusive attention away from humans and look also at nonhumans. Here they are, the hidden and despised social masses who make up our morality. They knock at the door of sociology, requesting a place in the account of society as stubbornly as the human masses did in the nineteenth century. What our ancestors, the founders of sociology, did a century ago to house the human masses in the fabric of sociology, we should do now to find a place in a new social theory of the nonhuman masses that beg us for understanding. (Latour, 1992b: 227)

166

The Invisible Revolution

So ANT is revolutionary, but not on behalf of oppressed humans, but on behalf of things.

Q: But what allows ANT to set itself up as a representative of the silent masses? Why should we trust ANT to speak the true interests of things? And what guarantee do the things themselves have of being truthfully represented by ANT?

At this point in the interrogation, I must report, ANT is clearly beginning to feel the pressure. If its replies until now have been consistent, although perhaps showing a somewhat twisted inner logic, they now begin to lose coherence. Once the revolution has been accomplished and the modern constitution has been replaced with a non-modern one, ANT does not want to be the spokesperson of the silent masses. In contrast to the Marxists, ANT has no ambition of setting itself up as some communist party that by definition speaks for the proletariat. Under the non-modern constitution, says ANT, the work of translation will be explicitly recognized. While this will preserve the moderns’ capacity to create new hybrids, their existence as hybrids will be acknowledged. Under the nonmodern constitution, therefore, the production of new hybrids will be regulated. Thus, the fourth guarantee of the non-modern constitution reads: “[T]he production of hybrids, by becoming explicit and collective, becomes the object of an enlarged democracy that regulates and slows down its cadence” (Latour, 1993: 141). This regulation cannot be one of human control over things, however, since that would be to reintroduce the modern distinction between Nature and Society. Instead, Latour suggests a “Parliament of Things:” In its confines, the continuity of the collective is reconfigured. There are no more naked truths, but there are no more naked citizens, either. The mediators have the whole space to themselves. The Enlightenment has a dwelling-place at last. Natures are present, but with their representatives, scientists who speak in their name. Societies are present, but with the objects that have been serving as their ballast from time immemorial. .... The imbroglios and networks that had no place now have the whole place to themselves. They are the ones that have to be represented; it is around them that the Parliament of Things gathers henceforth. ‘It was the stone rejected by the builders that became the keystone’ (Mark 12:10). (Latour, 1993: 144)

So the final destination of the non-modern revolution is the abolition of all difference. There will no longer be humans-among-themselves, with full citizen rights, and things-in-themselves, silenced

167

Holm

and disenfranchised. What remains are only the hybrids, which now have the place for themselves and are equally represented and spoken for in the parliament of things. This sounds like great news for the silent masses. Finally they get to be represented. Finally, they get to speak their minds. But wait a minute. Why would they want to? What do they have to gain from this proposed transition into non-modernity? What ANT proposes, is not a parliament of actants, but a parliament of hybrids, of actor-networks of some length and stability. The problems this construction is going to solve are the problems brought on by the moderns’success. Reminiscent of Marx’ analysis of capitalism, ANT’s diagnosis of modernity is that it undermines itself. The hybrids are pouring out of the laboratories at an increasing pace, overburdening modernity’s system of sorting and cleaning. This happens because the moderns, unlike anyone else, have forbidden themselves to think about the hybrids and their consequences. The parliament of things, then, is proposed as a way of discriminating among hybrids; a way of deciding which are good and can be allowed to enter into society, and which are bad and should be exiled. It would seem that ANT’s scheme is misnamed. It is not a parliament of things, but a court of things. Why would the hybrids be interested in such a system when under the modern constitution they have complete access to the kingdom of the moderns? If they are reluctant, perhaps suspicious that this is a system where only humans will be the judges and only hybrids will be the accused, who could blame them?

Q: What are the resources ANT has at its disposal? Does it have access to laboratories, as the natural scientists do when they pull off their remarkable reversals of strength? If ANT had taken up the strategies it attributes to the natural scientists it has studied, a laboratory would have been their weapon of choice: “Give me a laboratory, and I will raise the world” (Latour, 1983). In contrast to academic sociologists, who are satisfied to talk endlessly about society, natural scientists, through their laboratories, are in a position to change them: “Microbiology laboratories are one of the few places where the very composition of the social context has been metamorphosed” (ibid: 158). So if ANT wants to bring about a revolution, to change modern society into a non-modern one, it would have to do so by way of laboratories.

168

The Invisible Revolution

Alas, there are no laboratories. Despite ANT’s pretence about abandoning sociology, we see few real attempts in ANT to leave the safety of the seminar rooms. In this respect, ANT is not much different from sociology. All ANT does is tell stories. Sometimes fairy-tales and fables. Sometimes more realistic stories, firmly grounded in facts and figures. But these stories are never cast in solid materials and built into machines, as the natural scientists’ stories are. The main audience for ANT, after all, are humans, mostly academics of the sociological persuasion. ANT is a sociology.

Q: But what about theories and explanations. Does ANT command such wonderful machines for “acting at a distance”? Now, we know here that ANT distinguishes clearly between two forms of explanation (Latour, 1988c). In the first, explanation is involved in attempts at “acting at a distance.” Here, theories and explanations are parts of empire building: “the more powerful an explanation, the larger the empire and the stronger the material in which it is built.” This is the way main-stream science operates. The second form of explanation, which I suppose could be labled “description,” would be to display and render visible the processes and work by which the first type of explanation is made to work: The first way tries to abolish distance, the second feeds on it. In the first, power is reinforced and the represented elements disappear in their representatives. In the second, power is weakened and the initial elements are maintained in full view. The first is reductionist, because holding a single element of A is tantamount to holding all the elements of B. The second I call non-reductionist or ‘irreductionist’ because it adds the work of reduction to the rest, instead of subtracting the rest once the reduction has been achieved. The first creates a power, that is the impression that having one element involves having all the others ‘in potentia’. The second creates what I call a gradient of force (Latour 1988[b]: part II). In other words, the first starts with equivalences without telling through which instruments and through which metrology these equivalences are obtained; the second starts from translations and tries to present the work of rendering elements equivalent by setting up new instruments and keeping long metrological chains in alignment. In still other words, the first tradition accuses and allocates responsibilities, while the second regards accusations as unfair because they always fall back on an innocent scapegoat. The first is on the side of the knowing, the second is also on the side of the known. (Latour, 1988c: 163).

169

Holm

As the tone of this quote indicates, ANT wants to be on the “side of the known,” committed to “always look[ing] for weak explanations rather than for general stronger ones” (Latour, 1988c: 174). But this is not really what ANT is doing in practice, is it? On the contrary, much of ANT is written as an exercise in the noble art of persuasion, of setting up clever translation chains, making it hard for the reader to dissent. In a review of Science in Action, Steven Shapin has suggested that Latour demonstrates extraordinary talent in doing exactly what he claims the scientists under study are doing: Bruno Latour has been following scientists around for years. Now he wants us to follow him following them around. He offers students of science and technology a detailed map that will allow us to follow him easily. He defines the nature, scope and terms of the exercise; he even invents a name – ‘technoscience’ for its object of study. In unmistakenly French fashion he gives us ‘rules of method’ and ‘principles’, numbered and ordered. No one following Latour is meant to get lost or stray off the line of march. Stragglers will have no excuse. There has never been a programme for research in the social studies of science that has been presented in such a systematic and integrated way. (Shapin, 1988: 533).

If this is true for Science in Action (Latour, 1987), it is even more true for We have never been modern (Latour, 1993). Here, Latour’s claim is to have in his hand the explanation of modernity itself. Isn’t this an obvious attempt to “act at a distance,” robbing natural scientists of their objects, social scientists of their subjects, turning everything into creatures under ANT’s command? If there ever were a pure example of science as scapegoating, of setting up gradients and redistributing responsibility, this must be it (cf. Cohen, 1997). Technoscience, says ANT, is war conducted by much the same means (Shapin, 1988: 534). If you want to do battle with the moderns, you cannot refrain from using the most effective weapons available. The only strategy that can beat the army of the moderns is the strategy of science.

Q: But does not this corrupt the ANT enterprise of revealing how science works? How can ANT engage in scientific empire building, when the purpose from the start was to reveal such imperialist endeavors?

Yes, ANT will admit that there may be a problem here. With reference to Science in Action, Latour says: “…there is in my text an unresolved problem which is how to protect against unequivalence, against one part of the word/world winning over the other” (Crawford, 1993:

170

The Invisible Revolution

265). But this doubt seems to have been overcome in We have never been modern (see Kock, 1995). The solution lies in the notion of the non-modern constitution. The problem, it would seem, is not building empires per se, but building modern-type empires. As long as you build, as ANT wants, according to the non-modern constitution, you’re doing fine. Stepping into non-modernity, then, it becomes possible to apply all the weapons of the modern repertoire in the fight against modernity.

Q: What about ANT’s managerialist bias? Isn’t there a danger, despite such admirable commitment to the cause of the “silent masses,” that ANT cannot speak for the weak, that ANT will only lend voice to those who are already winning?

In answering this, ANT starts by pointing out that the scientists are not the only actors who have roles to play in ANT analyses. On the contrary, ANT recognizes a number of vastly different types of entities, humans, things, machines, symbols, etc. However, since most of these entities – like the dissenter Latour introduced in Science in Action – end up without enough resources to continue their dissent and hence become enrolled, translated, pacified, in the networks of the successful scientists, they do not really count. When translation chains have been stabilized, then, we end up with two main categories of actors, on the one hand those who do the translation and act at a distance, Pasteurs, Newtons, Einsteins, or Latours, and, on the other hand, those lowly creatures who become translated and acted upon, those who go along without protest. There might be a bias here, but it is not ANT’s. The problem is that science produces biases; it sets up differentials; it creates winners and losers. This is what translation is for. You cannot possibly blame ANT for the way science and modernity works?! ANT may focus on the winners, on the centers of networks, but only to show that the construction of networks is necessarily a collective enterprise. That is, when the Pasteurs, Newtons, Einsteins, and Latours of the world, by hagiographic storytelling are made to become larger than life, ANT wants to show that this is an illusion, created because they have succeeded in installing themselves as spokespersons in extended networks. Instead of reducing this network to a point, hiding all the work that must be undertaken to create the illusion that Pasteur is out of this world, ANT wants to display the work that goes into creating this very illusion, and hence reduce Pasteur back to normal size (see Latour, 1988b: Introduction).

171

Holm

Q: Perhaps so, but isn’t ANT’s own portrayal of networks, in a totalizing language that reduces everything to point forces and resistances, in itself a silencing strategy, only allowing entities to present themselves in specific ways?

To our great astonishment, ANT agrees with this objection. It admits, it would seem, that its attempt at establishing a final repertoire in order to include the “Other,” can turn into system of thought for colonizing all things (Lee and Brown, 1994). Since ANT focuses on the winners – those who do the translating rather than those who are translated – even if that is to display or uncover the process of translation, this might contribute to silencing those who are silenced, erasing the work of the secretaries, wives, laboratory technicians, and all sorts of associates (Star, 1991: 29; Elam, 1999). In order to avoid this problem, ANT immediately comes up with a number of less constraining vocabularies. One of these is borrowed from Deleuze and Guttari (1988), and contrasts striated space, in which “all lines … are subordinated to the points and are made to resonate to the point, and all points are made to resonate with the center” (Lee and Brown, 1994: 784), and smooth space, which is occupied by rhizomes, indeterminate collections of elements and “in which communication runs from any neighbour to any other, the stems and channels do not pre-exist, and all individuals are interchangable, defined only by their state at a given moment – such that the local operations are coordinated and the final global result synchronized without a central agency” (Deleuze and Guttari, 1988: 17, quoted from Lee and Brown, 1994: 785). ANT, as it readily admits, has been firmly focusing on striated space (Lee and Brown, 1994: 785). But no longer. From now on, ANT promises, it will ambulate between smooth and striated space; between centralized networks of power and rhizomes. It will, from now on rely more on metaphors and vocabularies that convey less centralized, homogeneous, one-dimensional, and certain situations than the metaphor of the network (cf. Latour, 1999c), be it heterogeneity and multiplicity (Star, 1991), fluids (Mol and Law, 1994), trails (Cussins, 1992), choreography (Cussins, 1996), or fractals (Law, 1999). In short, ANT no longer wants to be ANT. As Latour puts it: “…there are four things that do not work with actor-network theory; the word actor, the word network, the word theory and the hyphen! Four nails in the coffin” (Latour, 1999c: 15) In his paper at the workshop “Actor Network and After,” Law concluded in much the same way:

172

The Invisible Revolution

And this is why I would recommend actor-network theory. I would recommend it because it is weak. Because it is in dissolution. Because it has betrayed itself. Because it has turned itself from signal into noise. Because it no longer exists. Because it has dissolved itself into other ways of seeing, of writing, and of doing.” (Law, 1997: 14).

Our interrogation has come to an end. There is, it would seem, not much left of ANT. Whatever ANT is – but we can of course never be completely certain here – it does not pose a threat to modernity. Its pretence of strength, of having modernity itself on a leash, came down to a bluff. ANT never commanded the resources (laboratories, explanations) it would have needed, and could never represent those it claimed to represent. In the end, it also decided, having been deserted by all its friends, that it would not, after all, want to participate in the “deadly proof race” of scientific warfare, and instead chose weakness as its strategy. This must therefore be my recommendation. ANT did not betray modernity. If it committed crimes, these lay only in the outrageousness of its stories. Since it has also retracted its most fantastic claims and seems to be repenting, ANT should be free to go as it pleases.

ANT and the Invisible Revolution

ANT as irony ANT grew out of, and in important respects builds on, SSK. Perhaps the most important commonality between SSK and ANT is their resistance against natural realism. Both SSK and ANT reject the realist notion of correspondence between knowledge claims and the entities they refer to. Nevertheless, ANT quickly broke with SSK, and in particular its attempt to replace natural realism with its own brand of social realism. For ANT these two realisms are mirror-image perspectives representing the two Kantian repertoires – things-in-nature versus humans-in-society – with the aid of which the moderns impose order and meaning on their world. Rejecting both

173

Holm

natural and social realism, ANT proposes that reality emerges from heterogeneous networks, belonging to neither side of the Kantian divide. Science is involved in hybrid-making, in translation processes firmly grounded in the middle. Instead of understanding science as a process of “discovering” the things-in-themselves, as the natural realists will have it, or as “reflecting” social interests and cultural understandings, as the social realists will have it, ANT insists that both society and nature emerge when hybrid networks stabilize. This means that, in contrast to SSK, whose aim is to understand society in science, ANT’s object of study is science in society (Seguin, 2000). Instead of inquiring how society defines or causes science, as is SSK’s project, ANT is interested in the social role of science; in the ontological labor performed by scientific activity. For ANT, the question is not how social factors “explain” science, but quite the contrary, how society and nature result from science. This also means that ANT cannot be seen simply as a “sociology of science.” Since science and scientific activity play a crucial role in modern society, ANT turns into a general social theory, with particular focus on modernity. Summarized thus, ANT sounds like a serious, respectable analytical discipline, different in philosophical foundation from other social theories, perhaps, but very much part of the same academic/philosophical tradition. Nevertheless, much ANT writing – in particular the work of its main inventor, Bruno Latour – does not fit well with such a characterization. When confronted with this work, one is struck by its heroic posture and its provocational style. This work does not conform to standard academic norms such as intellectual modesty, that all claims should be grounded firmly in arguments and evidence, the obligation to give credit where credit is due, and the responsibility to focus on the issue and factual argumentation rather than engaging in strong emotional and moral rhetoric. ANT’s heroism is particularly obvious in its claim to have mastered modernity itself, to have revealed this greatest deception of all times, to have succeeded – singlehandedly – in slaying this most powerful of monsters so that it can be replaced by a creature of ANT’s own making. Such claims are, of course, in themselves on such a grand scale that they are difficult to take seriously. Where did ANT get the resources to pull off such a fantastic feat? How come ANT could gain such superhuman powers? Does ANT, like the science generals it so skillfully portrays, assume command of such resources, after first having amassed a great number of black boxes, by positioning itself on top of them? In a way, of course, it does. As has been argued in this chapter, ANT is indeed part of a broad academic tradition. If ANT’s authors had wanted to, they could easily have chosen to focus on the way it builds on the work of other social

174

The Invisible Revolution

theorists and philosophers. As it is, however, ANT – in crass contradiction of the tenets of the non-modern constitution – engages in hagiographic story-telling when it comes to its own exploits. ANT does not want to share the credit for its discoveries with anyone. It found the secret of modernity all by itself. ANT is the only one that has succeeded in leaving modernity behind. All other social theories are thus portrayed as contributors to modern-type purifications. It would seem, then, that ANT – much like SSK – has entangled itself in a self-referential quagmire. After having solemnly sworn to expose all purifications and dedicate its resources to restoring the honest work of translation, it immediately falls back on a modern-type purification strategy when accounting for its own claim to success. ANT’s heroic mastery of modernity conforms exactly to the modern-type empire-building that ANT observes – and wants to expose – in science. But how can ANT engage in a modern-type empire-building when the express purpose is to reveal such practices in others? How can we put faith in such double talk? The solution, I think, is that we cannot take ANT at face value. Instead, ANT’s heroism and provocations can only be seen as irony. ANT says: “See, I can do as the moderns. I can also build spectacular networks, balancing the whole world on a pin-head.” But the claims are exaggerated and press the issue into absurdity. Instead of making its claims as reasonable and continuous with other work as possible, it chooses to be maximally provocative, making its points by estrangement. In doing this, of course, ANT is helped along by its opponents, willingly contributing to the work of pushing ANT into absurdity, thus revealing modernity at work. In this interpretation, then, ANT has solved the problem of reflexivity by way of irony. ANT is indeed part of modernity. Its claim to non-modernity is a play of words. As part of science, ANT cannot do otherwise than the scientists it studies, always building heterogeneous networks and trying to stabilize them by all means possible. But when ANT does so, it does not for a second claim to have found the truth, to have discovered how reality really works. From its ironic position, it insists, as it performs modernity: “See what I am doing. Notice the techniques I am applying. This is modernity at work. Don’t for a second believe what I tell you!”

175

Holm

The invisible revolution This leaves us with the question about ANT and the invisible revolution. Can ANT help us where institutional perspectives and SSK failed? Is ANT able to make the resource management revolution visible again, without, like institutionalism and SSK, first turning the fish into pure social constructs? As you suspect, my answer to these questions is in the positive. Let us take a more detailed look at how ANT will deal with the invisible revolution, and how its answers differ from those of institutionalism and SSK. Recall the basic problem of the invisible revolution. I had observed important changes in practices, organizational structures, values, and perspectives connected with the introduction of fisheries resource management. I believed that these changes were so important and of such a systemic nature that they constituted a fundamental institutional transition, shifting the institutional set-up of the fisheries from one type of configuration to a very different type: An institutional revolution. The problem with this interpretation, however, was that virtually no-one else shared it. It was not that they did not see that changes had occurred, because that was obvious and easy to document. The problem was to see this as a revolution, thereby acknowledging these changes as man-made and conditional; that they were not inevitable and determined by the nature of things, or rather, the nature of fish stocks. Instead, resource management – given the vulnerability of the fishery resources, as documented by fishery science – was perceived of as necessary and rational. Since the established institutional order, the MSO system, could not protect the fish, and to some extent even comprised organizational principles, values and mechanisms that were counter-productive from a resource management perspective, it had to be abandoned. From this viewpoint, there had been changes, but no revolution. What had happened was a rational and necessary adaptation to the nature and situation of the fishery resources. As we recall from Chapters 2 and 3 above, both institutionalism and SSK were perfectly willing to grant reality to the invisible revolution. Both these perspectives instinctively rejected the realist interpretation. From both these perspectives, the premise on which fisheries resource management rests – that of limited and vulnerable fish stocks in need of management – is redefined from fact to fiction. Both institutionalists and SSKers refuse to attribute more faith to such a story than to the story on which the MSO system was based, featuring a world in which the basic problem is too much fish. Both institutionalists and SSKers will, of course, acknowledge

176

The Invisible Revolution

that the resource management story is based in science and that the MSO story is not. And this difference is also recognized as an important reason why the first gains credibility over the latter. But neither institutional theory nor SSK will for a second consider the possibility of accepting the first as objectively true and the second as false. In both perspectives, such stories are internal to the institutions in question. They arise as results of institutions, and therefore cannot explain a given institutional configuration, or a shift from one configuration to another. Here, both institutionalists and SSKers will be relativists, recording and comparing accounts and forms of knowledge, but never allowing that one of them is true and the other false. In the picture painted for us by these perspectives, then, there can be no direct confrontations between society and nature. So the revolution cannot be about this, as in the naïve realist interpretation. But if nature itself can never make any difference, accounts of nature surely can. In these perspectives, then, there can again be revolution because the two sides confronting each other are both human. On the one hand, there is the MSO system, featuring accounts that talk mostly about social relations. On the other hand, there is the resource management system, featuring accounts that talk mostly about nature. Our revolution occurs as the latter type of accounts beat the first type of accounts. In this context, institutionalists and SSKers are both heeding the tenet of symmetry, applying the same analytical framework for explaining nature and society. But this symmetry exists only because nature first has been totally transported into society. Having done this, there can again be revolution: Clashes of interests, conflicting viewpoints, incompatible forms of life. This move solves the initial problem. We clearly see the revolution; it has stopped being invisible. This solution comes with another problem, however. In order for it to work, we must accept the highly improbable notion that nature is a social invention; that the hardest of all realities, constantly asserting itself in everyday experience and confirmed by science in theory and practice, is essentially a construction hinging on our beliefs in it. ANT offers an alternative to such a problematic position. For ANT, both the MSO system and the resource management system are heterogeneous networks. They are different in configuration, length and stability, yes. But apart from this, they are basically the same. They are hybrids, imbroglios, or monsters, created in an ongoing work of translation.

177

Holm

The story in this thesis, as seen from ANT, is one in which one such network was replaced by another. Suddenly, over scarcely more than a decade, the MSO network lost its integrity and was replaced by the resource management network. Call it revolution if you like. Instead of such a political vocabulary, usually reserved for account of interactions among humans, ANT may have preferred a more neutral conceptualization, one that would be equally sensitive to the activities of non-humans as to the activities of humans. But revolution is all right, as long as this is not used in order to ignore the most important of our revolutionary heroes, the fish itself. In contrast to institutionalists and SSKers, then, ANT allows the fish to make a difference. Instead of throwing it out, or turning it into something entirely social, ANT lets it join in and contribute to the revolution. Of course, the fish cannot accomplish the whole thing by itself. But when it is allowed to play its part, we will finally be able to explain why the MSO system was replaced by resource management. But we need to be careful here. In the attempt to allow the fish to make a difference, we do not want to fall back into naïve natural realism. This means that we cannot simply talk about the fish as an objectively given entity. If we say, in the executive summary of our report of the invisible revolution, that the fish brought it about, deciding that the resource management system served its interests better than the MSO system, we must quickly add that such a shift of allegiance was the result of a long, cumbersome and costly process, involving, among other things, a complete transformation of the identity of the fish itself. While ANT insists, in its quarrel with SSK, that non-humans must be allowed to play a part, it is not retreating to a realist position. On the contrary. Central to ANT’s idea of fisheries resource management as a heterogeneous network is that a tremendous amount of work is required in order to construct and stabilize a suitable identity and role for the fish. If resource management is going to work, the fish must be transported from the bottom of the ocean and into board rooms and decision making processes. The fish does not need to be present itself, of course. Just like the fishers, the fishing industry, the scientists, the state, etc, the fish is made present by someone speaking on its behalf, explaining its interests, viewpoints and possible courses of actions by means of words, charts, tables, theories, anecdotes, and statistics. For ANT, then, a crucial step in the attempt to explain the resource management revolution is a careful description and documentation of the effort needed to set up translation chains through which the fish regularly and reliably can be transformed from elusive creatures of the ocean into effective participants in decision-making processes. As explained in

178

The Invisible Revolution

Paper 4, and elaborated in much finer detail in Smith (1994), this was indeed a complex and difficult process, taking a great number of talented people a century or so until a reasonable result was achieved. A couple of important points should be noted with regard to ANT’s analysis of this process. The translation chain by which the fish is transported from the ocean into decision-making processes should not be seen as simply representational, thus producing an accurate conceptual “picture” of fish in society standing for an objective entity out there in nature. Rather, for ANT, the fish is transformed as it is transported. All the representational work that must be undertaken in order for the fish to appear in decision-making processes helps define what the fish is under fisheries resource management, be it the procedures of sampling and measuring of fish; all the work involved in collecting and producing catch statistics; the cleaning up and modelling of data; the construction and authorization procedures of fish stock assessment; the work involved in predicting stock developments under different catch regimes, etc. These activities are seen by ANT as ontological work: They define the fish, which therefore should not be confused with some purified image of the fish as an eternal, objective entity, as the realists would have it. In the fisheries case, the realist misunderstanding is fairly easy to avoid. The scientific controversies are still raging, and the network has not yet reached a particularly stable configuration. The model of the fish that is used for assessment purposes is obviously a gross simplification, deliberately leaving out of the picture a number of factors that are of great importance. If translations always involve simplification and betrayal, these must be said to be unusually obvious and crass in the fisheries case. But so what? All this means is that the final move of the purification process, in which the identity of the being created by the network can be postulated as the cause of knowledge about it has not been reached. But this does not mean that we can switch from a realist understanding to the mirrorimage social constructivist understanding, holding that the model of the fish can be completely explained by reference to social factors. Against both these interpretations, ANT maintains that fisheries resource management as a heterogeneous network by necessity engages human as well as non-human entities. In the construction of fisheries management networks, non-human actors and processes are involved, like fish, ecosystems, recruitment, mortality and growth. But it also involves humans, like the fishers. If the construction of a fish fit for management involved drastic

179

Holm

reconstructive work, this is no less true for the fishers. Not only where their rationality and social role completely changed as the open fisheries commons were enclosed and turned into a state protectorate. When the fish was transported onto the negotiation tables as a numbered quantity, it could also be dished out in more or less fixed units, and tied to vessels and skippers by way of licenses and quota schemes. This, in turn, allowed fish to be bought and sold in markets, allowing not only a whole new development of vessel and fleet structures, but also impinging heavily on the distribution of political resources among the fishers (Paper 2). As this suggests, the reconstructions of the fish and the fishers go hand in hand within the fisheries resource management network. As a result of the ontological work performed by the construction and beginning stabilization of fisheries resource management, both humans and non-humans are fundamentally redefined. And the story does not end with the fish and the fishers. Also other entities are transformed, including the fishery scientists, who succeeded in establishing themselves in an obligatory passage point, monopolizing knowledge about the fish. The state was transformed, gaining control over resources to which it earlier had no access. Even the international political/economic order was to some extent transformed by fisheries resource management. Instead of fishery resources as a global common resource, freely accessible to everyone who had the means, the stabilization of fishery resource management came with a new oceans regime, in which the fishery resources were defined as the property of the coastal states. Is it reasonable to see the introduction of fishery resource management as a revolution? Of course it is. The construction and stabilization of resource management is a revolution. It transforms people – the fishers, the fishing communities, the economic and organizational structure of the sector, the scientific community, the state and the global political and economic order. But not only that. In addition, it transforms the fish, the ecosystems, nature itself. When fisheries resource management appears as a stabilized network, nature and society have been transformed and redefined: A revolution has occurred. We have reached our goal. With the help of ANT, the invisible revolution can finally appear in all its splendor. Since the creation and stabilization of networks is regarded as an ontological process, from which reality – society and nature – emerges, we no longer have problems seeing the invisible revolution. Our eyes are no longer clouded by the purifications offered us by natural and social scientists. We are not fooled by the realists, saying that there was

180

The Invisible Revolution

no revolution since fish fit for management had only to be discovered. Since we now know that that the fish fit for management is a creature invented by the scientists, we realize how profound the revolution has been. Nor are we fooled by the social constructivists, wanting us to believe that fishery resource management could not be that sensational, since it simply reflected dominant perceptions and interests in the social context. This attempt at “socialization” is just the mirrorimage of the “naturalizing” move by the realists, making fish and fisheries management blend into the environment. When the social constructivists want to talk about the introduction of this as a revolution, then, it is not to emphasize the depth and novelty of the institutional configuration that is established, but part of the strategy to shift emphasis and responsibility back to the human side of things. But from an ANT perspective, the revolutionary character of the transition is not only limited to the politics and power struggle involved. With the stabilization of the resource management network, society, culture and politics go through fundamental changes; they are recreated and constituted in new shapes. At the same time, the fish, the ecosystems, nature itself are metamorphosed. We cannot explain these changes as reflections of existing cultural configurations or established societal interests. Culture and society, as well as nature, are outcomes of these processes and can therefore not be used as explanations for them. Such is the true revolutionary character of this transition. The stabilization of fisheries resource management exemplifies the ontological processes capable of transforming nature and society from the most basic of levels, from where reality itself begins.

181

Holm

182

The Invisible Revolution

5. INVENTING THINGS THAT DO NOT EXIST

The Fiction of the Invisible Revolution During the 1980s and early 1990s, great changes took place in Norwegian fisheries. The MSO system, featuring a complex of corporatist organizational arrangements protecting the petty capitalist fisher-owner, was replaced by the resource management system, the key goal of which is to protect fish stocks. When the latter system gradually replaced the first, the consequences reached virtually every aspect of the fisheries. It redefined the fish and the natural environment in which it belongs. It gave the fishers new identities. It shifted power around, so that actors who used to be insignificant came into key positions, while actors who used to be powerful found themselves thrown out of office. It turned the sector-political system on its head, tearing up established coalitions, allowing the establishment of new ones. It redefined the basic values of the fisheries, focusing less on fishing as a way of life, and more on fishing as a rational economic enterprise. It changed the fisheries’ significance for and role in the larger society. In short, a revolution. But there was something peculiar about this revolution. It was as if it had never happened. It remained virtually imperceptible, uncommented on and apparently unnoticed, not only by the fishers and their political organizations, but also by journalists, bureaucrats and academics. A revolution had happened, but it had, for some unknown reason, disappeared from view. It was an invisible revolution. This was the basic puzzle this thesis postulated and wanted to resolve: The mystery of the missing revolution. This meant that the thesis had to confront two related problems. Not only would it have to document and explain the details of the transition, the identity of the sector system as it was before the revolution happened, the identity of the sector system as it was established afterwards, and the mechanics of the process of change itself. In addition, it would have to account for the fact that the process of change to such a large extent remained undetected and unaccounted for. There had been a revolution, but it had been made invisible. How could the thesis explain its invisibility, thus making it visible again?

183

Holm

This project turned out to be more difficult than it perhaps seemed at the outset. This is because the most reasonable explanation for the invisibility of the revolution is simply that there was no revolution. Think about the apparent incompatibilities of the two institutional configurations that make up the beginning and end points of the process of change in question. At the start you have the “hot” and explicitly political institutional forms tied to the MSO system. At the end you have the “cool” and scientifically rational forms of intervention connected to fisheries resource management. The two institutional configurations, the collision of which drives our process of change, are equipped with very different origin stories, modes of legitimation, and discursive styles. Where the former is explicitly political, going with the fishers in their struggle against exploitation, the second is rational, focusing on how to best utilize the fishery resources. Where the first takes sides where questions of value, equity and justice are concerned, the second insists on impartiality and neutrality and only seeks the common interest. Where the first resides wholly within society and politics, the second is securely tied to nature. If the institutional process of change that occurred when MSOs were pushed out by resource management came about as a collision between society and nature, then, we can no longer talk of revolution in a literal sense. Such a collision is not, in the first instance, about politics and power ploys, as are revolutions. Instead, it is about a society that had become overly optimistic in its faith in nature’s gifts. When nature subsequently stopped delivering and no longer could be taken for granted, society had to repent. The extravaganzas of the MSO system were called off. In their place we got collective restraint in the shape of resource management. There is no revolution here, only learning and adaptation, the replacement of an established institution that had stopped working with a new one that would. If we accept this interpretation, the thesis has a problem. Our key phenomenon, the “invisible revolution,” has turned into a metaphor. We cannot take it literally any more, and need to put it between inverted commas. What can we say to such an objection? Should the project be called off, even before it has started? No, of course not. On the contrary, this only makes it more challenging. The more naturalizing explanations that the claim of an invisible revolution encounters, the more convinced we becomes that it is up against a conspiracy. If they take so much trouble to deny that something out of the ordinary is going on, we must surely be on to something! The revolution didn’t go

184

The Invisible Revolution

missing because of some misunderstanding or simply by oversight. Not at all, the thesis must insist. As revolutions go, ours is spectacular and interesting. If it went unnoticed, it can only be because it was actively hidden. The reason why the revolution does not look like one, then, is not that there never was one, but that it was so successfully camouflaged. The resource management revolution was one in which the winner took all, including the history and self-understanding of the losing side. The revolution was, with the aid of great cunning and a huge amount of work, deliberately turned into a non-revolution. The mystery of a revolution that went missing. How was such a remarkable vanishing act accomplished? Who was responsible? Who are the heroes of our story? Who are the villains? What resources must be mobilized in order to solve the mystery and sort out truth from deception? What happens if we do? How can we know if we get it right? Such questions have been at the center of our efforts in the preceding pages, the findings of which can be summarized as follows: Science and scientific knowledge played an important role in making the invisible revolution disappear. That science was a key suspect for the cover-up was clear from the start. It did not take much effort to see that a driving force behind the changes that went on in the fishery during the 1980s and 1990s was science’s claim to have a rational solution to the resource management problem. Without solid scientific backing, the resource management system would not have been able to topple the existing MSO system. But if we accept the scientific claims on which resource management rests as simply the truth, the revolution is in danger. If we concede that the fish-as-fit-for-management is an eternal, objective entity discovered by science, we might as well have given up from the start. In order to keep the revolution intact, then, we would need to undo the claims of science. A tough challenge if there ever was one. But not quite impossible. Indeed, this seemed like a great opportunity to try out the potential and limits of the analytical repertoires developed in social theory for deconstructing hard-core rationalist and realist understandings. Following in the footsteps of my own struggle with these issues for the past 10 years, the thesis hence put to the test three different analytical perspectives: Neo-institutional theory, the sociology of scientific knowledge, and actor-network theory. Our first candidate and wannabe hero was neo-institutional theory. A basic thrust in this theory is a firm rejection of all rationalist-type reifications. In place of seeing institutions as

185

Holm

objective and external to society, institutionalists want to demonstrate that institutions are socially constructed and thus exist the extent that they are believed in and enacted. As such, neoinstitutionalist perspectives will be skeptical of the rational-scientific foundation of fisheries resource management, since this threatens to ground social institutions outside society. This also means that neo-institutionalism will be sympathetic to the invisible revolution. Nevertheless, we found that the neo-institutionalists, while in principle willing to write off scientific claims as cultural constructions, were quite helpless in confronting claims of scientific truth on a more practical level of analysis. Neo-institutionalists remained at a safe distance from science and scientific knowledge, and did not take on the analysis of science as an institution, perhaps because of their own ambition to be involved in scientific truth-making. Institutionalism, therefore, had not developed analytical techniques or methodologies for dealing with science. Our first candidate failed miserably and fell flat on its face. Our next candidate was the sociology of scientific knowledge. As a version of institutional theory specializing in scientific knowledge, this could bring us a good deal further towards our goal than the neo-institutionalists. Sharing the institutionalist rejection of realist reifications, SSK had no problem dealing with science’s claim of having “discovered” the fish-as-fit-for management. For SSK, such claims can and must be explained by reference to cultural and societal factors. If they are taken to be true, this is due to power constellations and societal mechanisms of authorization of belief, and has little to do with the true nature of those poor creatures swimming around in the ocean. While this perspective can indeed deal effectively with the realist objection to the invisible revolution, it does so in a way that is highly problematic. Can we really believe that the fish, the oceans, nature itself, are social constructions? Can we really take seriously the notion that nature is completely contained within society? Perhaps we can, if we really work hard and consistently at it. But this is a perspective that requires a highly developed imagination, and calls for extremely cumbersome analytical work and elaborate explanations in order to explain the simplest of everyday reality. So SSK has its problems. And then we haven’t even mentioned the problem of reflexivity, of how to understand SSK’s own claims to truth, when it follows from its perspective that all knowledge claims are social constructions. Our third candidate, and as we now know, the real hero of the story, is ANT. In contrast to the first two, ANT not only casts natural science as one of the villains in the mystery of the

186

The Invisible Revolution

invisible revolution, but also includes social science as a co-conspirator. Such cover-up activities as those of making the resource management revolution disappear are seen by ANT as a normal process of science in action. ANT’s label for it is “purification,” which is undertaken by natural and social sciences alike as part of their effort to stabilize heterogeneous networks of their own making. According to ANT, then, both the natural and the social sciences construct the objects featured in their knowledge claims. But this ontological work is then carefully hidden. The heterogeneous networks coming out of science must, in order to be acknowledged as real under the modern constitution, be referred back as belonging to either pure nature or transcendental society. These are the processes which allow the resource management revolution to go unnoticed. Institutional theories and SSK can perhaps help us reveal the way natural science will try to hide the resource management revolution behind the image of an eternal fish-as-fit-formanagement. But in its place, the social scientists will only offer us another purification, namely resource management as the enactment of rationalized, modern-type cultural configurations in the fisheries. With ANT, however, we have the resources to clear away both these mirror-image purifications. Finally, the resource management revolution comes into sight in its full splendor. The revolutionary character of the introduction of fisheries resource management can neither be captured in a language of natural scientific “discoveries” nor that of social constructivist “reflections.” As the result of long, hard labor to construct and stabilize networks, fisheries resource management constitutes a new nature and a new society. When fisheries resource management replaces MSOs, new realities are born. Such is the extent of the invisible revolution.

The Invisible Revolution is a Heterogeneous Network Are you amazed? Isn’t it fantastic how all the pieces fall into place. Isn’t it like magic? Who could have believed it? When our story about the invisible revolution began a few chapters ago, it was such an unlikely tale, with all the odds against it. Not only did no-one but its author believe in it, but for it to gain credibility, it would have to confront and deconstruct the most impervious and mighty items on the inventory of modernity – that of scientific truth and objectivity. Furthermore, as we found out along the way, the most obvious resources at hand – those offered by the social constructivist perspectives of institutionalism and SSK – could not really help us. In order for the

187

Holm

resource management revolution to be real, it required us to step out of modernity, leaving the safety of natural and social science alike. But once we risked that move, all the pieces suddenly fell into place. All the mysteries dissolved. Isn’t it great? Aren’t you impressed? I sure hope so. But I also hope that you realize how this impression was created. Did you think I started out, as the thesis pretended, with the observation of the invisible revolution? That I saw the puzzle, and then tried to find a way to account for it? You are not that naïve, are you? The invisible revolution never existed. It is a creature of my invention that should not be confused with the realities of Norwegian fisheries. If you have learned anything from reading this thesis, it should be this: There never was an invisible revolution in Norwegian fisheries in the 1980s and 1990s. The invisible revolution is pure fiction. This thesis is not in the documentary genre. It does not tell a true story. The invisible revolution is a science fiction fantasy, which for the moment has no existence outside the covers of this thesis. Hopefully, it is also an entertaining story. As you will surely have noticed, it borrows its basic plot from a Norwegian fairy tale, “The princess who always had to have the last word.” Instead of the task of silencing the Princess, however, the wannabe heroes of our story were made to take on the task of uncloaking the invisible revolution. Institutional theory and SSK were cast in the roles of Per and Pål, who predictably fell short of the challenge and were left with little honor. Our real hero, ANT, armed, just like Espen Askeladd, with a collection of discarded objects, succeeded against all the odds, and could claim the prize, in this case, explanatory control over the resource management revolution. Are you still amazed that all the pieces fall into place as our story concludes? Are you still entertaining the idea that this must be because the story of the invisible revolution is true, grounded in objective realities? No, of course not. The reason ANT can pull off its tricks in the story is simply that this was the way the story was written. When all the pieces were made to fall into place towards the end, this does not demonstrate any command over reality whatsoever. If ANT were made to succeed where institutional theory and SSK failed, this in no way proves that ANT is more resourceful than its two competitors. When the thesis let institutional theory and SSK fall on their faces, this has very little to do with their inherent character traits. The story of the invisible revolution was willfully constructed in order to make this happen. Institutional theory and SSK did not get fair, impartial treatment, but were portrayed from the perspective of ANT. Indeed, if you

188

The Invisible Revolution

read the thesis as it was written, starting with Chapter 4 on ANT, and then return to Chapters 2 and 3, you will see that the latter two are exemplary ANT-type analyses of its rival theories. As much as I would want to insist that the invisible revolution is real, and all I did was “discover” it, that simply will not do it. The invisible revolution is a network construction. For now, the network in which it lives scarcely manages to reach outside the covers of this thesis. Unless the thesis becomes widely read, believed in, cited, and the invisible revolution thus starts being enacted – a possibility that at the moment when these lines are written seems extremely remote – the invisible revolution will remain a fiction. As such it is not true. It has no separate existence. It never happened. But this does not mean that it cannot come true. Even if this is unlikely, it is not impossible. For this to happen, however, a tremendous amount of work has to be done. A tiny fraction of this work I can do myself, such as having this thesis published, preferably by a highquality international publishing house, doing the appropriate promotion work; use it as a basis for getting scholarly recognition – earning a Ph.D., getting promoted to a full professorship, perhaps, or even winning some academic prizes. If such strategies succeeds, it will boost my own academic authority, thus stabilizing the network in which the invisible revolution lives. Another set of strategies would be to try to make other people acquaint themselves with and use ANT, for instance by arranging courses and seminars, having students use it in their theses, perhaps inspiring new Ph.D. theses or research programs. If such strategies work, the perspectives and language in which the invisible revolution makes sense will gradually be spread. Even if such work is important, it will not amount to much. Even if I could go on doing such tasks forever, I could only accomplish a small fraction of the work needed in order for the invisible revolution to become real. If this is to happen, it has to be a collective achievement, undertaken by a great number of people, the first in line of which are the readers of this thesis. In the end, the invisible revolution can only become real if the readers of this thesis take it as real, and thus help enact it. But why would they do that when – as they should have learned from the thesis itself – they cannot believe that the story of the invisible revolution is true? I must admit that I have some doubt here. Perhaps it is not very efficient to mobilize people when the naturalizing strategies of scientific purification are not available. Perhaps it is impossible, after all, to make people 189

Holm

contribute to the building of networks when they realize that this is exactly what they are doing. I really do not know. But then, I – the author – am not the one to answer these questions. It is all up to the readers, isn’t it. Who am I to prefigure their moves? Instead of discussing such questions, then, it is more important to think about the consequences if the invisible revolution – against all odds – is carried outside the covers of this thesis and made real. Two related points can be made in answer to this question. The first starts with the fact that, in comparison to your average social science and natural science critique of fisheries resource management, an ANT analysis of the invisible revolution invites appreciation and admiration for the tremendous effort and achievement involved. When the long, hard work of constructing the fish-as-fit-for-management finally succeeded, it resulted in a large-scale, totally unexpected transformation of nature and society. The fish and their ecosystem changed; the fishers and their communities were transformed; the property rights regime in marine resources underwent fundamental changes; the structure and workings of the fishery were reformed; fisheries politics metamorphosed; the global oceans regime shapehifted. Can you avoid admiring this accomplishment? Even if you dislike the thrust of the changes – the destruction of coastal communities; the dispossession of the commoners; the redefinition of fishing from “way of life” into “productionist” practices; the substitution of scientists for local fishers as the real ocean experts – even if you think that this goes in the wrong direction, the pulling off of such a drastic transformation must still be appreciated and admired. And whatever its critics say, fisheries resource management is not a failure. How can it be, having accomplished all these highly unexpected, dramatic transformations? Perhaps it does not do all that has been promised on its behalf. Perhaps one can imagine regimes that could make a better job of it. Perhaps one can imagine management systems that would be able to achieve better results with less effort. Perhaps. But that it is not perfect, that it can and should be improved, does not mean that it is a failure. What would have happened if the resource management revolution had not happened? How would the status of fishery resources have been without it? How would the “arms race” of global capital against common fishery resources have ended without the intervention of the new management regime? How would the fish have fared if it had not been included as an effective partner in decision-making processes in the fisheries? Even if there are many objections to be raised against modern fisheries resource management, even if there are many ways in which it can

190

The Invisible Revolution

and should be improved, a critique that keeps quiet about its great accomplishments remains insufficient. This is our first point: ANT’s uncloaking of the invisible revolution allows for a better appreciation of fisheries resource management and its accomplishments. ANT reveals the translation process, clearing away the mirror-image purifications of natural and social science. Our second point follows directly from this. When fisheries management is revealed as a heterogeneous network, it also becomes easier to see how it can be improved upon. As seen from ANT, fisheries management escapes from its shadow existence, be it either as grounded in eternal, objective nature or as a secondary “reflections” of dominant cultural and social configurations. When this happens, and resource management can finally be appreciated as a heterogeneous network, it also becomes easier to change. And of course there is room for improvement. The established management configuration, despite all its admirable sides, also has a number of problems that should be fixed. While it is working well as seen from the perspective of the state and capitalist enterprises, it does not seem to be very good for the fish, deliberately and grossly ignoring some of its most precarious interests. It is extremely bad for coastal communities and small-type coastal fishers. It also seems to be far too costly, inflexible, and centralized. There is not much doubt in my mind that a sounder, more flexible, more democratic, and more efficient system can be found in the direction of co-management, as suggested by social scientists (Papers 7 and 8). From and ANT perspective, however, the strategy the social scientists employ in order to build co-management is all wrong. Instead of acknowledging comanagement as a heterogeneous network, as ANT would, the social scientists try to ground it in social relations only. Paradoxically, the first move in their chain of arguments is to steal the fisheries scientists’ invention, the fish-as-fit-for-management, transport it back in time and make it the property of a creature of their own invention, the fisher-manager (Paper 8). Not only do the co-management activists accept the purification offered them by the natural scientists. When the fishery scientists pretend they did not laboriously construct the fish-as-fit-for-management, but simply discovered it, the social scientists believe them, and add: “Yes, we know. Fisheries management is nothing new. The fishers did it all along. All we have to do is restore the management institutions of old.”

191

Holm

In contrast, ANT wants to recognize the construction of co-management as truly innovative. Co-management never did exist. Nor did the self-managing fishing communities as invented by the social scientists. Co-management cannot be, for ANT, about the re-discovery of age-old institutions, which can subsequently be restored. Co-management is a totally new thing, that could turn out to be no less revolutionary than the biologically based fishery management model. Furthermore, it cannot be built out of social materials alone, as the social scientists seem to think. In order for it to work, such a network can neither rely purely on traditional ecological knowledge, as constructed by the social scientists, nor on the single-stock model of the existing regime. A better starting point for fisheries co-management would be an explicit reinvention of the fish. While fishers, social scientists and others can surely contribute to this, it is hard to see how it can be done without the fishery biologists. Instead of alienating the biologists, trying to make the fisher-manager do their work, as the social scientists tend to do, they need to be in on the project from the start. This is my second and final point, then. If the invisible revolution is believed in and turned into reality, it will make it easier to pull off the next revolution, which may occur in the direction of co-management. Acknowledging the existing resource management regime, as well as comanagement, for what they are – heterogeneous networks – will make it easier to see and appreciate their strength as well as their faults. This is my only argument for the invisible revolution. It does not exist. It is not real. But if it is given a chance, it can do good work. It can help create a nature/society that will be better for fish and fishers to live in.

There is nothing more for me to do. It is all up to you.

192

The Invisible Revolution

References Amsterdamska, Olga, 1990. ‘Surely You Are Joking, Monsieur Latour!’ Science, Technology, & Human Values 15(4): 495-504. Apostle, Richard, Gene Barrett, Petter Holm, Svein Jentoft, Leigh Mazany, Bonnie McCay, Knut Mikalsen, 1998. Community, State and Market on the North Atlantic Rim: Challenges to Modernity in the Fisheries. Toronto: Toronto University Press. Arthur, Brian, 1990. ‘Positive Feedbacks in the Economy.’ Scientific American 262(2): 92-99. Ashmore, Malcolm, 1989. The Reflexive Thesis: Wrighting Sociology of Scientific Knowledge. Chicago: The University of Chicago Press. Ashmore, Malcom, Robin Wooffitt, and Stella Harding, 1994. ‘Humans and Others, Agents and Things.’ American Behavioral Scientist 37(6): 733-740. Asbjørnsen, Per Christian and Moltke Moe, 1992. ‘The Princess Who Always Had to Have the Last Word.’ In Per Christian Asbjørnsen and Jørgen Moe (eds.) Norwegian Folk Tales. Volume One: 85-87. Oslo: Grøndahl Dreyer. Barbalet, J.M., 1987. ‘Power, Structural Resources, and Agency.’ Current Perspectives in Social Theory 8: 124. Barley, Stephen R., 1986. ‘Technology as an Occasion for Structuring: Evidence from Observations of CT Scanners and the Social Order of Radiology Departments.’ Administrative Science Quarterly 31: 78-108. Barley, Stephen R. and Pamela S. Tolbert, 1997. ‘Institutionalization and Structuration: Studying the Links between Action and Institution.’ Organization Studies 18(1): 93-117. Barnes, Barry, 1974. Scientific Knowledge and Sociological Theory. London: Routledge & Kegan Paul. Barnes, Barry, 1977. Interests and the Growth of Knowledge. London: Routledge & Kegan Paul. Barnes, Barry, 1982. T. S. Kuhn and Social Science. London: Macmillan. Barnes, Barry, 1983. ‘Social Life as Bootstrapped Induction.’ Sociology 17:524-45. Barnes, Barry, 1985. ‘Thomas Kuhn.’ In Skinner, Quinten (ed.) The Return to Grand Theory in the Human Sciences: 83-100. Cambridge: Cambridge University Press. Barnes, Barry, 1992. ‘Realism, Relativism and Finitism.’ In Diederick Raven, Lieteke Van Vucht Tijssen, Jan de Wolf (eds.) Cognitive Realism and Social Science: 131-147. New Brunswick: Transaction Publishers. Barnes, Barry, 1998. ‘A History of Regression: Actors, Networks, Machines, and Numbers.’ Environment and Planning A 30:203-223. Barnes, Barry, David Bloor, and John Henry, 1996. Scientific Knowledge: A Sociological Analysis. London: Athlone. Barnes, Barry and Donald McKenzie, 1979. ‘On the Role of Interests in Scientific Change.’ In Roy Wallis (ed.) On the Margins of Science: The Social Construction of Rejected Knowledge:49-66. Sociological Review Monograph 27. Keele: University of Keele. Ben-David, Joseph, 1971. The Scientist’s Role in Society. Englewood Cliffs: Prentice-Hall.

193

Holm

Berger, Peter L. and Thomas Luckmann, 1967. The Social Construction of Reality. New York: Doubleday. Bhaskar, Roy, 1989. The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences. New York: Harvester Wheatsheaf. Bijker, Wiebe E., Thomas P. Hughes, and Trevor Pinch, (eds.) 1987. The Social Construction of Technology Systems: New Directions in the Sociology and History of Technology. Cambridge: MIT Press. Bloor, David, 1981. ‘Hamilton and Peacock on the Essence of Algebra.’ In H. Mehrens, H. Bos, and I. Schneider (eds.) Social History of the Nineteenth Century Mathematics: 202-232. Boston: Birkhäuser. Bloor, David, 1983. Wittgenstein: A social Theory of Knowledge. London: Macmillan. Bloor, David, 1991 [1976]. Knowledge and Social Imagery. Chicago: The University of Chicago Press. Bloor, David, 1992. ‘Left and Right Wittgenstainians.’ In Andrew Pickering (ed.) Science as Practice and Culture: 266-282. Chicago: University of Chicago Press. Bloor, David, 1996. ‘Idealism and the Sociology of Knowledge.’ Social Studies of Science 26: 839-856. Bloor, David, 1997. Wittgenstein, Rules and Institutions. London: Routledge. Bloor, David, 1999a. ‘Anti-Latour.’ Studies in History and Philosophy of Science 30(1): 81-112. Bloor, David, 1999b. ‘Reply to Bruno Latour.’ Studies in History and Philosophy of Science 30(1): 131-136. Bourdieu, Pierre, 1990. The Logic of Practice. Oxford : Polity Press. Breslau, Daniel, 2000. ‘Sociology after Humanism: A Lesson from Contemporary Science Studies.’ Sociological Theory 18(2): 289-307. Brint, Steven and Jerome Karabel, 1991. ‘Institutional Origins and Transformations: The Case of American Community Colleges.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 337-360. Chicago: University of Chicago Press. Brinton, Mary C. and Victor Nee (eds.), 1998. The New Institutionalism in Sociology. New York: Russel Sage Foundation. Brochmann, Bjørn, 1980. ‘Distrikts-Norge bør ikke opprettholdes gjennom sysselsetting som er verre enn å gjøre ingen ting!’ Fiskets Gang 66(49) 767-771. Brochmann, Bjørn, 1981. ‘Virkninger på lang sikt av statsstøtte til fiskeriene.’ Sosialøkonomen 2: 23-29. Brunsson, Nils, 1989. The Organization of Hypocrisy: Talk, Decisions and Actions in Organizations. Chichester: John Wiley & Sons. Callon, Michel, 1986. ‘Some Elements of a Sociology of Translation: Domestication of the Scallops and the Fishermen of St. Brieuc Bay.’ In John Law (ed.) Power, Action and Belief: A New Sociology of Knowledge?: 196-233. London: Routledge. Callon, Michel, 1987. ‘Society in the Making: The Study of Technology as a Tool for Sociological Analysis.’ In Wibe E. Bijker, Thomas P. Hughes and Trevor J. Pinch (eds.) The Social Construction of technological Systems: New Directions in the Sociology and History of Technology: 83-103. Cambridge: MIT Press. Callon, Michel, 1991. ‘Techno-economic Networks and Irriversibility.’ In John Law (ed.) A Sociology of Monsters: Essays on Power, Technology and Domination: 132-161. London: Routledge.

194

The Invisible Revolution

Callon, Michel, 1994. ‘Four Models for the Dynamics of Science.’ In Sheila Jasanoff, Gerald E. Markle, James C. Petersen, and Trevor Pinch (eds.) Handbook of Science and Technology Studies: 29-63. Thousand Oaks: Sage. Callon, Michel, (ed.) 1998. The Laws of the Market. Oxford: Blackwell. Callon, Michel, 1999. ’Whose Imposture?’ Social Studies of Science 29(2): 261-286. Callon, Michel and Bruno Latour, 1981. ‘Unscrewing the Big Leviathan: How Actors Macro-Structure Reality and How Sociologists Help Them to Do so.’ In K. Knorr Cetina and A.V. Cicourel (eds.) Advances in Social Theory and Methodology: Toward an Integration of Micro and Macro Sociologies: 277-303. London: Routledge. Callon, Michel and Bruno Latour, 1992. ‘Don’t Throw the Baby Out with the Bath School.’ In Andrew Pickering (ed.) Science as Practice and Culture: 343-368. Chicago: The University of Chicago Press. Callon, Michel and John Law, 1995. ‘Agency and the Hybrid Collectif.’ The South Atlantic Quarterly 94(2): 481-505. Callon, Michel and John Law, 1997. ‘After the Individual in Society: Lessons on Collectivity from Science, technology and Society.’ Canadian Journal of Sociology 22(2): 165-182. Campbell, John L. and Leon N. Lindberg, 1990. ‘Property Rights and the Organization of Economic Activity by the State.’ American Sociological Review 55: 634-647. Carruthers, Bruce G. and Wendy Nelson Espland, 1991. ‘Accounting for Rationality: Double-Entry Bookkeeping and the Rhetoric of Economic Rationality.’ American Journal of Sociology 97(1): 31-69. Christensen, Pål and Abraham Hallenstvedt, 1990. På første hånd: Norges Råfisklag gjennom femti år. Tromsø: Norges Råfisklag. Clawson, Mary Ann, 1989. Constructing Broderhood: Class, gender, and Fraternalism. Princeton: Princeton University Press. Coase, Ronald H.,1937. ‘The Nature of the Firm.’ Economica 16: 386-405. Coase, Ronald H., 1960. ‘The Problem of Social Cost.’ Journal of Law and Economics 3: 1-44. Cohen, Ira J., 1989. Structuration Theory : Anthony Giddens and the Constitution of Social Life. Basingstoke: Macmillan. Cohen, Michael D., James G. March and Johan P. Olsen, 1972. ‘A Garbage Can Model of Organizational Choice.’ Administrative Science Quarterly 17: 1-25. Cohen, Sande, 1997. ‘Science Studies and Language Suppression - A Critique of Bruno Latour’s We Have Never Been Modern.’ Studies in the History and Philisophy of Science 28(2): 339-361. Collins, Harry M., 1992 [1985]. Changing Order: Replication and Induction in Scientific Practice. Chicago: University of Chicago Press. Collins, Harry M., and Trevor Pinch, 1994. The Golem: What Everyone Should Know about Science. Cambridge: Cambridge University press. Collins, Harry M. and Steven Yearly, 1992. ‘Epistemological Chicken.’ In Andrew Pickering (ed.) Science as Practice and Culture: 301-326. Chicago: The University of Chicago Press. Collins, Randall, 1981. ‘On the Micro-foundation of Macro-Sociology.’ American Journal of Sociology 86: 984-1014. 195

Holm

Collins, Randall, 1988. Theoretical Sociology. San Diego: Harcourt Brace Janovich. Craib, Ian, 1992. Anthony Giddens. London: Routlegde. Crawford, T. Hugh, 1993. ‘An Interview with Bruno Latour.’ Configurations 2:247-69. Cussins, Adrian, 1992. ‘Content, Embodiment and Objectivity: The Theory of Cognitive Trails.’ Mind 101(404): 651-688. Cussins, Charis, 1996. ‘Ontological Choreography: Agency for Woman Patiens in an Infertility Clinic.’ In M. Berg and A. Mol (eds.) Differences in Medicine: 166-201. Durham: Duke University Press. Czarniawska, Barbara and Guje Sévon, (eds.) 1996. Translating Organizational Change. Berlin: Walter de Gruyter. David, Paul, 1986. ‘Understanding the economics of QWERTY: The Neccessity of History.’ In W. Parker (ed) Economic History and the Modern Historian: 30-49. London: Blackwell. Deleuze, G. and F. Guttari, 1988. A Thousand Plateaus: Capitalism and Schizophrenia. London: Athlone. DiMaggio, Paul J., 1986. ‘Structural Analysis of Organizational Fields.’ In Barry Straw and Lawrence Cummings (eds.) Research in Organizational Behavior 8 (Special Issue): 335-370. Greenwitch: JAI Press. DiMaggio, Paul J., 1988. ‘Interest and Agency in Institutional Theory.’ In Lynne G. Zucker (ed) Institutional Patterns and Organizations: Culture and Environment: 3-21. Cambridge: Ballinger. DiMaggio, Paul J., 1991. ‘Constructing an Organizational Field as a Professional Project: U.S. Art Museums, 1920-1940.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 267-292. Chicago: University of Chicago Press. DiMaggio,Paul J. and John Mohr, 1985. ‘Cultural Capital, Educational Attainment, and Marital Selection.’ American Journal of Sociology 90(6): 1231-1261. DiMaggo, Paul J. And Walter W. Powell, 1991a. ‘Introduction.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 1-38. Chicago: University of Chicago Press. DiMaggo, Paul J. and Walter W. Powell, 1991b [1983]. ‘The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 63-82. Chicago: University of Chicago Press. Douglas, Mary, 1986. How Institutions Think. New York: Suracuse University Press. Durkheim, Emile, 1951 [1897]. Suicide: Glencoe: Free Press. Durkheim, Emile, 1982 [1895]. ‘The Rules of Sociological Method.’ In Steven Lukes (ed) Durkheim: The Rules of Sociological Method and Selected Texts on Sociology and its Method: 31-163. London: The MacMillan Press. Durkheim, Emile, 1995 [1912]. The elementary forms of religious life. New York: Free Press. Eder, Klaus, 1996. The Social Construction of Nature: A Sociology of Ecological Enlightenment. London: Sage. Edge, David, (ed.) 1999. Social Studies of Science 29(2). Elam, Mark, 1999. ‘Living Dangerously with Bruno Latour in a Hybrid World.’ Theory, Culture & Society 16(4): 1-24.

196

The Invisible Revolution

Eythórsson, Einar, 1999. ‘Sannheter og makt i forskningen: Den sosialkonstruktivistiske kritikken av biologisk og samfunnsvitenskapelig fiskeriforskning.’ Prøveforelesning til graden dr. polit. Tromsø: Det samfunnsvitenspelige fakultet. Fevolden, Svein E and Jørgen S. Christiansen, 1999. ‘The Coastal Cod – N.E. Arctic Cod Enigma in Northern Norway.’ Abstract. Hirtshals: Third Nordic Symposium in Marine Biology. Fevolden, Svein E. and G.H. Pogson, 1997. ‘Genetic Divergence at the Synaptophysin (Syp I) Locus among Norwegian Coastal and North-east Arctic Populations of Atlantic Cod.’ Journal of Fish Biology 51(5): 895908. Finlayson, Alan Christopher, 1994. Fishing for Truth: A Sociological Analysis of Northern Cod Stock Assessments from 1977 to 1990. St.John’s: ISER. Fligstein, Neil, 1985. ‘The Spread of the Multidivisional Form among Large Firms, 1919-1979.’ American Sociological Review 50: 377-391. Fligstein, Neil, 1990. The Transformation of Corporate Control. Cambridge: Harvard University Press. Friedland, Roger and Robert R. Alford, 1991. ‘Bringing Society Back In: Symbols, Practices, and Institutional Contradictions.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 232-263. Chicago: University of Chicago Press. Garfinkel, Harold, 1967. Studies in Ethnomethodology. Englewood Cliffs: Prentice Hall. Giddens, Anthony, 1976. New Rules of Sociological Method. London: Hutchinson. Giddens, Anthony, 1979. Central Problems in Social Theory. London: MacMillan. Giddens, Anthony, 1984. The Constitution of Society. Berkeley: University of California Press. Gingras, Yves, 1995. ‘Following Scientists through Society? Yes, but at Arm’s Length!’ In Jed Z. Buchwald (ed.) Scientific Practice: Theories and Stories of Doing Physics: 123-148. Chicago: University of Chicago Press. Gingras, Yves, 1997. ‘The New Dialectics of Nature.’ Social Studies of Science 27(2): 317-3134. Goffmann, Erving, 1971 [1959]. The Presentation of Self in Everyday Life. Harmondsworth: Penguin Books. Golinsky, Jan, 1998. Making Natural Knowledge: Constructivism and the History of Science. Cambridge: Cambridge University Press. Granovetter, Mark, 1985. ‘Economic Action and Social Structure: The Problem of Embeddedness.’ American Journal of Sociology 91: 481-510. Grendestad, Gunnar and Per Selle, 1995. ‘Cultural Theory and the New Institutionalism.’ Journal of Theoretical Politics 7(1): 5-27. Gross, Paul R. and Norman Levitt, 1994. Higher Superstition: The Academic Left and Its Quarrels with Science. Baltimore: The John Hopkins University Press. Guneriussen, Willy, 1996. Aktør, Handling, Struktur: Grunnlagsproblemer i Samfunnsvitenskapene. Tromsø: Tano. Habermas, Jürgen, 1987. The Philosophical Discourse of Modernity: Twelve Lectures. Cambridge: MIT Press.

197

Holm

Hacking, Ian, 1983. Representing and Intervening: Introductory Topics in the Philosophy of Natural Science. Cambridge: Cambridge University Press. Hacking, Ian, 1991. ‘The Making and Molding of Child Abuse.’ Critical Inquiry 17: 253-288. Hacking, Ian, 1992. ‘The Self-Vindication of the Laboratory Sciences.’ In Andrew Pickering (ed.) Science as Practice and Culture: 29-64. Chicago: The University of Chicago Press. Hacking, Ian, 1999. The Social Construction of What? Cambridge: Harvard University Press. Hall, Peter A. and Rosemary C.R. Taylor, 1996. ‘Political Science and the Three New Institutionalisms.’ Political Studies XLIV: 936-957. Hobbes, Thomas, 1985 [1651]. Leviathan. London: Penguin. Holm, Petter, 1991. ‘Særinteresser versus almenninteresser i forhandlingsøkonomien: Om Hovedavtalen for norsk fiskerinæring.’ Tidsskrift for Samfunnsforskning 32: 99-119. Holm, Petter, 1999. ‘Fisheries Resource Management as a Heterogeneous Network.’ Paper presented to the Ecological Knowledge Working Seminar, St.Francis Xavier University, Nova Scotia, May 24th-31st, 1999. Holm, Petter and Jahn Petter Johnsen, 1990. Hovedavtalen for fiskerinæringa ved en norsk EF-tilpasning. Tromsø: Fiskernes EF-utvalg. Holm, Petter and Leigh Mazany, 1995. ‘Changes in the Organization of the Norwegian Fishing Industry.’ Marine Resource Economics 10: 299-312. Immergut, Ellen M., 1996. ‘The Theoretical Core of the New Institutionalism.’ Politics & Society 26 (1): 5-34. Jentoft, Svein, 1999. ‘Healthy Fishing Communities: An Important Component of Healthy Fish Stocks.’ Fisheries 24(5): 28-29. Jacob, Margaret C., 1998. ‘Reflections on Bruno Latour’s Version of the Seventeenth Century.’ In Noretta Koertge (ed.) A House Built on Sand: Exposing Postmodernist Myths about Science: 240-254. New York: Oxford University Press. Jepperson, Ronald L., 1991. ‘Institutions, Institutional effects, and Institutionalism.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 143-163. Chicago: University of Chicago Press. Kenshur, Oscar, 1996. ‘The Allure of the Hybrid: Bruno Latour and the Search for a New Grand Theory.’ Annals of the New York Academy of Sciences 775: 288-297. Keohane, Robert, 1988. ‘International Institutions: Two Research Programmes.’ International Studies Quarterly 32: 379-96. Knorr, Karin, 1981. The Manufacture of Knowledge: An Essay on the Constructivist and Contextualist Nature of Science. Oxford: Pergamon. Knorr-Cetina, Karin, 1994. ‘Laboratory Studies: The Cultural Approach.’ In Sheila Jasanoff, Gerald E. Markle, James C. Petersen, and Trevor Pinch (eds) Handbook of Science and Technology Studies: 140-166. Thousand Oaks: Sage. Koch, Robert, 1995. ‘The Case of Latour.’ Configurations 3: 319-347. Koertge, Noretta, (ed.) 1998. A House Built on Sand; Exposing Postmodernist Myths about Science. New York: Oxford University Press.

198

The Invisible Revolution

Krasner, Stephen D., 1984. ‘Approaches to the State: Alternative Conceptions amd Historical Dynamics.’ Comparative Politics 16(2): 223-46. Krasner, Stephen D., 1988. ‘Sovereignty: An Institutional Perspective.’ Comparative Political Studies 21 (1): 66-94. Kuhn, Thomas S., 1970. The Structure of Scientific Revolutions. Chicago: The University of Chicago Press. Lakatos, Imre, 1978. ‘History of Science and its Rational Reconstructions.’ In John Worrall and Gregory Currie (eds.) The Methodology of Scientific Research Programmes. Philosophical Papers. Volume 1: 102-138. Cambridge: Cambridge University Press. Latour, Bruno, 1983. ‘Give me a Laboratory and I Will Raise the World.’ In Karin Knorr Cetina and Michael Mulkay (eds.) Science Observed: 141-70. Cambridge: Harvard University Press. Latour, Bruno, 1987. Science in Action: How to Follow Scientists and Engineers Through Society. Cambridge: Harvard University Press. Latour, Bruno, 1988a. ‘A Relativistic Account of Einstein’s Relativity.’ Social Studies of Science 18: 3-44. Latour, Bruno, 1988b. The Pasteurization of France. Cambridge: Harvard University Press. Latour, Bruno, 1988c. ‘The Politics of Explanation: An Alternative.’ In Woolgar, Steve (ed.) Knowledge and Reflexivity: New Frontiers in the Sociology of Knowledge: 155-176. London: Sage. Latour, Bruno, 1990. ‘Drawing Things Together.’ In Michael Lynch and Steve Woolgar (eds.) Representation in Scientific Practice: 19-68. Cambridge: MIT Press. Latour, Bruno, 1991. ‘Technology is Society Made Durable.’ In John Law (ed.) A Sociology of Monsters: 103131. London: Routledge. Latour, Bruno, 1992a. ‘Pasteur on Lactic Yeast: A Partial Semiotic Analysis.’ Configurations 1: 129-145. Latour, Bruno, 1992b. ‘Where Are the Missing Masses? The Sociology of a Few Mundane Artifacts.’ In W.E. Bijker and John Law (eds.) Shaping Technology/Building Society: 225-258. Cambridge: MIT Press. Latour, Bruno, 1992c. ‘One More Turn after the Social Turn …’ In Ernan McMullin (ed.) The Social Dimension of Science: 272-294. Notre Dame, Indiana: University of Notre Dame Press. Latour, Bruno, 1993. We Have Never Been Modern. New York: Harvester Wheatsheaf. Latour, Bruno, 1998. ‘Ramsès II est-il mort de la tuberculose?’ La Recherche 307: 84-85. Latour, Bruno, 1999a. Pandora's Hope: Essays on the Reality of Science Studies. Cambridge: Harvard University Press. Latour, Bruno, 1999b. ‘For David Bloor… and Beyond: A Reply to David Bloor’s ‘Anti-Latour’.’ Studies in History and Philosophy of Science 30 (1) 113-129. Latour, Bruno, 1999c. ‘On Recalling ANT.’ In John Law and John Hassard (eds.) Actor Network Theory and After: 15-25. Oxford: Blackwell. Latour, Bruno and Steve Woolgar, 1986 [1979]. Laboratory Life: The Construction of Scientific Facts. Princeton: Princeton University Press. Laudan, Larry, 1982. ‘More on Bloor.’ Philosophy of the Social Sciences 12: 71-74. Law, John, 1994. Organizing Modernity. Oxford: Blackwell. 199

Holm

Law, John, 1997. ‘Traduction/Trahison: Notes on ANT.’ Centre for Social Theory and Technology, Keele University, UK. Http://www.keele.ac.uk/depts/staff/jl/pubs-JL2.htlm. Law, John, 1999. ‘After ANT: Complexity, Naming and Topology.’ In John Law and John Hassard (eds.) Actor Network and After: 1-14. Oxford. Blackwell. Layder, Derek, 1987. ‘Key Issues in Structuration Theory: Some Critical Remarks.’ Current Perspectives in Social Theory 8: 25-46. Lee, Nick and Steve Brown, 1994. ‘Otherness and the Actor Network.’ American Behavioral Scientist 37(6): 772-790. Lee, Nick and Paul Stenner, 1999. ‘Who Pays? Can We Pay Them Back?’ In John Law and John Hassard (eds.) Actor Network Theory and After: 90-112. Oxford: Blackwell. Lorber, J., and S.A. Farrell, 1991. The Social Construction of Gender. Newbury Park: Sage. Lukes, Steven, 1982. ‘Introduction’. In Steven Lukes (ed) Durkheim: The Rules of Sociological Method and Selected Texts on Sociology and its Method: 1-31. London: The MacMillan Press. Lynch, Michael, 1985. Art and Artifact in Laboratory Science: A Study of Shop Work and Shop Talk in a Research Laboratory. London: Routledge & Kegan Paul. Lynch, Michael, 1992. ‘Extending Wittgenstein: The Pivotal Move from Epistemology to the Sociology of Science.’ In Andrew Pickering (ed.) Science as Practice and Culture: 215-265. Chicago: University of Chicago Press. March, James G. and Johan P. Olsen, 1989. Rediscovering Institutions: The Organizational Basis of Politics. New York: Free Press. March, James G. and Herbert A. Simon, 1958. Organizations. New York: Wiley. Masterman, Margaret, 1970. ‘The Nature of a Paradigm.’ In Imre Lakatos and Alan Musgrave (eds.) Criticism and the Growth of Knowledge: 59-89. Cambridge: Cambridge University Press. McKenzie, Donald, 1978. ‘Statistical Theory and Social Interests: A Case Study.’ Social Studies of Science 8: 35-83. McMullin, Ernan, 1992. ‘Introduction: The Social Dimensions of Science.’ In Ernan McMullin (ed.) The Social Dimension of Science: 1-26. Notre Dame: University of Notre Dame Press. Merton, Robert K., 1996a [1938]. ‘Science and the Social Order.’ In Piotr Sztompka (ed.) R.K.Merton On Social Structure and Science: 277-285. Chicago: The University of Chicago Press. Merton, Robert K., 1996b [1942]. ‘The Ethos of Science.’ In Piotr Sztompka (ed.) R.K. Merton On Social Structure and Science: 267-276. Chicago: The University of Chicago Press. Meyer, John W., 1992a [1983]. ‘Centralization of Funding and Control in Educational Governance.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 179-198. Newbury Park: Sage. Meyer, John W., 1992b [1983]. ‘Organizational Factors Affecting Legalization in Education.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 217-232. Newbury Park: Sage.

200

The Invisible Revolution

Meyer, John W., 1992c [1983]. ‘Innovation and Knowledge Use in American Public Education.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 233-260. Newbury Park: Sage. Meyer, John W., 1994a. ‘Institutional and Organizational Rationalization in the Mental Health System.’ In Richard Scott and John W. Meyer (eds.) Institutional Environments and Organizations: Structural Complexity and Individualism: 215-227. Thousand Oaks: Sage. Meyer, John W., 1994b. ‘Social Environments and Organizational Accounting.’ In Richard Scott and John W. Meyer (eds.) Institutional Environments and Organizations: Structural Complexity and Individualism: 121136. Thousand Oaks: Sage. Meyer, John W. and Brian Rowan, 1991 [1977]. ‘Institutionalized Organizations: Formal Structure as Myth and Ceremony.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 41-62. Chicago: University of Chicago Press. Meyer, John W. and Brian Rowan, 1992 [1983]. ‘The Structure of Educational Organizations.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 71-98. Newbury Park: Sage. Meyer, John W., W. Richard Scott, (eds.) 1992a [1983]. Organizational Environments: Ritual and Rationality. Newbury Park: Sage. Meyer, John W., and W. Richard Scott, 1992b [1983]. ‘Centralization and the Legitimacy Problem in Local Government.’ In Meyer, John W., and W. Richard Scott, (eds.). Organizational Environments: Ritual and Rationality: 199-216. Newbury Park: Sage. Meyer, John W., W. Richard Scott and Terrence E. Deal, 1992 [1983]. ‘Institutional and Technical Sources of Organizational Structures: Explaining the Structure of Educational Organizations.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 45-67. Newbury Park: Sage. Meyer, John W., W. Richard Scott, and David Strang, 1994. ‘Centralization, Fragmentation, and School District Complexity. In Scott, W. Richard and John W. Meyer (eds.) Institutional Environments and Organizations: Structural Complexity and Individualism: 160-178. Thousand Oaks: Sage. Meyer, John W., W. Richard Scott, David Strang, and Andrew L. Creighton, 1994. ‘Bureaucratization Without Centralization: Changes in the Organizational System of U.S. Public Education, 1940-1980.’ In Scott, W. Richard and John W. Meyer (eds.) Institutional Environments and Organizations: Structural Complexity and Individualism: 179-206. Thousand Oaks: Sage. Mohr, John W., 2000. ‘Introduction: Structures, Institutions, and Cultural Analysis.’ Poetics 27: 57-68. Mol, Annemarie and John Law, 1994. ‘Regions, Networks and Fluids: Anaemia and Social Topology.’ Social Studies of Science 24: 641-71. Mulkay, Micahel, 1991. Sociology of Science: A Sociological Pilgrimage. Milton Keynes: Open University Press. Mulkay, Michael, J. Potter, and S. Yearly, 1983. ‘Why an Analysis of Scientific Discourses is Needed.’ In K. Knorr Cetina and M. Mulkay (eds.) Science Observed: Perspectives on the Social Study of Science: 171-203. London: Sage. Nee, Victor, 1998. ‘Sources of the New Institutionalism.’ In Mary C. Brinton and Victor Nee (eds.) The New Institutionalism in Sociology: 1-16. New York: Russel Sage Foundation. Norges Høyesterett, 1995. Høyesterettsdom av 11. Desember 1995 i sak lnr 157/1995, snr 81. Norsk Retstidende 30/1995: 1901-3.

201

Holm

North, Douglass C.,1981. Structure and Change in Economic History. New York: Norton. Oliver, Christine, 1991. ‘Strategic Responses to Institutional Processes.’ Academy of Management Review 16:145-179. Ostrom, Elinor, 1986. ‘An Agenda for the Study of Institutions.’ Public Choice 48: 3-25. Ostrom, Elinor, 1990. Governing the Commons: The Evolution of Institutions for Collective Action. New York: Cambridge University Press. Pálsson, Gísli, 1991. Coastal Economies, Cultural Accounts: Human Ecology and Icelandic Discourse. Manchester: Manchester University Press. Pels, Dick, 1995. ‘Have We Never Been Modern? Towards a Demontage of Latour’s Modern Constitution.’ History of the Human Sciences 8(3): 129-141. Perrow, Charles, 1986. Complex Organizations: A Critical Essay. New York: Random House. Pickering, Andrew, 1984. Constructing Quarks: A Sociological History of Particle Physics. Edinburgh: Edinburgh University Press. Pickering, Andrew, 1992. ‘From Science as Knowledge to Science as Practice.' In Pickering, Andrew (ed.) Science as Practice and Culture: 1-26. Chicago: The University of Chicago Press. Pickering, Andrew, 1995. The Mangle of Practice: Time Agency & Science. Chicago: The University of Chicago Press. Pickering, Andres, 1999. ‘In the Land of the Blind…’ Social Studies of Science 29(2):307-311. Pickering, Andrew, 2000. ‘The Objects of Sociology: A Response to Breslau’s “Sociology After Humanism”.’ Sociological Theory 18(2): 308-316. Pinch, Trevor, 1986. Confronting Nature. Dordrecht: D.Reidel Publishing Company. Pinnick, Cassandra L., 1998. ‘What’s Wrong with the Strong Programme’s Case Study of the “Hobbes-Boyle” Dispute?’ In Noretta Koertge (ed.) 1998 A House Built on Sand: Exposing Postmodernist Myths about Science: 227-239. New York: Oxford University Press. Pinnick, Cassandra L., 1999. ‘Caught in a Sandy Shoal of the Shallow: Reply to Shapin and Schaffer.’ Social Studies of Science 29(2): 253-256. Pope, Whitney, Jere Cohen and Lawrence E. Hazelrigg, 1995 [1975]. ‘On the Divergence of Weber and Durkheim: A Critique of Parsons’ Convergence Thesis.’ In Peter Hamilton (ed.) Emile Durkheim: Critical Assessments. Volume 8, Section 4: 229-242. London: Routledge. Powell, Walter W., 1991. ‘Expanding the Scope of Institutional Analysis.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 183-203. Chicago: University of Chicago Press. Powell, Walter W. and Paul J. DiMaggio, (eds.) 1991. The New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press. Preda, Alex, 1999. ‘The Turn to Things: Arguments for a Sociological Theory of Things.’ The Sociological Quarterly 40(2) 347-366. Riker, William H., 1980. ‘Implications from the Disequilibrium of Majority Rule for the Study of Institutions.’ American Political Science Review 74: 432-446.

202

The Invisible Revolution

Røvik, Kjell Arne, 1992. Den “syke” stat. Oslo: Universitetsforlaget. Røvik, Kjell Arne, 1996. ‘Deinstitutionalization and the Logic of Fashion.’ In Czarniawska, Barbara and Guje Sévon, (eds.) Translating Organizational Change: 139-172. Berlin: Walter de Gruyter. Røvik, Kjell Arne, 1998. Moderne organisasjoner: Trender i organisasjonstenkningen ved tusenårsskiftet. Bergen: Fagbokforlaget. Schaanning, Espen, 1996a. ‘Diskursens materialitet – Del I: Foucault.’ Idehistorisk tidsskrift1:1-23. Schaanning, Espen, 1996b. ‘Diskursens materialitet – Del II: Latour.’ Idehistorisk tidsskrift 2:1-21. Schaffer, Simon, 1991. ‘The Eighteenth Brumaire of Bruno Latour.’ Studies in History and Philosophy of Science 22(1): 174-192. Scott, W. Richard, 1987. ‘The Adolescence of Institutional Theory.’ Administrative Science Quarterly, 32: 493-511. Scott, W. Richard, 1991. ‘Unpacking Institutional Arguments.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 164-182. Chicago: University of Chicago Press. Scott, W. Richard, 1992a [1983]. ‘Health Care Organizations in the 1980s: The Convergence of Public and Professional Control Systems.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 99-114. Newbury Park: Sage. Scott, W. Richard, 1992b [1983]. ‘Reform Movements and Organizations: The Case of Aging.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 115-128. Newbury Park: Sage. Scott, W. Richard, 1992c. Organizations: Rational, Neutral, and Open Systems. London: Prentice-Hall. Scott, W. Richard, 1995. Institutions and Organizations. Thousand Oaks: Sage. Scott, W. Richard and John W. Meyer, 1991a. ‘The Rise if Training Programms in Firms and Agencies: An Institutional Perspective.’ Research in Organizational Behavior 13: 297-326. Scott, W. Richard and John W. Meyer, 1991b. ‘The Organization of Societal Sectors: Propositions and Early Evidence.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 108-140. Chicago: University of Chicago Press. Scott, W. Richard and John W. Meyer, 1992 [1983]. ‘The Organization of Societal Sectors.’ In John W. Meyer and W. Richard Scott (eds.) Organizational Environments: Ritual and Rationality: 129-153. Newbury Park: Sage. Scott, W. Richard and John W. Meyer, 1994a. Institutional Environments and Organizations: Structural Complexity and Individualism. Thousand Oaks: Sage. Scott, W. Richard and John W. Meyer, 1994b. ‘Environmental Linkages and Organizational Complexity: Public and Private Schools.’ In Scott, W. Richard and John W. Meyer (eds.) Institutional Environments and Organizations: Structural Complexity and Individualism: 137-159. Thousand Oaks: Sage. Seguin, Eve., 2000. ‘Bloor, Latour, and the Field.’ Studies in History and Philosophy of Science 31(3): 503-508. Selznick, Philip, 1957. Leadership in Organization: A Sociological Interpretation. Berkeley: University of California Press. Selznick, Philip, 1980 [1949]. TVA and the Grass Roots: A Study of Politics and Organization. Berkeley: University of California Press. 203

Holm

Serres, Michel, 1995 [1990]. The Natural Contract. Ann Arbor: The University of Michigan Press. Shapin, Steven, 1979. ‘The Politics of Observation: Cerebral Anatomy and Social Interests in the Edinburgh Phrenology Disputes. In R. Wallis (ed.) On the Margins of Science: The Social Construction of Rejected Knowledge: 139-178. Sociological Review Monograph 27. Keele: University of Keele. Shapin, Steven, 1982. ‘History of Science and its Sociological Reconstructions.’ History of Science 20: 157139. Shapin, Steven, 1988. ‘Following Scientists Around.’ Social Studies of Science 18: 533-50. Shapin, Steven, 1996. The Scientific Revolution. Chicago: Chicago University Press. Shapin, Steven and Simon Schaffer, 1985. Leviathan and the Air Pump: Hobbes, Boyle, and the Experimental Life. Princeton: Princeton University Press. Shapin, Steven and Simon Schaffer, 1999. ‘Response to Pinnick.’ Social Studies of Science 29(2): 249-253. Shepsle, Kenneth A., 1986. ‘Institutional Equilibrium and Equilibrium Institutions.’ In H. Weisburg (ed.) Political Science: The Science of Politics: 51-82. New York: Agathon. Simon, Herbert A., 1976 [1945]. Administrative Behavior. New York: Free Press. Sisimondo, Sergio, 1996. Science without Myth: On Constructions, Reality, and Social Knowledge. New York: State University of New York Press. Smith, Tim D, 1994. Scaling Fisheries: The Science of Measuring the Effects of Fishing, 1855-1955. Cambridge: Cambridge University Press. Sokal, Alan, 1996a. ‘Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity’ Social Text 46/47: 217-252. Sokal, Alan, 1996a. ‘A Physicist Experiments with Cultural Studies.’ Lingua Franca (May/June): 62-64. Sokal, Alan, and Jean Bricmont, 1998. Intellectual Impostures: Postmodern Philosophers’ Abuse of Science. London : Profile. Star, Susan Leigh, 1991. ‘Power, Technology and the Phenomenology of Conventions: On Being Allergic to Onions.’ In John Law (ed.) A Sociology of Monsters: Essays on Power, Technology and Domination: 26-56. London: Routledge. Stinchcombe, Arthur L., 1968. Constructing Social Theories. Chicago: The University of Chicago Press. Sturdy, Steve, 1991. ‘The Germs of a New Enlightenment.’ Studies in History and Philosophy of Science 22(1): 163-173. Sztompka, Piotr, 1996. R.K. Merton On Social Structure and Science: 267-276. Chicago: The University of Chicago Press. Thelen, Kathleene and Sven Steinmo, 1992. ‘Historical Institutionalism in Comparative Politics.’ In Sven Steinmo, Kathleen Thelen and Frank Longstreth (eds.) Structuring Politics: Historical Institutionalism in Comparative Analysis: 1-32. Cambridge: Cambridge University Press. Thomas, George M. and John W. Meyer, 1984. ‘The Expansion of the State.’ Annual Review of Sociology 10: 461-82.

204

The Invisible Revolution

Thomas, George M, John W. Meyer, Francisco Ramirez, and John Boli, (eds.) 1987. Institutional Structure: Constituting State, Society and the Individual. Newbury Park: Sage. Thompson, Michael, 1992. ‘The Dynamics of Cultural Theory and Their Implications for the Enterprise Culture.’ In Shaun Hargreaves Heap and Angus Ross (eds.) Understanding the Enterprise Culture Themes in the Work of Mary Douglas: 182-202. Edinburgh : Edinburgh University Press.

Tolbert, Pamela S., 1985. ‘Resource Dependence and Institutional Environments: Sources of Change in Institutions of Higher Education.’ Administrative Science Quarterly 30: 1-13. Tolbert, Pamela S. and Lynne G. Zucker, 1983. ‘Institutional Sources of Change in the Formal Structure of Organizations: The Diffusion of Civil Cervice Reform, 1880-1935.’ Administrative Science Quarterly 28: 22-39. Traweek, Sharon, 1988. Beamtimes and Lifetimes: The World of High Energy Physicists. Cambridge: Harvard University Press. Trigg, Roger, 1978. ‘The Sociology of Knowledge.’ Philosophy of the Social Sciences 8: 289-298. Vardø herredsrett, 1994. Utskrift av domsprotokoll fra rett holdt i Vardø herredsrett på sorenskriverkontoret i Vardø, 17. desember 1994, S487A 407.3. Weber, Max, 1968. Economy and Society: An Outline of Interpretive Sociology. Guenther Roth and Claus Wittich (eds.) Berkeley: University of California Press. Weick, Carl, 1969. The Social Psychology of Organizing. Reading: Addison-Wesley. Weick, Carl, 1976. ‘Educational Organizations as Loosely Coupled Systems.’ Administrative Science Quarterly 21:1-19. Wildawsky, Aaron, 1987. ‘Choosing Preferences by Constructing Institutions: A Cultural Theory of Preference Formation.’ American Political Science Review 81(1):1-21. Williamson, Oliver E., 1975. Markets and Hierarkies. New York: Free Press. Williamson, Oliver E., 1985. The Economic Institutions of Capitalism. New York: The Free Press. Wittgenstein, Ludwig, 1968. Philosophical investigations. Oxford : Blackwell. Whitley, Richard, 1992. ‘The Social Construction of Organizations and Markets: The Comparative Analysis of Business Receipts.’ In Michael Reed and Michael Hughes (eds.) Rethinking Organizations: New Directions in Organization Theory and Analysis: 120-143. London: Sage. Wrong, Dennis, 1961. ‘The Oversocialized Conception of Man in Modern Sociology.’ American Sociological Review 26(2): 183-193. Young, Oran R., 1986. International Regimes: Towards a New Theory of Institutions. World Politics 39:104-22. Zucker, Lynne G., 1983. ‘Organizations as Institutions.’ In S.B. Bacharach (ed.) Advances in Organizational Theory and Research: 1-43. Greenwich: JAI. Zucker, Lynne G. 1987. ‘Institutional Theories of Organization.’ Annual Review of Sociology 13: 443-64. Zucker, Lynne G., 1988. ‘Where Do Institutional Patterns Come from? Organizations as Actors in Social Systems.’ In Lynne G. Zucker (ed) Institutional Patterns and Organizations: Culture and Environment: 2349. Cambridge: Ballinger.

205

Holm

Zucker. Lynne, G., 1991 [1977]. ‘The Role of Institutionalization in Cultural Persistence.’ In Walter W. Powell and Paul J. DiMaggio (eds.) The New Institutionalism in Organizational Analysis: 83-107. Chicago: University of Chicago Press. Zukin, Sharon, and Paul J. DiMaggio, (eds.) 1990. Structures of Capital: The Social Organization of the Economy. Cambridge: Cambridge University Press. Åslund, Arnfinn, 1994. Fanden i nøtta: Innføring i litteraturteori. Oslo: Samlaget.

206