ISSUES POSITION PAPERS

BACK TO ISSUES POSITION PAPERS' TABLE OF CONTENTS

 

Realism and Constructivism in the Science of Ecological Knowledge

Petter Holm

Norwegian College of Fishery Science
University of Tromso
Tromso, Norway

"The Truth is Out There" (X-files)

Introduction

I suppose it is easy to agree that the science of TEK, LEK, IK, or whatever label you prefer (LEK in the following), is somewhat ambiguous when it comes to its own position on fundamental ontological and epistemological issues. In other words, locating typical exemplars of the science of LEK on an axis going from pure positivism, or realism on the one hand to radical relativism, idealism or social constructivism on the other hand is not as easy as one perhaps would think. Some authors identify LEK as part of a post-modern inclination, and thus as far out on the constructivist side:

The present interest in traditional folk knowledge or indigenous knowledge on nature can be seen as a reflection of post-modern scepticism towards the scientific approach to environmental conservation and management of natural resources. The idea of indigenous knowledge pictures a sort of non-scientific 'authentic' knowledge rooted in premodern and precolonial culture and tradition. (Eythorsson, 1988: 197).

Other authors point out that while scientists involved in the collection of LEK engage in rhetorics that suggest a constructivist position, their research practices bring them close to western (realist) science:

In their desire to find an elevated status for indigenous knowledge, they attempt to use the same instruments that western science uses. In so doing, they undermine their own assertions about the separability if indigenous from western knowledge in three ways: 1) They want to isolate, document, and store knowledge that gains its vigour as a result of being integrally linked with the lives of indigenous people; 2) they wish to freeze in time and space a fundamentally dynamic entity Ð cultural knowledge; and 3) most damning, their archives and knowledge centers privilege the scientific investigator, the scientific community, science, and bureaucratic procedures. (Agrawal, 1995: 428)

Perhaps the science of LEK combines elements of realism and constructivism? Wilson's characterization of the 'TEK model' suggests that this is the case:

The TEK model both incorporates, and is a challenge to, the deference model [i.e. one in which "Scientists are seen as the experts and the best way to get an accurate picture of nature is to rely on their professional judgement."]. No one claims that TEK produces generalizable, scientific information. The emphasis is on local information that can supplement the generalizable knowledge produced by mainstream science. In this sense the TEK model builds upon the deference model. However, the TEK model makes two important claims. The first is that traditional, local knowledge reflects the thinking of local people, including fishers, and therefore, is critical to producing a picture of nature that will be accepted as legitimate. The second claim is that the local knowledge, while not generalizable, is just as valid as that produced by formal science. From the perspective of science as a social process, this is its crucial difference with the deference model: the scientist is no longer assigned the role of the gatekeeper who has the final word on what is or is not known to be true about nature. In this model, the scientist now has one among perspective on what is true among others." (Wilson, 1999: 10)

While building on the 'deference model', in that the TEK model presupposes that LEK can supplement scientific knowledge, the TEK model at the same time challenges scientific knowledge by claiming that LEK is equally as valid as scientific knowledge.

I suppose it might be possible to establish a position somewhere between realism and constructivism. But at the moment, it seems that the argument for the validity of such a position is not very well developed. My suspicion here, which Wilson's characterization does very little to ease, is that the TEK model is trying to combine positions that remain contradictory. How can local knowledge, if it is different yet equally valid as scientific knowledge, supplement the latter? While this may be possible when seen from a relativist position, it would seem very difficult from the normal (positivist or realist) scientific position, in which there can only be one truth about nature. In this paper, I will try to contribute to a discussion about the different perspectives one can take in studying different types of Ecological Knowledges (EK), whether scientific (ScEK) or local (LEK). (.... Unfinished....)

Realism and Constructivism

In order to construct methodologies for the study of EK, whether this is local (LEK) or scientific (ScEK), we first need to make up our mind about the epistemological status we want to ascribe to such knowledge. In other words, are we going to take such knowledge claims literally, as true or false statements about the world, or are we going to study them as facts in themselves, regardless off their truth content, as clues to the culture, values, and world views of the society to which they belong? In order to see the relevance of this question, we can mobilize the conceptual pair of realism on the one hand and idealism or social constructivism on the other. Realism is based on the assumption that there is a real world out there and that it is - in principle and in practice Ð possible to obtain true knowledge about that world. If you apply the proper methods and procedures (science and perhaps science-like procedures) true knowledge about the world can be obtained. Such knowledge will be true to the extent it corresponds to reality. By this view, science, when done correctly, mobilizes nature to act as a judge, sorting true from false knowledge claims.

On the other hand, idealism or social constructivism, holds that nature, the real world, while it may be out there, can never be made to intervene in human affairs. It will always be possible to interpret nature in many different ways. Nature itself is a blank screen, which will reflect everything society projects onto it. Since nature is indifferent and refuses to be a judge about different theories about it, it will always be social factors Ð culture, power or interests Ð that enter the picture when groups of people decide to hold one interpretation of nature as the true one instead of an another interpretation. By this view, all knowledge is socially constructed. Nature is too soft and malleable to decide the truth of knowledge claims. Only Society can do the job.

Now, let us see how this pair of concepts can be used in order to sort out different approaches to ScEK and LEK.

First Position: ScEK is realist, LEK is constructed

First it should be noted that the conventional scientific approaches to the study of nature (ScEK) are predominantly realist. Consider, for instance, the position of science in the conflict over modern whaling in Norway towards the end of the 19th and the beginning of the 20th century. As you perhaps will know, modern whaling technology, which opened up the large stocks of fin whales for commercial exploitation, was invented by a Norwegian, Svend Foyn. The area where this technology was developed was the coast of Northern Norway. Towards the end of the 19th century, a conflict over whaling developed here, between the local fisher population, and the whaling companies, which came from southern Norway. In the center of this conflict was a LEK claim from the fishers. They belived that the cod fishery was dependent on healthy whale stocks. The whales would chase the cod's bait, herring and caplin, towards the coast. The cod would follow. Without whales, the caplin and herring would stay out at sea. So would the cod, which then would be out of reach for the fishers' small coastal vessels. The fishers believed that the whaling operations threatened the whale stocks, and thus the fishery.

So, a nice ecological theory - let us call it, following Ethorsson (1998), the 'sheepdog theory' - held by the fishers. Because the intensity of the fisher mobilization (this conflict led to the first member of the Labour Party in being elected to the Norwegian Parliament) on the one hand, and the enormous presige of the whaling industry (one of Norway's few contribution to the industrial revolution) on the other, the Government intervened. How? By appointing a natural scientist to find out whether the sheepdog theory was true or false. Enter Johan Hjort. What did he do? (... details ..). So, after checking out the facts of the matter, with the methods and resources available to him, he concluded that the sheepdog theory was false. While the whale could be important in order for the fishers to locate the fish, it was not important in order to bring the fish to shore.

So, here we see realism at work. Nature is believed to be out there. When a knowledge claim is launched, its truth can be checked by finding out whether it corresponds to nature or not. Now, the point here is that this approach can be found in much of the contemporary approach to LEK. LEK can generate interesting hypotheses, which then can be checked out, verified or falsified, by science. (... fill out examples, e.g. Johannes (1993) etc). If something in this approach is different from what Johan Hjort Ð or a number of other natural scientists like him Ð have done, is the more systematic effort to collect instances of LEK hypotheses. This is an important difference, of course. But when it comes to the fundamental position as to the different statuses assigned to the knowledge claims involved Ð LEK must be processed by science before it can be judged as false or true Ð the approach is very much the same. (This is important to say only because of the impression created that the LEK studies are radically different from main-stream science.)

If the above is true about the LEK studies, they are just a variety of normal ScEK; one that will apply a different methodology (consult the people, not the fish!) as a starting point for their research projects. But this does not seem to cover the whole story, since it perhaps could be argued that LEK studies represent a variety of social science. In order to pursue this, we could start asking what would be the characteristic mode of investigation of a LEK knowledge claim by a social scientist. Imagine, for instance, what a social scientist typically will have to say about the fishers' sheepdog theory in the case of Norwegian whaling. For a social scientist, I suggest, the most typical angle here is not to focus on whether the fishers sheepdog theory is true or false. Instead s/he would inquire Ð particularly if natural science proved the sheepdog theory was wrong Ð why the fisher would stick to this belief. Note here that the focus is moved here from the truth content of the knowledge (although it adds interest if the knowledge happens to be wrong), to the knowledge as a product of a different process. So Eythorss—n (1998: 199) concludes:

"In the whaling dispute, by the turn of the century, the strength of the fishermen's knowledge was not primarily in its substance but in the political power placed behind it. As the protests of whaling took on a form of class unrest in a multi-ethnic area where the loyalty of the population to the Norwegian state was considered somewhat uncertain the Government was forced to listen."

The social scientist typically will look beyond the substance of the knowledge, whether it corresponds to some feature of reality or not, and instead look at to the interests behind it, how it relates to class struggle; how it is used in a broader context of political and social mobilization. In other words, the social scientist will presume that the LEK claims are socially constructed. If s/he can understand the culture, the social processes, the politics of the groups involved, s/he will also be able to understand the group's knowledge claims. It is important to note here, despite the derogative tone social scientists often will apply towards natural scientists, that this is very much the same perspective as the natural scientist will apply. For the natural scientists, the main thing is whether the sheepdog theory is true or false. Nevertheless, when Hjort concluded that it was wrong, he was quick to move towards an analysis of why the fishers would stick to a false theory. His suggestions here was, as also Eythorsson suggests, that it had to do with the poverty of the fishers in relation to the wealth of the whalers, the fishers problems in switching from coastal to sea-going vessels, the political organization of the campaign, which turned the whaling issue from one that could be settled by science into question of ideology, etc, etc. In short, he turned himself into a sociologist/political scientists in order to answer this question. Neither was Hjort reluctant to talk to the fishers; he asked them about their theories, held meetings with them.

ScEK reserves realism for its own knowledge claims, while it assumes that LEK is socially constructed. When checking whether a given knowledge claim is true or false, scientists are realists. When checking why local peoples believe what they believe (particularly when they differ from science), scientists turn to social constructivism.

We should note here that social science's constructivist perspective on cultural beliefs is one of the discipline's fundamentals. Without it, it would be quite impossible, or at least somewhat meaningless, to engage in a sociological analysis of, say, religion. If the analysis are made dependent on the truth content of peoples beliefs Ð whether God exists or not Ð all the analytical tools of the sociologist are immediately in jeopardy. On the contrary, the basic idea of such an analysis is that God will not intervene, and that such beliefs can be completely accounted for by reference to social factors. God must be socially constructed for a sociology of religion to be possible. In general, suspending the idea that people's beliefs actually corresponds to, and perhaps are caused by, reality, is of great importance to social science. Without this trick of the trade, social scientists would have great difficulties in analyzing things like tastes and fashions, why some people would believe in white supremacy and the like, why people hold that women are different from men, or why economists and others tend to believe in the rationality and efficiency of ITQs.

The point here is that there is almost complete compatibility between the classical natural science inquiry into LEK claims (true or false?), and the social science inquiry into such claims (why do they believe what they believe, regardless of the truth content of this belief?). So, the first, completely plausible and defensible version of LEK studies, which will be applicable from both natural and social science alike, will be one that reserves realism (the truth is out there) for ScEK, and holds LEK to a constructivist position (the local's knowledge claims are socially constructed). My feeling, although I have yet to do a thorough literature review to substantiate this claim, that a great deal of contemporary LEK studies fit into this position. The good thing, of course, is that such a standpoint is well founded in main-stream science; and that it allows complementary roles for natural and social scientists. However, it may be appropriate to borrow Wilson's term of a "deference model" to characterize this position, for two reasons. First, LEK claims and their holders, while in focus, is relegated to a deferential position in relation to ScEK and scientists. (.. illustration needed ...?). Second, social science and social scientific knowledge take on a deferential role in comparison to ScEK and natural scientists. This is because ScEK knowledge remains the center around which the whole enterprise revolves. Besides contributing to the methodology of collecting LEK Ð social scientists are reputed for being better to talk to people Ð social scientists' role here is mainly to account for mistaken LEK. This is the same role as sociology will have in Mertonian sociology of science, where sociological insights are only called for in explaining the content of scientific knowledge when this knowledge goes wrong. True scientific knowledge has no sociological cause; false knowledge can only be explained sociologically.

Perhaps the deferential role for LEK and social science in this position is the explanation of the tendency in the literature to over-communicate the deficiencies and weaknesses of natural science, as well as the contribution and insights offered by LEK? In the end, it is tempting to say, the whole enterprise of LEK studies in this variety is to "deconstruct and reconstruct" LEK, that is, rip it to pieces, pick out the good bits and rebuild it according to ScEK standards. In order to be able to do that without loosing all legitimacy with the main suppliers of LEK (and accommodate social scientists inferiority complex towards natural scientists?), it would seem a good and necessary rhetorical strategy to talk endlessly about the deficiencies of ScEK and the richness, holism, truthfulness, usefulness, ...etc, of LEK.

Second Position: ScEK is socially constructed, LEK is realist

Some of the LEK studies read as if LEK claims should be taken literally, as true or potentially true statements about the world, whereas ScEK should be regarded as socially constructed and hence not sensitive to the true workings of nature. Here, the normal rank-order of ScEK/LEK is turned on its head. For instance, in today's local paper (Nordlys June 7, 2000 editorial), I read, in a comment on the cod crisis in the Barents sea: ... "that the ministry actually was warned Ð by the fishers. But the inherited knowledge of the fishers, who used to take one and one fish from the sea, harvested it like a peasant harvests his herd, has been pushed out by a science that definitively has demonstrated its shortcomings."

It is very easy to imagine that sometimes, on some issues, LEK will turn out to be true, while ScEK will turn out to be false. But could we argue that this usually is the case, and hence, that a realist position should be reserved for LEK, while we should approach ScEK from a constructivist angle? That we can use LEK as the standard from which to judge the truth of ScEK knowledge claims?

I must admit that I cannot see any sound analytical reasons for taking such a position. If there are procedures within LEK that allows access to the truth about nature, it must be possible for science to follow the same procedures. My conclusion here is that the reverse constructivist/realist position is not stable, and that it at the most can be used to score cheap rhetorical points. But that, as we know, is sometimes important enough. (... whole section needs to be deconstructed and then reconstructed ...)

Third Position: ScEK and LEK are both socially constructed

From this perspective, ScEK and LEK are viewed symmetrically, as two (or rather, a multitude of) different local knowledge constructions, none of which has privileged access to the truth about nature. Now, the problem here is not to see LEK in this way, since this conforms to the conventional perspective social science takes on local cultures. The problem is to apply this perspective on ScEK. While problematic for many reasons, which I will not go into here, it is of course not impossible. Social constructivist sociology of science does these things on a routine basis. In the fisheries, Finlayson's analysis of the Canadian cod crisis is one example (Finlayson, 1994).

Such perspectives on ScEK are an important resource for LEK studies, since the delegitimation of ScEK, which is an important outcome of such exercises, is crucial to make space for LEK. However, the symmetrical position of ScEK and LEK as socially constructed, without any truth to serve as arbiter between them, turns the whole business of studying LEK into a very different exercise than it is in position 1. Why would one want to collect LEK if it does not tell you anything else than local truths? In this version, then, the focus shifts from nature back towards society. Science is important not because it is true, but because it has so much power behind it. LEK is usually disregarded, not because it is less true than ScEK, but because it has less power behind it. Perhaps it could be argued that collecting LEK will serve to empower it. See Maurstad (2000) for a discussion on this. And as Agrawal has argued:

Ex situ conservation, even if it is successful in unearthing useful information, is likely to benefit the richer, more powerful constituencies Ð those who possess access to international centers of knowledge preservation Ð thus undermining the major stated objectives of the neo-indegenistas Ð to benefit the poor, the oppressed, and the disadvantaged. Withess the appropriation of ethnobotanical knowledge by pharmaceutical and biotechnology multinationals. Knowledge freely available to all does not benefit all equally. (Agrawal, 1995: 432)

What is needed is not to collect LEK, but to mobilize power behind the communities that hold LEK. The whole thing turns into politics. I think that Agrawal's paper (and perhaps also Eythorsson's?) is an example of the constructivist/constructivist (or symmetrically relativist) position:

The appropriate response from those who are interested in preserving the diversity of different knowledges lie in attempting to reorient and reverse state policies and market forces to permit members of threatened populations to determine their own future, and attempt, thus, to facilitate in situ preservation of indigenous knowledges, In situ preservation cannot succeed without indigenous populations gaining control over the use of lands in which they dwell and the resources on which they rely. Those who are seen to possess knowledge must also possess the right to decide on how to save their knowledge , how to use it, and who shall use it. (Agrawal, 1995: 432)

Forth Position: ScEK is realist, LEK is realist

The final possible position of realist/constructivist perspectives would be to regard both ScEK and LEK as procedures that have the potential for producing truth about nature. The problem now, since we do not have one knowledge system that is superior, is to arrive at some way to make the judgement as to when each of them will produce the truth. One criterion here would be that when ScEK and LEK produces the same results, this wold probably be the truth. However, how shall we proceed when ScEK and LEK differ, or on issues where just one of them forward knowledge claims?

Here we could turn to Merton and the idea of a scientific "ethos" (Merton, 1996). When scientists adhere to this "ethos", truth will be produced. When this ethos for some reason or other is perverted and external factors such as interests and politics are allowed to enter into science (think of the Lysenko affair) science's truth-producing capabilities are threatened. We could also imagine that a similar normative and organizational system to be constructed for LEK knowledge production. If LEK knowledge is produced according to this system, it would then tend to be true. If the system for some reasons is distorted, it will tend to produce false knowledge.

This may perhaps seem to be somewhat far-fetched. And it is indeed difficult to find any examples of authors having taken such a position. But Felt's (1994) paper on the indigenous knowledge among Atlantic Canadian salmon fishers, although it comes very close to position 1 (ScEK is realist, LEK is constructed), can perhaps be read as an approximation.

The observation Felt makes is that different groups of salmon fishers hold different knowledge about the status of the salmon stock. In other words, some fishers believe that the salmon stock is severely depleted. Others believe that the salmon stock is in good shape. While this is unproblematic from a constructivist or relativist position, which is not be very interested in the truth content of knowledge claims, it is very much so when one believes Ð as Felt does - that indigenous knowledge has relevance and utility. The challenge, then, since different local knowledges differ and contradict each other, is to find some ways to sort the good knowledge from the bad:

In this chapter I shall argue that the lack of consensus, and even contradiction, among the fishers need not negate the relevance and utility of their knowledge for resource management. The key to successful utilization, however, lies in first understanding the processes and context within which local knowledge is produced. In contemporary language, the knowledge must be deconstructed and then reconstructed (Pinch 1986) for potential management use. Fishers' knowledge, in other words, must be primarily understood as a social construction in which particular experiences are given meaning within a specific cultural context. As such, the process whereby something comes to be known assume at least as much importance as the knowledge claim itself. (Felt, 1994: 253)

This sounds of course a lot like a constructivist program on LEK. Furthermore, the reference, Pinch, 1986, is a social constructivist text within the sociology of science. However, there is an important difference in the way Felt proposes to sort out good LEK from bad LEK from that proposed from within position 1. From the standard realist/constructivist position, LEK claims serves as a source of hypotheses, which must be appropriately tortured and interrogated before some of it can be accepted as true knowledge. What Felt seems to suggest, is that we can sort out the good from the bad LEK by checking out the social process by which it has been produced. Instead of looking whether the resulting knowledge claims correspond to nature or not, as you would do (at least in principle) from position 1, you will look into the processes by which the knowledge claims were constructed:

Perhaps the overall message of this analysis is that enumerating and codifying such knowledge is not enough. Before it can be fully utilized for sustainable management, the social conditions and constraints under which it is produced must also be appreciated. (Felt, 1994: 282)

In practical terms, Felt finds that the differences in beliefs about the status of the salmon stock vary systematically according to the fishers' geographical location in relation to the stock. The fishers that believed that the salmon stock was healthy (Profile 1), were located in zones where the salmon was available for harvest for only a short time each year (2-3 weeks). This would result in limited knowledge of the fish's life cycle. Further, because of the large number of salmon migrating past this location, they would perhaps notice any reductions in their number. The fishers that believed that the stocks are threatened (Profile 2), typically live closer to the salmon rivers:

The geographical proximity of salmon river and fishers allows for more comprehensive, integrated understanding of salmon. Rather than a fleeting encounter as they pass by, fishers encounter the salmon at various stages of their life cycle. (Felt, 1994: 267)

As this analysis suggests, Felt has arrived at the conclusion that the Profile 1 belief (salmon is healthy) is false, while the profile 2 belief (salmon is threatened) is true. There is a problem here, since Felt does not give his reason for this conclusion. If he at this point trusts natural scientists, we are back in position 1. However, if he follows his own recommendation, and put his trust in the Profile 2 people because of their better position for collecting this type of knowledge, we may still be in the realist/realist position.

But Felt has an additional comment to make; one that allows me to get back to Merton and the ethos of science. First, Felt notes the strength in the denial of the possibility of stock decline among the profile 1 fishers. Then:

The explanation of such adamancy and denial of any stock decline amongst these fishers lies in the relationship between their partial, instrumental knowledge, on the one hand, and the consequences of certain important external factors on the other. Two general forms of external factors are critical: (1) changes in state fishery regulations and (2) participation in the union representing fishers in the province. In combination with an instrumental understanding, they lead fishers to reject claims of stock decline and to perceive salmon management as almost exclusively a political battle. For fishers possessing the more holistic view presented in Profile 2, these factors either have minimal consequences or are absent completely. (p 269)

Profile 1 fishers, to cut the story short (and somewhat simplified) have been more directly hit by the regulations (they are not disinterested), and have been more directly involved in NFFAW, which have engaged itself in a battle against angler interests (they are political agents). In short, profile 1 knowledge have been distorted by interests and politics, while profile 2 knowledge remain pure.

As you perhaps will remember, disinterestedness is one of the norms in Merton's "ethos of science". As I read Felt's paper, it contains the rudiments of an equivalent model for LEK. There are some requirements as to the frequency and intensity of the fisher's encounters with the object for which knowledge is sought. Distortions in the form of economic interests and political mobilization must be absent. There are of course also important differences here. Merton's ethos takes the form as technical and normative rules; they are believed in and made binding for scientists by the community of scientists. Since the fishers are not involved in the business of knowledge production as an overall goal, you will not find a comparable ethos of LEK among fisher. So, establishing an explicit model that predicts when LEK can produce true knowledge and when it will produce distorted knowledge is still some way ahead. But there certainly are ideas, in Felt's paper and elsewhere, that could be used here. One such idea, complementing Felt's idea of geographical proximity and frequency and intensity in the encounters with the fish, is that the fishing technology plays an important role for knowledge generation (Eikeland, 1998). The fishing gear can be viewed as an instrument of observation. Hence, trawler fishers will see different things in nature than gill-net and long-line fishers. (... more after a search of the literature ....).

It is of course not irrelevant here that, in contrast to position 1, which I characterized as the "deference model", the realist/realist position discussed here allows a much more gratifying role for social scientists. Instead of assisting natural scientists with the collection of raw LEK, and accounting for mistaken LEK, the realist/realist position grants to social science the role of certifying LEK knowledge claims on its own. As we see from Felt's paper, the social scientists can here establish themselves in what could be labeled an Obligatory Passage Point (OPP) between local communities carrying LEK, some good and some bad, on the one hand and fisheries resource managers in need of only good LEK on the other. The only way (that is if consulting nature directly is too expensive and carries too little legitimacy) for resource managers to gain access to the good LEK without getting the bad LEK in the same package, is to go through the social scientists, who has the instruments and methods to deconstruct raw LEK and then reconstruct good LEK out of the bits and pieces.

Concluding remarks

What can we learn from this? Well, I do think that the positions we take on LEK/ScEK pair have methodological implications. This is fairly obvious, I think, in the contrast between position 1 (realist/constructivist) and 3 (symmetrically constructivist). With position 1, the whole point is to collect LEK and test whether it holds true. In position 3, the purpose of the LEK exercise, which probably will not even take the form of collecting TEK, is political empowerment. Here, the value of LEK remains a matter of faith since its treatment as hypotheses that may be wrong, will be taken as an assault on the autonomy and cultural worth of its holders. The contrast between position 1 and position 4 (realist/realist), besides the implication for the role of social science, is also one of methodology. Whereas position 1 does not require much except good interview technique and survey methodology, position 4 is quite challenging since it requires construction of a model of LEK knowledge production that can be used for sorting good from bad LEK, not by its results, but by its process of construction. (... a bit more development here, perhaps ...)

As I hope to have shown in the preceding, it may be useful to distinguish between the different realist/constructivist combinations in the approaches to LEK. It might not be a good idea to try to construct methodologies for LEK studies as if these are of one kind only.

The ScEK/LEK literature is characterized by frequent shifts between realist and constructivist repertoires. To some extent, this is quite appropriate, since different knowledge claims must be treated in different ways. However, some of these shifts seem to be grounded on rhetorical rather than analytical reasons, so that a realist position is reserved for oneself and ones allies (what I/we claim is true), while the knowledge claims of ones adversary are taken as social constructions (your claims must be explained by reference to interests and politics). When such tactics are mingled in with the legitimate, but usually implicit, shifts between realist and constructivist repertoires for analytical reasons, the resulting mix becomes quite unsound.

References

Agrawal, Arun (1995) Dismantling the Divide Between Indigenous and Scientific Knowledge. Development and Change 26: 413-439.

Eikeland, Sveinung (1998) Flexibility in the Fishing Commons. In Svein Jentoft (ed) Commons in a Cold Climate, 97-114. Paris/New York: UNESCO/Parthenon.

Eythorsson, Einar (1998) Voices of the Weak Ð Relational Aspects of Local Ecological Knowledge in the Fisheries. In Svein Jentoft (ed.) Commons in a Cold Climate, 185-204. Paris/New York: UNESCO/Parthenon.

Felt, Lawrence F. (1994) Two Tales of a Fish: The Social Construction of Indigenous Knowledge Among Atlantic Canadian Salmon Fishers. In Christopher L Dyer and James R. McGoodwin (eds.) Folk Management in the World's Fisheries, 251-286. Niwot: university Press of Colorado.

Finlayson, Alan Christopher (1994) Fishing For Truth. St. John's: ISER.

Johannes, R.E. (1993) Integrating Traditional Ecological Knowledge and management with Environmental Impact Assessment. In Julian T. Inglis (ed) Traditional Ecological Knowledge: Concepts and Cases, 33-40. Ottawa: International Program on Traditional Ecological Knowledge and International Development Research Centre.

Maurstad, Anita (2000) "Trapped in Biology" Ð An Interdisciplinary Attempt to Integrate Fishers' Knowledge into Norwegian Fisheries Management. In In B. Neis and L. Felt (ed.): Finding Our Sea Legs: Linking Fishery People and their Knowledge with Science and Management. St.John's: ISER.

Merton, Robert K. (1996) The Ethos of Science. In Piotr Sztumpka (ed) Robert K Merton: On Social Structure and Science, 267-276. Chicago: The University of Chicago Press.

Pinch, Trevor (1986) Confronting Nature: The Sociology of Solar-Nutrino Detection. Dordrecht: D. Reidel.

Wilson, Douglas C. (1999) Fisheries Science Collaborations: The Critical Role of the Community. Paper presented at the Conference on Holistic Management and the Role of Fisheries and Mariculture in the Coastal Community, November 11-11, 1999, at TjŠrnš Marine Biological Laboratory, Sweden.