Culture and Holism in Ethnography

BACK to Home page to Writing index Posted: 24 August 01 Last edited: 12 July 2006

Culture and Holism in Ethnography and Exogenous

Factors in Economics

An Approach to the Meaning of Ethnography in relation to Economic Theory

CYRIL BELSHAW

NOTE 2001: Although written in 1986, the article was not published until 1992 in Archivio per l'Antropologia e la Etnologia

PAROLE CHIAVE: Antropologia economica, variabili esogene, olismo, scambi sociali, etnografie, razionalità, tautologia, falsificazione.

RIASSUNTO II lavoro aifronta Ia questione. come puô l’etnografia rapportarsi all’analisi di variabili esogene? Lo studio sostiene l’opportunita di incorporare l’etnografia nei modelli centrali dell’economia, suggerendo che le etnografie costituiscono ciò che gli economisti chiamano elenchi di preferenze. L’etnografia ribadisce che i pararnetri di scelta prendono in considerazjone costi non materiali.

Una volta che questo è fatto, la via è aperta per la necessaria incorporazione. II carattere tautologico del modello di razionalità degli economisti non è in opposizione a questa procedura.

Si sostiene anche che ciò puo essere ancora utile per previsioni pratiche e teoretiche.

KEY WORDS: Economic anthropology, exogenous variables, holism, methodology, interdisciplinarity, choice, social exchange, preferences, ethnographies, rationality, ends and means, tautology, falsification.

SUMMARY — The paper addresses the issue, how can ethnography relate to the analysis of exogenous variab1es? It argues for the incorporation of ethnography into the central models of economics, suggesting that ethnographies constitute what economists call schedules of preferences. Ethnography insists that the parameters of choice take into account non-material objectives, and that decisions also take into account non-material costs. Once this is done, the way is open for the requisite incorporation. The tautological character of the economists’ model of rationality is not in opposition to this procedure. It is also argued that it may still be useful for practical forecasting and theoretical prediction.

In 1986 Prof. Ulf Himmelstrand, the Swedish sociologist, organized a symposium, on behalf of the International Social Science Council and Unesco, to explore an interdisciplinary approach to the treatment of what economists call <<exogenous variables>>. The meeting consisted fifty per cent of economists and fifty per cent of representatives of other disciplines. The results have not been published.

The present paper consists of a slightly amended version of my contribution. In writing it I came to realize that, for the most part, we as anthropologists take for granted the economists’ image of where our discipline lies, and often accept that as truth, without questioning the fundamental theoretical concepts. Although my task was to talk to economists, I found myself re-affirming and re-thinking the consequences of a postulate I have held throughout my academic career, namely that there is a philosophical and theoretical unity between the core ideas of economics (as distinct from some of the contemporary elaborations and compromises with that core) and the operations of ethnography (as distinct from its contemporary theoretical apparatus).. Thus I found myself writing about matters that I feel need re-thinking in anthropology as much as they do in economics. And it is for this reason that I hope the arguments I now present will have some future value to colleagues in my own discipline, as well perhaps as in economics.

Ulf Himmelstrand, in his summing up of the meeting, inferred that I was accepting a great deal of what has come to be known as the <<imperialist>> position of economics. This is not the case. What I do accept is the desirability of using some of the earlier theoretical ideas of economics. But then I depart from conventional economics almost completely, and most certainly do not accept its version of the world around it or even within it. And in accepting the earlier versions of these concepts I argue for their transformation in a way that is not usual within economics itself or, I may say, anthropology. There have been some brave attempts by a very few economists to extend their concepts into areas covered by other social sciences in a non-imperial manner, but, at least insofar as anthropology-ethnography are concerned, without making use of what we have to offer in a positive way. That is partly our fault, since we ourselves have done little to show what our observations have to say in a framework that permits theoretical dialogue.

The present essay is an attempt to hint at possibilities.

In a background paper contained in a Newsletter to the participants, Ulf Himmelstrand grappled with the meaning of <<exogenous variables>> in economics. In essence they are those variables which lie outside the basic self-contained formal model. Since such models, though basically similar in contemporary economics, are not completely unified, there is room for differences of opinion as to what is exogenous and what is not. In practice, as Himmelstrand points out, the term frequently refers to factors which must be considered as given (my italics) in a certain context of economic decision-making, regardless of whether it is an economic or non-economic factor, though in this context we are referring mainly to so-called <<non-economic>> factors. We could think of family structure, ethos, religion, power, class structure, propensity to innovate, the environment — indeed almost everything that has cultural and environmental reality in determining choices and decisions. In economic model-building it is so easy when all that is regarded as external to the model.

So I set out to try to give one answer, among many possible ones, to the question, what can anthropology contribute that will cast light on the nature of exogenous variables, as conceived in economics?

To take the question literally would mean accepting the validity of the analysis which posits the usefulness of systems which can logically function only if there are exogenous factors not merely as a temporary accommodation to the limits of observation, but as an affirmation that such systems may be seen to work well.

From the perspective of anthropology, such a position has to be fundamentally flawed with respect to adequate models of reality, and, I would argue, is inconsistent with the classical axioms of economics, which are still fundamental to it. Hence I partially transform the question by asking the question, can the models be increased in explanatory power if the notion of exogenous variables is discarded, or at least substantially modified?

First, some background explanation. Within social and cultural anthropology and even more within conservative ethnography, abstract model building of the kind economists perform is extremely rare. It is overdue, even with the branch known as economic anthropology. Although we deal with social relations and social structure and organization, anthropology derives its basic concerns from the overriding importance of cultural systems which permeate all thought and decisions, and are essential to the definition of cultures and peoples who create variations of economic behaviour.

I believe that this is, or can be, done in a manner which can be related rather precisely to at least the earlier axioms of micro-economics. And despite some field work choices, we are concerned with humanity in all its expressions, not limiting ourselves, in the popular and economists’ view of us, to the apparently exotic. (The economists present at the meeting were totally surprised to find that anthropologists had anything to say about modern health delivery systems, industrial workers’ attitudes and cultures, modern urban systems, city planning, or anything of that ilk. Fortunately I believe this perception is changing amongst some.)

Most anthropologists, without thinking about it, subscribe to the idea that the discipline is holistic (sometimes spelling the term <<wholistic>>). Major theorists have suggested that beyond any macro-culture there is a level of principle which can be used to explain a cultural system in terms which go beyond its parts (cf. the <<superorganic>> of A.L. Kroeber (1917), some versions of the <<value systems>> of Clyde Kluckhohn (1953), and now the structuralism of Claude Levi-Strauss (1963).

Such constructions are far from commanding total acceptance. The bulk of anthropology is not holistic in that sense, but deals rather with the functioning of parts of the system within the systemic whole. While constructs such as society, culture, polity, institution, even economy, are not in fact organic as are biological organisms — there is for humans only one self-contained system, namely the global society, and even that can be debated — yet it sometimes suits our purposes to use an organic analogy, and to think of the parts as necessary to the total system. Although our language is very different, this is not far removed in principle from the mechanical systems of economists.

Also, holism as used in anthropology perhaps unthinkingly implies a different dimension, namely that the only real system embraces global society in all its aspects, which ultimately must be understood through the use of massive data (including <<thick>>, or highly detailed and evocative, ethnography) and insights drawn from all disciplines. Since such a task is in reality clearly impossible, abstracting mini-systems out of the maximum, and the selection of a few interdisciplinary themes out of the infinity, is justified — and there are as many ways of doing this as there are anthropologists.

Anthropology differs from mainstream economics in that we do not concentrate on one limited way of doing this, and (although we have much dogma) it is usually evident to us that such partial systems cannot give complete results, even within their terms of reference. What is left out can be as significant as what is put in, and, contrary to much of economics, we welcome the additional perspectives and linkages provided by those who are working with an adjacent dimension.

It is with such attitudes that anthropologists approach the kinds of issues that are embodied in the idea of exogenous variables. The idea implies that there is more at stake than heuristic expediency; the core of the model has a special primacy, a key significance, for enlightenment and additions to knowledge. It would be extremely difficult for anthropology to accept this, since it appears that the conclusions drawn from the model, whether explanatory or practical, are bound to be false. The only chance of correcting them appears to be to incorporate the exogenous variables into the core model (leaving aside the question as to whether the model itself should be replaced, a question to be determined by results and criticisms).

Granted, this would reduce such mathematical power as the models seem to hold, would reduce some of the elegance, and make the life of analysis much more complicated. But who can pretend that social life is not complicated? And is it not out of complications that the over-riding themes and principles and theories emerge, as we search for the ultimate simplicities? And if mathematically inclined economists were to take the anthropological perspectives more seriously, treating them as challenges to the improvement of the models rather than as curious nuisances, it might well be that the game of mathematicization would not be as hopeless as at first sight it seems to be (cf., for example, Becker, 1981; Belshaw, 1959, 1969, 1978; Blau, 1964). Anthropologists could certainly learn from the attempt.

I should give some grounds for making such brash assertions.

The notion of choice is central to neo-classical economics. As I have mentioned, a few anthropologists, and the now somewhat discredited national character school (e.g. Gorer and Rickman, 1949), have attempted to seek super-cultural principles which would constitute what the sociologist Talcott Parsons (1935, 1937) called normative orientations which govern the way preferences are chosen and means are used to satisfy them. If such a search were successful, and most anthropologists believe that it is something more than an intellectual game producing stimulating insights, it would alter the whole paradigm. All exogenous factors, translated into behaviour, could then be predicted on the basis of their immediately preceding state and their linkage to known superior forces. Nothing like this has happened, or is likely, even with the intrusion of biological and genetic knowledge into the considerations. Man and his variations are too complex, there is strong evidence that any posited principles change in given cultures over time in such significant ways that, while continuity may be there, [are] less than mechanistic.

On the other hand we do, at a lower level of abstraction, speak of cultural systems and social structures, and make sense of their dynamic forms. Depending on analytical circumstances, these can be presented as constraints on choice, in that, through various processes such as socialization, limitations of perceptions, strong moral valuation, and many more, they put boundaries around the supposed infinity of wants and around the definition of resources and means. The latter are not so much technologically given, but are defined as perceptions within the culture. What is a resource to one people is unknown or irrelevant to another, whether or not it is physically present.

Within such boundaries, choice is wide open: society is free to change or not to change, to create new religions or stay with the old. And by the exercise of individual choice within society and culture, aggregated and ramified, social systems and cultures themselves change and adapt, sometimes through external, i.e. foreign or environmental, pressures and influences, which are interpreted through the same processes, and just as much through their internal activities, tensions, resolutions.

Although many ethnographic descriptions read statically, all contain the data which anthropologists and economists can tease out to reveal the exercise of choice.

The fact of choice is even more central to anthropology than to economics, since the totality of the subject deals with its empirical manifestations. The major stream of analysis originating with the work of Marcel Mauss (1925) on the gift and its prestatory status, would derive social structure in any society at any given time from the cumulation of individual choices. (There is an echo here of the perspective of Radcliffe-Brown, that social structure is a statistical phenomenon). This has led to a grouping of theoretical stances variously labelled <<social exchange theory>>. <<Transactional analysis>> and <<social network theory>> (e.g. Bailey, 1969; Barth, 1966, Blau, 1964; Kapferer, 1976; Orans, 1968; Ortiz and Howard, 1971).

Individual choice extends to the choice of persons with whom to interact, perhaps within defined categories (but you can always choose not to interact if you are willing to pay the price of non-interaction). The interaction includes the flow of symbolic messages. material things, emotions, and what we would call services. Of course choices are influenced by culturally emphasized preferences, including many the conventional economist leaves out of consideration, such as salvation, prestige (Veblen should be re-read), love of family, beauty, aggression, habit versus innovation, and all the others you can think of. In turn, the choices result in a reaffirmation or a change in the culture.

Such analysis, despite limited manpower directed to its extension, has been used for such diverse subjects as kinship and family patterns (Homans and Schneider, 1955), ceremony and the social order (numerous deriving from Mauss), political parties in Britain (Bailey, 1969), leadership and power (most abstractly by Blau, 1964; Bailey, 1971), the social organization of the Swat Pathan (Barth, 1959 a and b), fishing entrepreneurs (Barth, 1963 and his followers), regularities in peasant markets (Mintz, 1959, 1961), pre-monetary exchange systems (Belshaw, 1965), Mediterranean feuding (Bailey, 1971), to name but a few.

How analytically, can this situation be tied into the problem before us? I shall take, one by one, a selection of theoretical concepts presented for discussion by Kenneth Hartley, an economist at the University of York.

He states that <<Utility or preference functions are central [to economic analysis]... and are assumed to be given>>. That is, there is no discussion in the central model of how individuals or groups choose preferences or determine utility (in the special economists’ sense of that word). There then follows what, to an anthropologist, would be an off-hand sketch of what in fact is a limited range of influences upon those givens, which we need not go into.

To explore the genesis and dynamics of such givens, which surely must be essential to a knowledge of the dynamics of an economy, is, as we know, extremely complex and fundamental to any dynamic interpretation of social systems. Even ethnologists often abjure digging into the dynamics, preferring a static description. But to assume the functions to <<be given>> is to relegate the act of choice to the administration of resources only — how choices, given, are satisfied whereas the overwhelming characteristic of human culture is choice between each of the ends, objectives, goals, wants and ambitions.

This of course is not overlooked in economics, though there is a tendency, except in fields such as welfare economics, to subsume linked goals under a broader classification. Thus the goal of maximized profitability can accommodate the consideration of competition between the goal of making automobiles and that of making snowmobiles. But it is more difficult to include in the consideration that an individual, an entrepreneur, a firm (large or small), a civil servant, may have alongside of strict profit a concern for an optimal mix of return embracing such matters as profit, prestige and social status, expansion for its own sake, successful gamesmanship, environmental protection (not simply as a conservation of means but as a goal in itself), aesthetics of work and product, philanthropy, family support, worker support.

Apart from providing a false simplicity of argument, the notion that preferences are given enables the economist to avoid what he may see as the trap of cultural relativism. Relativistic explanations are indeed insufficient, but non-relativistic explanations of cultural difference, and of the articulation of differences of culture within a system, must be sought deliberately. It is highly significant that no attention whatsoever was paid to such issues in the economists’ contributions to this particular symposium, even when considering exogenous factors. It is as if history and culture do not exist.

Yet surely it is very clear (as some economists do indeed recognize) that Japanese, French, Russian and British, to say nothing of Indian or Uzbek, preference functions differ enormously between each other and change significantly; and also that with each of these societies there are ethnic, class, professional, cultural, religious and other significant groups who choose differently from other citizens. Production, trading, marketing, consumption, yes and working and organization, are reflections of such differences, and of the articulative interaction between them.

Most economists I am sure would concede this, the point at issue being whether they are, or should be, exogenous in their thinking. The hold up, as I see it, is not in the potential mathematical technique, but in the relative vagueness of some of the concepts, in problems in getting agreement on operational definitions, and in lack of attention to modes of measurement, particularly when the link to monetary price is not there or not apparent.

Alternatives to monetary price have been suggested (e.g. Belshaw 1959, 1965, 1969, 1980), and are legion in studies of non-monetary societies. Whatever the practical difficulties — and there are many — it appears there are few difficulties of theoretical principle. Every economics freshman student is told that maximization in relation to consumption behaviour is not to be thought of in terms of a single commodity, but of optimal mixes in a preference schedule. Thereafter, in the handling of such institutions as the firm, the principle gets thrown away, profit being, usually, the single determinant. But if the principle of optimal mix were applied to all behaviour, an understanding of the work place, of contracts, of exchange relationships, or interactive dynamics would be more revealing, if not as <precise> as that provided in econometric models.

To some extent the raw data are there.

In anthropology, in the beginning is ethnography. Despite holistic principles, no ethnography can be culturally complete. In all such studies the data are selected in accordance with theoretical or interpretive canons of significance, which vary according to problem. Yet an ethnography, with or without quantification, is a statement of an effective demand schedule for the people concerned, that is a broad statement of what individuals actually consume at the moment of observation. The effective demand schedule is an essential tool in economic thinking, which the conventional tools of economics simply cannot approach with the same richness and directness as is achieved in ethnography. Nor can conventional economics take into account an enormous range of <<satisfactions>> which are routinely handled in ethnography — the demand for religious ritual, story-telling, recreation, even sleep, sport, work as a value in itself, learning, and on and on. How, one may ask, can economists have a realistic view of the competition and linkages that exist between elements in a demand schedule without knowing about such things?

Ethnographies — though here, customarily, they are not as strong as they might be — also have things to say about the potential demand schedule, that is those things people would like to have, do, experience - those things they would direct their actions and energies towards obtaining by way of satisfaction, if the costs were right. This ethnographies do by reference to enduring values, ideologies, structures, ambitions, frustrations and so forth.

Few economists have the patience to read ethnographies, and few ethnographic studies revealing the cultural behavior of entrepreneurs and wheeler-dealers of the kind responsible for economists’ data have been made. They should be and will be.

Ethnographic data are far too diffuse, even when quantified, to permit of immediate incorporation into models which make major use of national statistics. Yet the effort is immensely important. I would like to see economists and anthropologists, both, working on the translation of such types of data into models which would avoid the concept of the exogenous in its current forms.

Similar arguments apply to the consideration of costs, where nonmaterial issues are, for example, often behind the so-called <conservatism> of populations confronted with the possibilities of technical innovation, or with choices between immediate maximal exploitation of a resource or its husbanding for the future (Belshaw, 1976).

The work of the economist Gary S. Becker (1976, 1981) suggests that, to some degree, the extensions I have advocated are feasible within the terms of economists’ forms of reasoning. The work is far too little read among anthropologists, who would benefit their own discipline by responding to its stimulus. How can it be carried forward? Where is it wrong? What can be done about incorporating <soft> ethnography? Nevertheless, it is but a brave beginning, and needs development and extension in two directions.

In The Economic Approach to Human Behaviour (1976) Becker rejects the Robbinsian approach to <<the economic>> -- that which represents the application of scarce means to infinite ends — as being embarassingly broad, without sufficient focus, and holds that <<the economic>> should be that which relates to the market. He mostly maintains this perspective in his essays, though it is pretty well abandoned, as it has to be, for some of his papers in A Treatise on the Family (1981).

The so-called <substantive> school of economic anthropology would tend to agree, classifying <economies> into those based on reciprocity, redistribution, and market forces (Polanyi, 1957). (But if <<<the economic>> is that which is connected to markets, how then can reciprocal systems be <<economic>>???).

Others, including myself, criticize such a distinction in its purist form, since it over-emphasizes the differences rather than the similarities and continuities in the social and historical process. Scarcity, supply and demand, exchange transactions, and similar processes are present in all societies: so is the operation of choice.

To consider transactions and choices together is, for example, to reveal interactions between behaviour in a monetary market and behaviour outside that market, which is fundamental to the understanding of the market itself (Mintz, 1959). Further, the significance of Marcel Mauss’ work in Essai sur le don lies not in his treatment of gifts qua gifts, but in laying the groundwork for contemporary social exchange theory, in which all human interactions can be seen as exchange transactions, with attendant costs, benefits and real, if not monetary, prices. (The principle has even been extended, with advantage, into animal and insect biology). To make the most of social exchange theory, we ethnographers would do well to adapt Becker’s models, and ask ourselves how we can make ethnography useful in that adaptation (not necessarily retaining all the logical formalism of the models).

The second direction of extension and development is to wrestle with the ethnographic real. Becker realizes he must do this when be treats of institutions such as polygyny, which are mostly (not by any means entirely) foreign to Western culture, but it has to be (and I do not fault him) entirely superficial at this stage. He uses appropriate non-economists’ literature when dealing with primates. But he does not refer to or incorporate either ethnographic or theoretical work in anthropology on such areas as altruism, social interactions, divorce or family evolution. On other matters such as division of labour the references are unsystematic and less than minimal. The problem is not with Becker’s brilliant pioneering work (some have suggested worthy of the Nobel Prize in Economics), but with the failure of anthropologists to take up the challenge in a form that forces economists to take their work seriously — without abandoning its integrity.

It has frequently been said that the basic classical model of choice and rationality, which makes no moral judgements as to ends, and accepts that when people choose it represents their valuations at that time and place and bearing the applicable costs, material or non-material in mind, together with an evaluation, conscious or unconscious, of alternatives is tautological. And hence of no value. Tautological it certainly is. But it shows the value of tautology, since indeed one can use it, if not to discover final causes, then to forecast future events and to predict logical extensions of theory. Hartley states that Becker’s approach is <<often guilty of adhocery and of failing to provide distinctive hypotheses capable of being refuted>>. Not so.

Such charges are also levelled against the extension of the model. It appears to be a sin to embrace an enormous range of data into a basically simple, single, model, although in other sciences the canons of preferred explanation call for such abstraction and simplicity. Yet the power the model gives to explanatory elucidation of empirical material is considerable, and general, testable (provided some form of quantitative judgement is possible) hypotheses abound.

Some examples. In applied or developmental work, or work analyzing social change empirically, it more often than not turns out that forecasts can be made on the basis of cost benefit analysis, or something similar, taking into account the wider and non-material factors I have indicated as significant. The forecast fails. The model then enjoins the search for <<the missing factors>>, or the factors that were not appropriately weighted in the original scheme. The awareness of the need to adjust the schedule of preferences accordingly is surely fundamental to subsequent forecasts, and if used properly adds to our knowledge of the interplay between elements of culture which affect choices.

Second example. The assumption of rationality in the existing model is frequently challenged (e.g. Godelier, 1966). But such challenges ignore the special meaning of rationality in economics, which has nothing to do with psychologically rational processes or otherwise in the selection of goals. Usually, with such challenges, <non-rationality> can be broken down into component parts, such as search for prestige, altruism, anti-social behaviour, asceticism, fanaticism. But in fact all we are doing here is to add elements to the preference schedule, saying that people want these things, that they are different perhaps from the norms of Western society (perhaps...) and in so doing we add distinctive, sometimes curious, production functions for their satisfaction.

Third example. From anthropology one can pull out numerous general hypotheses, dealing with the way cultural elements dynamically articulate. That is, in them, we account for adjustments in preference schedules over time and in response to new conditions. Some are at the level of deductions close to the basic model (Belshaw, 1969), as <<Where actors perceive that ends are incompatible, conflict, synthesis and adjustment will occur with a view to making them compatible>>. Surely this sets the stage of the possibility of dis-proof (i.e. the search for contrary cultural examples) and refinement, that is defining the conditions under which the various alternative empirical paths will be followed.

Other hypotheses derive from the independent examination of social phenomena. Thus the study of millenarian movements posits specific conditions under which, among other things, populations will engage in the immediate destruction of resources in anticipation of special kinds of gain (Jarvie, 1964). (Ethnocentric observers call this <irrational>, but in the economic model it is highly rational, in that the people are behaving in accordance with perceptions that state that destruction is the means to the desired salvation). Such generalizations about millenarian movements can be extended, with appropriate modification, to embrace wider phenomena such as more conventional social movements, and other kinds of social protest, into one over-riding formula, though this has not yet been completely done. And it has been used, albeit clumsily, for an attempt at historical forecasting, which was in fact partially falsified (Belshaw, 1969, 1976). (I used the model to forecast the outburst of millenial movements in the Palestinian diaspora, whereas in modified form it burst out in Iran —- possibly in accordance with the model, but not as I stated it).

Many such hypotheses, since they deal with the articulation of aspects of society and culture, institutions, social and natural environmental factors, cultural categories, and so forth, speak directly to regularities of inter-action which the symposium papers identified as linkages (where anthropology and economics have much to say to one another). (Linkages = articulation in Belshaw, 1969).

A final issue. Economists are busy trying to find effective ways of handling variations in the state of knowledge, to which the crucial concept of innovation is closely related. From its very beginning, though in disguised language, anthropology and ethnology, through the study of social and cultural change, even of evolution, have been preoccupied with similar questions. <<The state of knowledge>> is of course nothing more nor less than the state of culture, including symbols, perceptions, ideologies and moral codes.

Anthropologists are now chary of jumping to conclusions that given societies or groups are inherently more innovative than others, though development economists sometimes cling to the despair of such an idea. Our empirical materials do concede that innovations may be concentrated in varying areas of culture at any given time (religion as against agricultural technology, for example). Our theories include general statements about the probability of innovations being produced and/or accepted or rejected, and the ways in which they are necessarily adapted when adopted. There are also general statements about what causes innovation and what factors are relevant to the prediction of a rate of innovation (Belshaw, 1969). If economists are to include knowledge and/or innovation as factors of production, anthropologists have much to say, both theoretically and empirically.

Economists continue to ask, what relevance does ethnology have to modern society? I need not belabour the point to this readership, since so many of our colleagues are directly involved in such societies. But there is of course a different relevance also. It was the study of the potlatch among North American Indians and of the kula of the Trobriand Islands which enabled Marcel Mauss to develop ideas which are now being applied to Western political parties, personal and subjective relations in the behaviour of entrepreneurs, and the enormous effect that ceremonial exchange has within the confines of the capitalist economy. Ways of thinking about small bounded groups enables analogies to be used in the analysis of boundaries and articulations between ethnic groups and social classes, between sub-cultures (including those of professionals, entrepreneurs and bureaucrats) and the nation state. Studies of small scale entrepreneurs carried out in Fijian villages contain points entirely compatible, historically, with considerations affecting OPEC oil ministers, and the Swat Pathan were not irrelevant for classic studies of Norwegian fishermen.

To conclude, the enormous material now present in the theory and practice of anthropology is essential to the consideration of economists’ exogenous variables. It would ultimately be more effective if variables could be incorporated directly into the basic models. Nevertheless, even if conservatism prevents this, each exogenous variable, whenever defined, is open to ethnographic data and anthropological hypotheses, and can be refined into many traceable parts. To achieve this is a programmatic matter, an issue of scientific strategy and priorities, rather than one of insuperable intellectual obstacles.

I wish to acknowledge the support of Unesco and the International Social Science Council for participation in the symposium, and of the Wenner Gren Foundation for Anthropological Research for a general writing grant that enabled me to proceed with these matters.

REFERENCES

Acheson,James M. (1981) Anthropology of Fishing. Annual Review of Anthropology, 275-316. Bailey, FG. (1969) Stratagems and Spoils. Oxford.

— ed. (1971) Gifts and Poisons, the Politics of Reputation. Oxford.

Banton, M., ed. (1965) The Relevance of Models for Social Anthropology. London. Barnett, H.G. (1983) Qualitative Science. New York.

Barth, F. (1959a) Political Leadership Among the Swat Pathans. London,

— (1959b) Segmentary opposition and the theory of games: a study of Pathan organization. Journal of the Royal Anthropological Institute, 89: 5-2 1.

(1963) The Role of the Entrepreneur in Social Change in Northern Norway. Bergen.

— (1966) Models of Social Organization. London.

Becker, Gary 5. (1976) The Economic Approach to Social Behavior. Chicago.

— (1981) A Treatise on the Family. Cambridge. Mass.

Belshaw, CS. (1959) The identification of values in anthropology. American Journal of Sociology, LXIV: 555-562.

— (1965a) Traditional Exchange and Modern Markets. Englewood Cliffs, N.J.

— (1965b) The cultural milieu of the entrepreneur, in Hugh Aitken. ed., Explorations in Enterprise. Cambridge, Mass. Pp. 139-162.

(1966) Evaluation of technical assistance as a contribution to development. International

Development Review, viii, 2: 2.6 and 23.

— (1967) Theoretical problems in economic anthropology, in M. Freedman, ed. Social Organization. London.

—~ (1969) The Conditions of Social Performance: an Exploratory Theory. London.

— (1976) The Sorcerer’s Apprentice: an Anthropology of Public Policy. New York.

— (1978) On scale, organization and performance. Paideuma 24: 11-23.

— (1959) Toward a comparison of value-emphases in different cultures, in L.D. White, ed., The State of the Social Sciences. Chicago. 116-132.

— and Kluckhohn, Clyde, eds. (1952) Culture: a critical reviewof concepts and definitions. Papers of the Peabody Museum of Archaeology and Ethnology, 47, vol. 1. Cambridge.

Levi-Strauss, Claude (1958) Anthropologic Structurale. Paris. (English translation 1963 New York).

Mauss, M. (1925) Essai sur Ic don, forme archaique de l’fchange, in Sociologie et anthropologic, Paris. Reprinted 1950. English translation 1954 et seq as The Gift, London.

Mendelsohn, E. and Elmana, Y. (1981) Sciences and Cultures: Anthropological and Historical Studies of the Sciences. Dordrechc.

Mincz, Sidney (1959) Internal market systems as mechanisms of social articulation. Proceedings of the American Ethnological Society, 20-30.

— (1969) Pratik: Haitian personal economic relations. Proceedings of the American Ethnological Society, 54-63.

Nash, J. (1981) Ethnographic aspects of the world capitalist system. Annual Review of Anthropology, 393-424.

Orans, Martin (1968) Maximising in Jalmani Land: a model of caste relations. American Anthropologist, 70: 875-897.

— and Howard, A. (1971) Decision-making and the study of social process. Acta Sociologica 14: 213-226.

Parsons, T. (1935) The place of ultimate values in sociological theory. International Journal of Ethics, 45: 282-316.