MODELS AND INTERPRETATIONS
John Barnes's collection of essays, published over the last forty years, covers a variety o...
109 downloads
1753 Views
7MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
MODELS AND INTERPRETATIONS
John Barnes's collection of essays, published over the last forty years, covers a variety of topics in sociology and anthropology, including lineage systems, social networks, colonialism, underlying assumptions of social science, and the significance of time in social analysis. Together they identify the author's particular view of social science: he is primarily interested in 'what really happens'. Rather than revamp articles written with a distinctive set of assumptions to bring them into line with current intellectual fashions, Professor Barnes has chosen to let them stand as they are, products of identifiable theoretical stances and modes of exposition. But introductory notes to each chapter set out the context in which the piece was originally written and draw attention to later publications and events that bear on it. A new introduction discusses in detail the author's view of social science as the construction of models rather than a search for social laws, while the final chapter presents a model of the modelling process itself. Models and Interpretations will be of interest to sociologists, anthropologists and philosophers of social science.
MODELS AND INTERPRETATIONS SELECTED ESSAYS
J. A. BARNES
The right of the University of Cambridge to print and sell all manner of books was granted by Henry VIII in 1534. The University has printed and published continuously since 1584.
CAMBRIDGE UNIVERSITY PRESS Cambridge New York Port Chester Melbourne Sydney
CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521366533 © Cambridge University Press 1990 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1990 This digitally printed first paperback version 2006 A catalogue recordfor this publication is available from the British Library Library of Congress Cataloguing in Publication data Barnes, J. A. (John Arundel), 1918Models and interpretations: selected essays / J. A. Barnes. p. cm. Bibliography. Includes index. ISBN 0-521-36653-4 1. Sociology — Methodology 2. Anthropology — Methodology 3. Paradigms (Social sciences) I. Title. HM24.B343 1990 3 0 1 \ 0 1 - d c 2 0 89-7046CIP ISBN-13 978-0-521-36653-3 hardback ISBN-10 0-521-36653-4 hardback ISBN-13 978-0-521-02493-8 paperback ISBN-10 0-521-02493-5 paperback
CONTENTS
Preface Acknowledgements Introduction: social science in practice
page vi viii i
I MODELS OF THE REAL SOCIAL WORLD
1 2 3 4 5
Genetrix : genitor :: nature : culture ? (i973g) African models in the New Guinea highlands (1962a) Agnatic taxonomies and stochastic variation (197m) Class and committees in a Norwegian island parish (1954a) The righthand and lefthand kingdoms of God: a dilemma of Pietist politics (1971k) 6 Indigenous politics and colonial administration, with special reference to Australia (1960a) 7 The perception of history in a plural society: a study of an Ngoni group in Northern Rhodesia (1951c)
29 44 56 67 88 103 120
II MODELS OF SOCIAL SCIENCE
8 Feedback and real time in social inquiry (1967b) 9 Timeflieslike an arrow (197if) 10 Kinship studies: some impressions of the current state of play (1980c) 11 Sociology in Cambridge: an inaugural lecture (1970m) 12 Social science in India: colonial import, indigenous product or universal truth? (1982c)
133 150 169 181 196
III A MODEL OF MODELLING
13 Modelling: for real or for fun? (1983b)
215
Postscript: structural amnesia (1947: 52-3)
227
Bibliography Index
229 260
PREFACE
All the essays constituting the chapters of this book have previously appeared in print. If I were to start afresh on any of the topics they cover, I would make substantial changes. Nevertheless I have as far as possible left them in their original published form. Attempts to revamp articles written with a distinctive set of assumptions, to bring them into line with current intellectual fashions, are seldom successful; I prefer to let them stand as they are, products of identifiable theoretical stances and modes of exposition. I have also left unchanged my use, following the convention prevailing when the earlier essays were written, of masculine nouns and pronouns to indicate both genders; this obsolescent usage should serve to remind the reader, as it does me, that most of the essays were first drafted many years ago. It is not only fashions in theoretical concepts that change with the times. I have however placed a note at the head of each chapter, setting out the context in which the essay was written and drawing attention to later publications and events that bear on it. Chapters 6 and 7, where colonialism is discussed, have been revised to the extent of replacing the present tense by the past, since the system they focus on has largely disappeared, despite its apparently firm prospects for the future at the time the essays were written. I have done the same with Chapter 4, for the social conditions I found in western Norway in 1952 changed radically a few years later following the collapse of herring fishing and the discovery of oil in the North Sea, changes that I touch upon in Chapter 5. Throughout the book I have updated references where appropriate, as well as adding a few new sources. In a few places I have been unable to resist the opportunity to amend glaring infelicities, and I have removed a few faded topical allusions. It is difficult to make adequate acknowledgement to the very many people who through the years have helped me with these essays. I am much indebted to thefifteenresearch and university institutions in which I vi
PREFACE
Vii
have worked, for the support I have received from them, and to the foundations that have provided generous financial assistance. They have helped me towards achieving my aim of being a dilettante in the original sense, someone who delights in the work he or she does. I am very grateful to the Ngoni people of Zambia and Malawi and the residents of Bremnes, in western Norway, for allowing me the privilege of learning from them, in their very different ways. Among the students, colleagues and friends from whom I have gained so much I must mention by name L. R. Hiatt, M. J. Meggitt, D. M. Schneider, W. Ullman, W. K. Whitten and D. E. C. Yale for their help with Chapter i; the late Ely Devons for comments on Chapter 4; A. L. Epstein for help with Chapter 5, and the late W. E. H. Stanner for comments on Chapters 5 and 9; M. Godelier also helped me in drafting Chapter 9. Jack Goody encouraged me to start on this collection of papers; Geoffrey Hawthorn and Tim Ingold made valuable suggestions on how it should be put together. To Frances Barnes I am greatly indebted not only for comments on Chapters 1 and 5 but also for constant support throughout the years. Beverley Bullpitt helped nobly with the preparation of a presentable typescript.
ACKNOWLEDGEMENTS
Chapters 2, 11 and 12 are reprinted from Man, vols. 62 (1962); n.s.15 (1980); n.s.6 (1971) by persmission of the Royal Anthropological Institute of Great Britain and Ireland. Chapter 3 is reprinted from Anthropological Forum III, 1 (19 71) by permission of The University of Western Australia Press. Chapters 4 and 7 are reprinted from Human Relations, vol. 7 (1954), and vol. 4 (1951) by permission of the Tavistock Institute of Human Relations. Chapter 6 is reprinted from Comparative Studies in Society and History, vol. 2 (i960) by permission of the journal editor. Chapter 10 is reprinted from Indigenous Anthropology in Non-Western Countries (1982), by permission of the Wennner-Gren Foundation for Anthropological Research Inc. Chapter 13 is reprinted from Connections by permission of the journal editor.
Vlll
INTRODUCTION: SOCIAL SCIENCE IN PRACTICE
i Preamble The thirteen papers reprinted in this book were written over a period of nearly forty years, in response to a variety of requests and opportunities, and draw on a very mixed bag of empirical data. Some are my contributions to symposia I was asked to join, others were written for Festschriften, and a couple were delivered as addresses to conferences. Yet although none was designed as a component in an orderly exposition of some gradually specified theoretical stance, they do all imply, some more strongly than others, a view of the enterprise of social science that differs from the views held by many of my colleagues. This view has gradually become clearer to me as I have argued with them. The time has come to make explicit the many assumptions that I have silently, and to a large extent unconsciously, made in writing these papers. In this introduction I try to meet the challenge; in particular I endeavour to present a picture of social inquiry that takes account of the inherent differences between the natural and the social branches of science, and that distinguishes social science from the humanities. In building up this picture, my starting-point is the practice of social science, rather than the philosophy of social theory from which so many other views have been derived. I am concerned with 'the social sciences, as actually practised and identified in contemporary societies', as Gellner (1984: 567) puts it. In this way I am following my own intellectual development, for, like many social scientists of my generation, I began making empirical inquiries about the social world blissfully untroubled by questions of ontology and epistemology. I was content to shelter behind Fortes and Evans-Pritchard's (1940a: 4) bald and misunderstood assertion that they considered the theories of political philosophers to be 'of little scientific value'. Only later, when I found myself obliged to defend what I
2
INTRODUCTION
was doing against critics from many quarters - notably natural scientists, historians and philosophers - did I attempt to work out in coherent fashion the hitherto somewhat muddled analytic framework that I had been taking for granted. I am aware that philosophers of social science like Richard Rorty (1980: 315) can proclaim 'the demise of epistemology' but the continuing flood of books and articles on the philosophical foundations of social science indicates that, even if buried, the corpse of epistemology is still kicking. More importantly, the significant, and in my view largely deleterious, impact of theoretical orientations such as ethnomethodology and reflexivity on the praxis of social inquiry shows that the philosophy of social science is too important a matter to be left to the philosophers. Given the limitations of possible human achievement and the diversity of human preferences, we cannot hope to avoid a division of labour between those who enjoy learning about the real world - whether this be the fantasies of Neasden grandmothers or the bargaining in Melanesian markets-and those who concentrate their attention on discussing whether and how anyone can learn anything about the real world. Each activity has its own autonomy, and each set of practitioners pays at most only intermittent attention to the pronouncements and discoveries of the other. In the natural sciences the divide is even sharper. The contempt in which many practising natural scientists hold the philosophy of science is notorious; they regard the subject as a soft option that good students should be encouraged to avoid in favour of real science courses. With natural science in mind Medewar (1969: 12) comments that Science, broadly considered, is incomparably the most successful enterprise human beings have ever engaged upon; yet the methodology that has presumably made it so, when propounded by learned laymen, is not attended to by scientists, and when propounded by scientists is a misrepresentation of what they do. Only a minority of scientists have received instruction in scientific methodology, and those that have done so seem no better off. In most branches of social science there is more open recognition that some amount of theoretical, and even philosophical, sophistication is necessary to prevent curiosity about the real world leading only to lowlevel empiricism and mindless number crunching. Indeed Henri Poincare is said to have commented 'The natural sciences talk about their results. The social sciences talk about their methods'. Nevertheless the divide between the empirical investigators and the theoreticians exists even in social science; in Chapter 13 I indicate why, in my view, it is a permanent feature and why advances in understanding are made sometimes mainly on one side of the divide and sometimes mainly on the other. But that is an Olympian account of the scene. For the moment, if I am to defend the view that those of us whose primary interest is in 'what really happens' need to have some understanding of the extent to which we can
SOCIAL SCIENCE IN PRACTICE
3
possibly satisfy that interest, I have to set out the broad outlines of my own understanding of the matter. In other words, I have to construct a model of social science. 2 Sciences and scientists Let us begin by considering the relation between ideas and the individuals who propose, oppose or defend them. This relation is sometimes crucial for understanding ideas, and sometimes quite irrelevant. For example, the history of scientific ideas may be handled in many ways depending, roughly speaking, on the relative weight given to the interactions of disembodied ideas and to the histories of the individuals who propagated them. Notably, the history of mathematics can be presented satisfactorily in terms of the sequence of the discovery of its theorems, with minimal reference to the people who discovered them; the history of graph theory, a branch of mathematics becoming of interest to social scientists, has been written in this way (Harary 1973). The natural sciences, which are less austere than is mathematics, can be viewed historically from either internalist or externalist standpoints, giving less or more weight respectively to the wider social environment within which scientific discoveries were made and paradigms discarded and adopted. However, when we come to the social sciences, we have to recognize that the social context in which theories are propounded and inquiries undertaken is not merely a background factor which, at the whim of the analyst, can be either incorporated into the analysis or omitted from it. To assess the rationale, the utility and the success of a social theory we are forced to take account of the social context within which the theory was developed and the political, religious, economic and even demographic characteristics of those propounding and opposing the theory. Nowhere is the significance of the broad social context made clearer than in communist-led countries, where drastic changes in the theoretical interpretations that are allowed, and in the topics that can be investigated, and indeed the very establishment or abolition of social science disciplines have all depended on political vicissitudes (e.g. Cheng and So 1983). These considerations apply clearly to the history of ideas but have to be heeded also in discussing the present status of ideas, at least when the ideas under scrutiny are those of social science. For many of the features that distinguish social scientists from, say, natural scientists and humanists and from the lay public derive directly from the specific differentiae which mark off the social sciences from other branches of systematized knowledge. The epistemological status of the social sciences has been controversial in the past and remains controversial today, and in like fashion the status of their practitioners, the social
4
INTRODUCTION
scientists, has been and continues to be a matter of debate not only among the practitioners themselves but also in the wider public arena. The roots of this latter debate are to be found, at least in part, in the unsettled status of the social sciences themselves. There is thus a dialectical relation between the intellectual characteristics of the social sciences and the social attributes of their practitioners, each influencing the other. Indeed, the existence of this dialectical or two-way relation is one of the features that mark off social sciences and social scientists from the natural sciences and natural scientists. For example, the speed of light is the same, whether it is measured in Washington or Moscow. Atomic reactors, not to mention atom bombs, work according to the same principles on both sides of the Iron Curtain. In the Soviet Union Marxist physicists enjoy a monopoly of research opportunities, whereas in the United States they are denied them; yet Marxist and bourgeois physicists are in agreement about the inventory of elementary particles. On the other hand, in social science the same differential treatment of Marxist and bourgeois practitioners in the two countries is paralleled by a lack of agreement between them about, say, the class structure of advanced societies and the nature of the state. To discuss what differentiates the social from the natural sciences we must therefore pay attention also to the differences between social scientists and natural scientists. 3 The attributes of social science The strength of the two-way relation between science and scientist is not the only attribute of social science that particularly bears upon the social status of its practitioners, the social scientists. The next that should be mentioned is the fact that the subject-matter of the social sciences is made up of phenomena that are familiar to the general public. This is a point that has been repeatedly emphasized by Giddens, as for instance when he writes that 'the production and reproduction of reality ... has to be treated as a skilled performance on the parts of its members' (Giddens 1976: 160). Giddens would certainly not claim to be thefirstto make this point, and the same viewpoint was adopted earlier, and in a more uncompromising form, by young and enthusiastic ethnomethodologists who, in the early '70s, tended to claim that 'members' were at least as skilled in the analysis of their own social environment as any social scientist who tried to study them (cf. Mehan and Wood 1975: 37). This stance of epistemological populism was particularly marked in sociology, but an analogous viewpoint was found also in social anthropology where, under the rubric of 'cultural relativism', it has had a longer history and a wider theoretical underpinning. In both disciplines this viewpoint has had serious consequences for the debate about whether, or to what extent, social scientists
SOCIAL SCIENCE IN PRACTICE
5
are professionals and not merely verbose citizens. Substantial numbers of practitioners in the two disciplines nowadays embrace the many-headed hydra of 'reflexivity' and risk becoming engulfed in the black hole of an infinite regress (e.g. Rose et al. 1987). There is no corresponding phenomenon in the natural sciences, for most citizens are not skilled performers in any of them, and notions of epistemological populism are nowhere entertained. Many routine inquiries in social science, of course, do not depart substantially from the natural science paradigm, in which, among other things, there is a clear distinction between the observer and the observed, and where whatever is observed supplies only data, not analysis (cf. Barnes 1980a: ch. 2). Nevertheless, we cannot overlook the significance of the fact that social scientists are the only scientists who gather most of their data by talking to people and asking them questions. A third intellectual feature that affects the social position of social scientists is the quality of the language used in social science. On the whole, the language of natural science is incomprehensible to the untutored lay person. Incomprehensibility arises in two ways, from the words that are used and from the meanings attached to the words. Chemistry seems to constitute the extreme case, for many of the words in its special vocabulary are so long and segmented that they cannot be assimilated into ordinary discourse, and the meanings attached to these words also seem hard for the lay person to grasp. Physics makes much greater use of words that are drawn from ordinary speech, as for example the so-called 'colour' of quarks, but gives these words meanings that are both precisely defined and significantly different from their meanings in lay speech (Hey and Walters 1987: 156-7). Social science also uses ordinary words and continually makes attempts to give these words precise and technical meanings, as for example with 'class', 'honour' and 'status'. But there is a striking difference between, say, physics and sociology in the way in which their use of language is seen by outsiders. No one complains because physicists nowadays use, for example, the word 'atom' in a way different from its Shakespearean usage; there is no lay assumption that there is a real meaning to the word 'atom', and that any departure from this usage by physicists is preposterous jargon. However, when social scientists use words like 'power' and 'authority', the lay public is always ready to protest that these words are not being used properly, and not being given their real meanings. These protests occur despite explicit statements by social scientists that these words are, in the context concerned, to be understood not in their everyday usage but in some other, usually more restricted, and analytically more useful, sense. Yet at the same time the wish, or need, to make some impact on the wider world, and in particular to make a contribution to the formulation of public policy, constrains social scientists to use, as far as they can, the language of the market-place and the
6
INTRODUCTION
political arena. As Rorty (1981:5 74) remarks, 'Predictions will do ''policymaking" no good if they are not phrased in the terms in which policy can be formulated'. Nevertheless one defence against this kind of lay criticism is to concoct technical terms that initially do not have any clear lay meaning at all. An example of this ploy is the invention of the term 'socio-economic status'. The adjective 'socio-economic' was first used as early as 1883 (Ward 1883: 525) and the expression 'socio-economic status' was used later to avoid some of the confusion over the diversity of lay and scientific meanings attached to the word 'class'. Forty years ago the former term was still seen by the laity as a typical instance of unpleasant sociological gobbledegook, but nowadays it raises no eyebrows. For with this and similar technical terms a process of diffusion operates which is fundamental for understanding the relation between social science and the wider public. Robert Merton notes that some sociological technical terms drift into ordinary speech (Merton 1981) but there is more to this process than simply drifting. Would-be trendy writers in high-class literary magazines and Sunday newspaper supplements take up new technical terms from social science for their own purposes, and thus begin the gradual modification of the denotation of these terms. The precise definitions with which these terms may have been first introduced are soon forgotten, and by the time the words have passed into ordinary everyday speech they have acquired all the polysemic attributes that make lay speech so vivid and ambiguous at the same time. 'Socio-economic status' can now be heard on lay lips all over Australia, and I guess elsewhere in the English-speaking world, along with 'J-curves', 'charisma' and 'relative deprivation' (cf. p. 144). The process of transformation may in some instances go a stage further, whereby a technical term acquires a lay meaning far removed from the meaning intended for it by the social scientists who first used it. A good example is provided by the term 'network'. The word has been part of the English language since Shakespearean times but I think it was not used in social science in a technical sense until 1954, when I published an article which hinged on a technical use of the word (Chapter 4). Some twenty-five years later interest in the topic among social scientists had grown sufficiently for the launch of a new international journal with the title Social Networks. This title conforms to the 1954 definition, but by that time the idea of a social network had already passed into ordinary speech and had lost the precise denotation I had given it. This did not surprise me, but I must admit to being surprised at the emergence of 'networking' as a new English verb, and to the appearance of journals and other publications advocating 'networking' as a new form of happiness, as Muller (1986) claims in a poem used as promotional material by the American-based
SOCIAL SCIENCE IN PRACTICE
7
Networking Institute Incorporated. On the other hand, the publication of whimsical in-jokes, based on the 1954 concept of network, and incomprehensible to the general public, in the increasingly technical pages of the journal Social Networks (Keul and Freeman 1987), is strong evidence for professional closure in this sector of quantitative sociology. Thus we have a somewhat paradoxical situation. In general, natural scientists are not criticized by lay writers for writing incomprehensibly in their technical journals, and for making extensive use of technical jargon, whereas the publications of social scientists are continually subjected to this kind of criticism from lay reviewers, or indeed from fellow social scientists belonging to rival factions (cf. Becker 1983: 578). But at the same time there is a continual trickle of words from the technical language of social science into ordinary speech, with changes in meaning occurring along the way. Commenting on this process, Runciman (1965: 47) remarks that 'jargon is not jargon if it succeeds: like treason, jargon is by definition a failure'. Social science thus not only lacks a vocabulary of technical terms with stable meanings; in addition its persistent attempts to create such a vocabulary are derided by the general public as insufferable goobledegookery. Maybe this criticism would not seriously impair the aspirations of many social scientists towards public recognition as professionals and as scientists, were it not for another feature of the relation between them and the public. For not only do the technical terms of social science pass into the wider arena of ordinary speech, albeit with modified meanings, but the findings of social inquiry are also taken up into the corpus of general knowledge and common sense. In part, of course, they are there already, for a great deal of empirical inquiry in social science is directed towards discovering which popular assumptions about the workings of society are true and which are false. Fortunately many are true. At least it is fortunate for society that many citizens are substantially correct in the assumptions they make about the social environment in which they have to operate and survive; but at the same time it is unfortunate that social scientists are often criticized for wasting time and effort on discovering what everyone knows already. There are, however, other instances where empirical inquiry demonstrates that a popular assumption is wrong and where, therefore, social scientists might expect praise from the laity for their efforts in revealing the truth. But, alas, the praise, if it is accorded at all, does not last long. After a lapse of time the research finding becomes incorporated into popular wisdom. The stimulus for the change is forgotten, so that in retrospect the inquiry may well be seen as yet another instance of wasting money to tell us what we know already. The research conducted during the Second World War by Stouffer and his colleagues (1949) on morale in the United
8
INTRODUCTION
States Army is an example of this process. Their finding, that morale was higher in army units where promotions were comparatively rare, was seen at the time as unexpected and contrary to popular belief, but forty years later is part of the accepted wisdom of every personnel manager, if not also of members of the wider public. For these reasons I have argued elsewhere that, at least for sociology: Building an edifice of professional achievement is like building on quicksand; the scientific achievements of one moment are buried as common sense in the next. I go on to describe 'the continual denigration of sociology in the popular media, in literary journals and by colleagues in other disciplines' as 'a rational defence by the laity against the claims of sociologists to know best' (Barnes 1981c: 22). Because so much of sociological inquiry is directed parochially, at the society to which the sociologists concerned themselves belong, the discipline is particularly vulnerable to this type of criticism; social anthropology is on safer ground when its inquiries are directed at exotic others, but when it tackles the local scene it encounters the same denigration. The absorption of social science findings into the pool of common sense explains a contrast to which Gellner calls attention. He notes, somewhat over optimistically in my view, that on the one hand the social sciences are characterized by The presence of well-articulated hypotheses and their systematic testing. Precise quantitative measurement, and the operationalization of concepts. Careful observation by publicly checkable methods. Sophisticated and rigorous conceptual structures, and great insights. Shared paradigms, at any rate over sizeable communities of scholars, and persisting over prolonged periods. On the other hand, in terms of 'the impact on our cognitive world', the social sciences are quite unlike the natural sciences, for the former obviously lack 'a generally overall consensual cognitive activity, radically discontinuous from the insights and techniques of ordinary thought, and unambiguously cumulative at an astonishing and unmistakable rate' (Gellner 1984: 584). I would doubt whether cognitive activity in the natural sciences is, or ever has been, 'radically discontinuous' from ordinary thought, but I agree with Gellner that the discontinuity is greater for most natural sciences than it is for most social sciences. Common sense absorbs quickest those scientific findings that form part of the common way of life. Cambridge provides good evidence of the extent to which those who are not social scientists expect the literature of social science to be written in language that all can understand, and also expect its findings to be in
SOCIAL SCIENCE IN PRACTICE
9
conformity with their lay understandings. The Press Syndicate of the university has the fortnightly task of deciding what books the Press shall publish. Apart from two ex-officio members, the other sixteen Syndics are drawn from a range of disciplines intended to match the range of books currently published by the Press. While I was serving as a Syndic there was an obvious hierarchy in the way in which proposals for publication were discussed. In most cases proposals in mathematics were quite unintelligible to all Syndics except for a solitary mathematician, and only he spoke to these proposals. Natural science proposals were discussed more widely, not only by the Syndic who was expert in the discipline concerned but also by his fellow natural scientists. Proposals in the humanities were assessed mainly by the humanists, but with significant contributions from several other Syndics whose professional qualifications lay elsewhere but who were also well versed as amateurs in music, painting and literature. Proposals in the social sciences, particularly in sociology, were however seen as fair game by all Syndics; indeed there was, I think, a feeling that every well-educated person ought to have an opinion about the merits and demerits of social science proposals and that no Syndic, whatever his or her speciality, could escape sharing full responsibility for deciding whether or not to publish. 4 The dearth of social laws In reviewing the relevant differences between the social and the natural sciences I have mentioned so far the dialectical relation between social scientists and the theories they have supported or contested; the competence of lay citizens in the subject-matter of social science, contrasted with their comparative lack of competence in the natural sciences; the impossibility of preserving a technical language in social science uncontaminated by lay connotations; the propensity of common sense to soak up the achievements of empirical inquiries. All these are features of social science that entail a social and intellectual status for social scientists that is significantly different from that occupied by natural scientists. There is, however, one more contrasting feature that many natural scientists, as well as philosophers of science, regard as inescapably fatal to the aspirations of social scientists for recognition as 'real' scientists. By and large, the corpus of accumulated social science understandings lacks propositions analogous to the laws, that is to say hypotheses that have survived attempts at refutation, that characterize natural science. Instead of logically interrelated scientific laws, social science understandings are made up of interpretations and models, together with a vast but only minimally interrelated array of empirical findings. Most social scientists in earlier generations saw this state of affairs as a
IO
INTRODUCTION
temporary condition, a shortcoming arising from the immaturity of the social sciences that would gradually be remedied with the passage of time, the accumulation of more evidence and the development of more powerful theories. But a later generation has abandoned this Whiggish prospect and views even Merton's famous tactic of concentrating on middle-range theories as barking up a blind alley. Giddens for instance argues that although there are valid generalizations in social science, 'they have a logical form substantially discrepant from laws in the natural sciences, whether these be universal laws or whether they be statistical-type laws . . . In social science to generalize about something means generalizing also about people's knowledge of the circumstances of the action involved' (Mullan 1987: 107). I am in broad agreement with Giddens on this point, though I would express it somewhat differently. As I see the difference, in the natural sciences law-like generalizations help us to understand how things work, whereas in social science law-like generalizations help us on the way towards making things work differently. The main consequence of this difference in intellectual structure between the two branches of science is the existence of a plurality of models and interpretations in social science, in contrast to the prevalence of monolithic orthodoxy in natural science. The contrast is of course not absolute, for there are plenty of rival interpretative models in natural science also. But these are all to be found on the frontiers of inquiry, while behind the expanding frontier is an essentially orthodox terrain where scientific controversies have been resolved and are now only of antiquarian interest. In this light the whole of social science is a frontier zone, with rival interpreters contesting every square inch of territory. Hence we have not a unified and integrated set of social science propositions, supported by evidence, but rather a plethora of rival schools, appealing to partially overlapping bodies of empirical evidence or, in some cases, not bothering very much about empirical support at all. Consequently social scientists continually criticize one another's interpretations, contradict one another's assertions in courts of law, and offer conflicting advice to governments and other clients. Answers to examination questions in natural science can be easily sorted into right and wrong, whereas wrong answers in social science typically consist of, say, attributing to Weber the views held by Durkheim rather than asserting a false interpretation of events. I realize that these contrasts fit some social sciences more closely than others. In some branches of psychology, scientific laws have indubitably been discovered and tested. But it can be argued that the label 'psychology' covers an area of inquiry that straddles the natural-social divide, and that scientific laws are to be found largely or solely on the natural science side of the subject. Indeed, recently in Britain a joint committee of the Royal
SOCIAL SCIENCE IN PRACTICE
II
Society and the British Academy decided to partition psychology, some branches coming under the aegis of the Royal Society as natural sciences, and other branches being allocated to the British Academy as social sciences. At the other end of the social science continuum the conflicting interpretations of the same scene by economists are notorious. Indeed, the condition of economics is a salutary reminder that quantification and mathematical sophistication do not constitute a panacea for achieving consensus among professionals. 5 Scientists? The absence of a corpus of generally accepted, universally applicable and logically interrelated propositions, more than anything else, prompts natural scientists to deny the label of scientist to their colleagues in social science. The general public follows their lead, so that in ordinary speech the contrast between natural and social science is expressed as a contrast between science and social science; natural science is, as the linguists would say, the unmarked term; the default value of 'science' is 'natural science'. This usage is reinforced by the attitude taken by many social scientists themselves. At one stage in his career the Oxford anthropologist Evans-Pritchard (1951: 7) placed his discipline firmly among the humanities rather than among the sciences (cf. Redfield 1953), and more recently a book appeared with the intriguing title A poetic for sociology (Brown 1977). I do not know of any attempts to place psychology or economics among the humanities, but the close connexion between political science and philosophy in the more conservative universities casts doubt on which side of the dividing line the former discipline should fall. The science/non-science distinction generates a great deal of heat and emotion, particularly when funds for research and teaching are at stake, but it is a distinction belonging to academic Realpolitik rather than to intellectual debate. My sympathies are with Dahrendorf, when he says of sociology 'I couldn't care less whether it's called a scholarly discipline, an intellectual discipline or a scientific discipline' (Mullan 1987: 43). The distinction between scientist and non-scientist in its Anglophone sense simply vanishes when translated into French or German, despite the Platonic or normative connotations of the term 'science' in English to which Gellner (1984) draws attention. The contested status of the social sciences may even have political or ideological implications, as we shall see in Chapter 12, where we look at what has happened to the social sciences in India. The claim to recognition as a branch of science can also be manipulated for reasons that have no connexion with the structure of knowledge, but which are related to more immediately pressing breadand-butter issues. I once worked in a university anthropology department
12
INTRODUCTION
that resolutely defended the position it had acquired, by historical accident, as part of the pre-clinical medical faculty, and did so for the very sound reason that it was logistically more advantageous to remain a cheap science rather than to become an expensive art. It is only sciences, cheap ones included, that can afford flunkeys in white coats to serve morning tea. Some theoreticians of the knowledge industry, like many politicians, are fond of drawing up comprehensive battle plans for the attack on ignorance and the transmission of understandings and skills, but on the battlefield itself the disposition of troops is largely determined by local commanders with little regard for logical neatness. The chronically and deliberately obscured division of labour between sociology and anthropology illustrates this principle admirably. For the argument about the status of social scientists as scientists to make any sense at all, we must specify the term with which 'scientist' is being contrasted. The obvious candidate for the contrasting term is 'humanist', following Evans-Pritchard's suggestion. 'Humanist' is a term even more slippery than 'scientist' (Scott 1987a) and to do full justice to the contrast would require an unwarrantedly lengthy discussion of the distinguishing marks of the social sciences and the humanities as groups of disciplines. For present purposes, it must be sufficient to adopt the distinction I put forward in Chapter 11. The humanist endeavours to enter into the thoughts, perceptions and actions of the people under scrutiny, 'to see the world through their eyes', whereas the social scientist tries 'to develop systematic theories to explain the underlying causal interactions' among the same set of phenomena (cf. Fay and Moon 1977: 227). Where the humanist focuses on empathy, the social scientist seeks falsification, as expressed in typically robust fashion by Homans in his delineation of the hallmarks of a science: 'when nature, however stretched out on the rack, still has a chance to say "No!"-then the subject is a science' (Homans 1967: 4). In this sense there is no doubt that all but the most parochial social scientists are indeed scientists rather than humanists. 6 Professionalism and the praxis of social science The popular stereotype of a scientist is someone in command of a body of esoteric knowledge, able to pronounce authoritatively on relevant specialist questions, and a member of a delimited collectivity with shared understandings. This stereotype is held not only by the laity, but also by most natural scientists. The proper name of this stereotype is, however, not 'scientist' but 'professional', or more precisely 'a member of a learned profession' if we need to exclude professional criminals, professional snooker players and the like. For not only natural scientists but also lawyers, doctors and specialists of many other kinds conform to the
SOCIAL SCIENCE IN PRACTICE
13
stereotype. Given that the status of social scientists as scientists may be contested, we have then to ask whether their status as professionals likewise is subject to dispute? In this connexion it is interesting to note that almost all the discussions by philosophers about the nature of social science concentrate on theory construction, even though this activity contributes only a fairly small part to the total praxis of social scientists. Most of them, most or all of their time, are not engaged in theorizing at all; indeed many of them make no explicit use of social theory, even though they do in fact implicitly draw on the models and interpretations offered by one or other version of social theory. Mullan (1987: 5) has recently commented on the fact that a traditional mode of activity for sociologists is to carry out small-scale empirical research using survey techniques and interviews that have implications for public policy and which possibly are linked to middle-range theories. Indeed it is not so long ago that an earlier generation of empirical sociologists, perhaps less self-conscious than their successors today, were wont to deny that they had any need for theory at all. The picture is much the same for anthropology. Anthropologists win their spurs not by proposing developments in social theory but by gathering fresh data about social life at first hand, preferably under hazardous or discouraging conditions, and then reporting theirfindings.It is true that the ethnographic monograph without theoretical pretensions is no longer fashionable in social anthropology, and that new recruits to that discipline have much freer licence to indulge in theorizing than do their counterparts in sociology. But even so it is the contribution to the ethnographic corpus that constitutes the necessary ingredient which secures admission to a college of practitioners. For once, Kuhn's model of the scientific enterprise fits an aspect of the praxis of social science; most of what social scientists do is routine science, and consists of empirical inquiry rather than developing theory. The dominance of routine science in everyday praxis is even more marked in psychology and economics than it is in sociology or anthropology. Furthermore, it is routine science that employers in general are interested in. Sociology graduates, for example, who seek employment as sociologists outside tertiary education are judged by their ability to draw samples and to use the latest software for multiple regression rather than by their knowledge of the works of Althusser and Habermas. The concentration on fresh empirical findings rather than on fresh theories has important consequences for the relation between the social scientist and the general public. For although citizens, or 'mundane actors', as Giddens calls them, may be 'knowledgeable about what they do and why they do it' (Mullan 1987: 99), they are in general not knowledgeable about how to check whether others act in the same way, or whether
14
INTRODUCTION
their stated reasons for acting as they do are valid. In other words the relation between social scientists and the mundane actors, i.e. the laity, varies depending on whether we are looking at, on the one hand, social theory embodying in varying degrees the findings of social inquiry, or, on the other, the process of social inquiry itself. In the former case, there is the dialectical relation discussed earlier whereby social theory and lay perceptions of the social order mutually influence one another, and whereby the findings of empirical inquiry are continually absorbed into common sense. In the latter sense, however, the relation is much more lopsided. There is some diffusion of techniques from the experts to the amateurs, as evidenced by such things as self-assessment quizzes in popular magazines, and opinion surveys conducted by the secretaries of golf clubs, but in the eyes of the experts these are in general unsophisticated and not to be taken seriously. Likewise beachcombers and other untrained travellers may publish accounts of exotic communities which win popular acclaim, but to expert anthropologists these accounts are valuable only as raw data and are not mistaken for valid analyses. Worsley (1974: 4-5) has rightly complained about the 'mystique of professionalism' which leads some sociologists to think that they may have a monopoly on thinking about society, and thus unjustly to reject the interpretations offered by scholars from other disciplines. There is also, alas, a good deal of mysticism in this debased sense to be found with the techniques of empirical inquiry, for example in the attitudes of journal editors towards levels of statistical significance. Nevertheless, when we are dealing with techniques of inquiry and the presentation of results, we are on much firmer ground in being cautious about the credentials of those who would persuade us of the validity of their reports. Few of his colleagues would support the statement made by Donald MacRae, a senior British sociologist, that most empirical research in social science is easy, and 'could be done by well-designed mechanical mice' (MacRae 1970: 360). Even with a personal computer in every home and on every school desk, it seems to me highly unlikely that, for example, loglinear analysis will ever become part of common sense. The diffusion of investigative and analytic procedures certainly exists but is significantly less pervasive than the diffusion of technical vocabulary. Darrell Huff's book How to lie with statistics (1954) is unfortunately not as revealing as its title suggests, but his basic idea is correct. Nor is the possibility of misusing techniques of data analysis confined to quantitative data; qualitative data can also be manipulated in pseudo-scientific fashion, as the economist Ely Devons showed long ago in his book Planning in practice (1950). Nevertheless the distinction between making inquiries and utilizing results is easily and frequently overlooked, as exemplified by the title of a recent article in
SOCIAL SCIENCE IN PRACTICE
15
the Times Higher Education Supplement: 'We are all social scientists now' (Scott 1987b). Thus my thesis is that the basis of the claim by social scientists to be professionals lies not in the theoretical interpretations of the social world that they offer to the public but rather in the way in which they go about the task of testing these interpretations by empirical inquiry. Furthermore, their claim is reinforced by the fact that most of the time most social scientists are engaged on this empirically oriented task rather than on developing social theory. 7 Shifting identities If this thesis is correct, we might expect that social scientists, or at least all those who engage in empirical inquiry, would be united and consistent in claiming professional status. In fact this is far from being the case. In this matter there are differences between one discipline and another, and between one moment of time and another. This is not the place for a full review of the history of the social sciences, but a few highlights are worth mentioning. At one end of the continuum is psychology. I have already mentioned its bridging position between the natural and social sciences. As far as I know, psychologists have never had any doubts about their claim to professional status. More than any other social science discipline, psychology is marked by rigid and formalized conditions of entry to its professional organisations, and by attempts to enforce uniform curricula in training institutions. Effective restrictions have been placed on the sale of its tools of trade, notably IQ testing schedules, to non-members. Alone among the social science disciplines, psychologists have in some jurisdictions secured legal sanctions to protect their monopoly of the use of the disciplinary label, as for example in Victoria and more recently in the United Kingdom (THES 1988). At the other end of the scale economists seem never to have been interested in professional closure and each centre of teaching has felt free to determine its own syllabus and definition of the discipline. A few years ago in Britain, when a central body with funds at its disposal tried to establish certain uniformities in post-graduate teaching of economics, the move was perceived, at least in my own university, as a completely unwarranted and ill-informed attempt to curtail the autonomy of the faculty. With both sociology and anthropology the story is more complicated. When the Sociological Association of Australia and New Zealand was founded in 1963 anyone could join, and anyone still can. What is interesting is that many people hired as sociologists do not do so. The same phenomenon has occurred in Britain, though in that country there was for
16
INTRODUCTION
a while an attempt to establish within the umbrella organization a hard core of really professional sociologists, an attempt that failed (Barnes 1981c). In both countries a significant number of academic sociologists do not belong to the national association. Their reasons for not doing so vary from one individual to another but I think the common ground is a dislike or distrust of the version of sociology promoted by the national body concerned. The sociology of professionalism stresses the enforcement of closure, that is of restrictive recruitment, as one of the essential attributes of any profession, so that by its own criteria sociology, at least in Australia and Britain, does not qualify as a full-blown profession. In the United States sociologists are better organized, but even there there has been some ambivalence about claims for professional status. This was manifested in a delightful incident in 1958 at a national meeting of sociologists in Chicago. Following a paper by Talcott Parsons on 'Sociology as a profession', Everett Hughes spoke to insist that the organization arranging the meeting was not a professional body but, rather, a learned society. His intervention from the floor evoked 'a warm response'. Yet at the very same meeting the organization resolved to change its name from the American Sociological Society to the American Sociological Association, 'lest', as Gouldner (1962: 207) puts it, Its former initials evoke public reactions discrepant with the dignity of a profession'. We lack, unfortunately, any succinct account of the history of social anthropology in either Britain or Australia and indeed, with the laws of libel as they are, it would be foolish to hope for one. Nevertheless I must risk a few comments. If we think of anthropology as a discipline aspiring to professional status, we might well be surprised to find that in 1953 the president of the Royal Anthropological Institute of Great Britain and Ireland, at that time the largest and most prestigious association for promoting anthropology in the United Kingdom, should choose for his presidential address the title 'Anthropology as a hobby' (Mills 1953). The Institute's president held a readership in the University of London which he had acquired on his retirement from the Indian civil service. Until the demise of empire, the Institute was dominated by travellers, ex-colonial civil servants and philanthropists. It was not until 1946 that the Association of Social Anthropologists was formed as a deliberately closed body, aimed at securing any available university posts for academically trained anthropologists rather than for retired colonial district officers. The (British) ASA has remained a closed body and appears to have retained its comprehensive membership despite a major split over the intellectual, political and ethical merits and demerits of applied anthropology. A branch of the ASA, its only branch ever, was established in Australia in 1956 with entry conditions which, in effect, were even more restrictive than those in force in Britain, for Australian aspirants had
SOCIAL SCIENCE IN PRACTICE
17
to secure approval from a British-based committee often imperfectly acquainted with their work. There was a brief period, following a colonial revolt, when the branch transformed itself into the Australian Association of Social Anthropologists. However, the wave of epistemological populism which affected sociologists and anthropologists in Britain and elsewhere was felt also in Australia. The AASA soon changed into the Australian Anthropological Society, an explicitly non-elitist organization open to anyone interested in promoting the cause of anthropology and anthropologists. Then, with the dismantling in the late '70s of the Australian internal colonialist structure of Aboriginal protection and welfare, and the creation of possibilities for Aboriginal groups to acquire land rights, the market situation for anthropologists changed significantly. Whereas hitherto there had been few jobs for anthropologists as such outside the universities, now non-academic jobs as consultants to Aboriginal groups and mining companies came on the market. In some instances those hired as anthropological consultants had comparatively little formal academic training in the subject, and persons labelled as anthropologists found themselves confronting one another in court as expert witnesses. The more experienced anthropologists involved with Aboriginal land claims saw a need to protect the reputation of their speciality as well as to preserve the market for applicants with fuller academic credentials. As a first step in this direction they urged the conversion of the AAS back to a closed professional association. The proposed change was opposed by many of their colleagues who were not involved with Aboriginal claims for land rights, or with other instances of applied anthropology, but the change was eventually carried through. The moral to be drawn from this story is, I think, that, as with positions on the science/non-science divide, attitudes towards professionalism may depend as much on market forces as on intellectual considerations. 8 Citizens? Citizenship is a concept embedded in the discourse of sociology and political science, but the involvement of social scientists as citizens stems from the content of their studies as a whole rather than from their scholarly concern with the concept of citizenship. Natural scientists are, of course, like the rest of us, also citizens, but it is the exception rather than the rule that their attributes as citizens bear directly on their techniques of inquiry and their mode of presenting their findings. The exceptions, even though they are only exceptions, are nevertheless important, for in the natural sciences even the most unlikely discoveries in so-called basic research may later be found to have significant political and moral implications. We should remember that when Pearson published his
18
INTRODUCTION
Grammar of science (1892: 29-30), he chose Hertz's discovery of what we now call radio waves as an example of a result 'of striking interest to pure science' which 'seems yet to have no immediate practical application'. But, he said, 'It is impossible to say of any result in pure science that it will not some day be the starting-point of wide-reaching technical applications'. Scientific discoveries 'in the end profoundly modify the conditions of human life'. Since the 1930s (Kuznick 1987), and particularly since 1945, natural scientists have become increasingly conscious of the moral implications of their inquiries, and have formed bodies such as the (British) Society for Social Responsibility in Science and Scientists against Nuclear Armaments. With social science, however, the connexion between empirical research, theoretical interpretation and public policy is much closer. As indicated earlier, most empirical research in sociology is undertaken with some expectation that its findings will have a bearing on public policy. Even though these expectations are not always fulfilled, the sociologist contemplating a possible inquiry has always to decide whether the likely outcome is one that he or she as citizen would like to see implemented, and if it is not, whether the inquiry constitutes an instance of an unwelcome but necessary service provided by an impartial professional for a client who can pay for it. Similarly, at the level of theory, competing models of society are underdetermined by the available empirical evidence and, in choosing between them, social scientists cannot avoid being influenced at least partially by their views of what ought to be. For example, the controversy over Jencks' book Inequality (Jencks et al. 1972), about the relation between family background and educational attainment in America, illustrates how, when the evidence is unclear, social scientists tend to divide in terms of their political views of what ought to be the case. Alvin Gouldner (1962) argued strongly against what he described as 'The myth of a value-free sociology', and as long ago as 1944 Gunnar Myrdal called on all social scientists to introduce their 'valuations' into their analyses 'as explicitly stated, specific, and sufficiently concretized value premises' (Myrdal 1944: 1043). Myrdal's 'valuations' are of course moral, political, aesthetic and religious valuations as much as purely technical ones. The intimate connexion between the values of the citizen and the praxis of the professional is proclaimed even more clearly in Howard Becker's address with the uncompromising title 'Whose side are we on?' (Becker 1967). The connexion between political views and professional praxis is clearest of all in advocacy research, whereby the sociologist or anthropologist identifies herself or himself wholeheartedly, though perhaps also critically, with one side in a political struggle and endeavours to facilitate victory by the deployment of technical expertise (cf. Touraine 1965; Paine
SOCIAL SCIENCE IN PRACTICE
19
1985), the payoff to the social scientist being partly political and partly a greater understanding of the social processes of conflict. To some extent the connexion between the theoretical preferences of social scientists, the kinds of empirical inquiry they seek to undertake, and their political beliefs, is enhanced by the processes of recruitment to the various specialisms. The received explanation for the burgeoning of empirical inquiries in social science during the first half of the nineteenth century is that they were undertaken to facilitate the formulation of public policy at a time of profound social change. Likewise, in more recent times, the social sciences in communist-led countries are, as indicated earlier, used or discarded by reference to their utility for aiding the tasks of government. Thus, for example, the restoration of sociology as an academic discipline in China in 1979 has been justified by claiming that sociology 'can serve China by helping to maintain the social stability necessary for any political regime' (Cheng and So 1983: 485, citing Wang 1981: 26). In the West, however, most sociologists, to a greater extent than their colleagues in other social science disciplines, have long ago distanced themselves from the governmental embrace. As taught in most Western universities, sociology is intrinsically an iconoclastic discipline, perpetually querying the validity of cultural myths, so that it both attracts recruits who are already nonconformists and promotes scepticism and nonconformity among its students. Likewise anthropology, at least in so far as it retains its traditional role of studying others (Levi-Strauss 1963b: 361-2), tends to recruit those who already perceive themselves as marginal in their own societies, and certainly promotes this perception in those who study the subject. On the other hand my casual observations suggest that in Western capitalist societies economics recruits mainly students who accept capitalism and wish to learn more about how to operate successfully within it. Students are likewise selectively recruited to the various natural sciences but the selection appears to be based on attitudes towards structures of knowledge rather than on political attitudes to society. Despite these factors making for uniformity through selective recruitment, all the social sciences, from psychology to economics, are manned by practitioners who hold diverse political views as citizens and diverse theoretical preferences as scientists. It is, I think, this double diversity that continually hinders the development of strong professional associations in the social sciences. The fragility of these associations is shown not only by the significant number of senior social scientists who do not belong to the appropriate professional association or who stay away from its meetings, but also by the likelihood of protest resignations when a professional association takes, as a collectivity, a positive political stance. This phenomenon was seen on more than one occasion among American anthropolo-
2O
INTRODUCTION
gists during the course of the Vietnam War. In Australia differing attitudes to the war were reflected in tensions within the anthropological profession (American Anthropological Association 1976) and a mild instance of the same phenomenon occurred some twenty years ago when the Australian Institute of Aboriginal Studies awarded a research grant to an anthropologist regarded by the Commonwealth government as a communist (Barnes 1969a). Both anthropology and, to an increasing extent, sociology make use of the technique of so-called 'participant observation' for gathering data. Under these conditions all of the personal attributes of the social scientist gender, age, marital status, religious affiliation, and even physical ability may be relevant to the success of the inquiry. These attributes can be manipulated for purposes of the research only to a limited extent and impose an additional constraint, along with the limits already mentioned, stemming from the scientist's political beliefs, on the range of inquiries that will be undertaken. Participant observation, of all modes of inquiry in social science, shows most forcefully the inapplicability of the natural science paradigm to the practice of social science. 9 Scientists sui generis? The ways in which social scientists differ from scholars in the humanities, natural scientists and lay members of the public have been highlighted by the illustrations given so far. If instead we had chosen as our paradigm of social science a latter-day anthropologist reporting in empathic rather than analytic terms what he or she had been told about sorcery somewhere deep in the jungle, we would have seen little difference between the report and a contribution in the humanities. On the other hand, if we had made our paradigm a street survey of preferences in breakfast cereals, we would have discovered few if any differences between the actions of the interviewer with her, or more rarely his, clipboard and questionnaire schedule and the behaviour of a research assistant counting atomic collisions in a Wilson cloud chamber in a physics laboratory. Either job can be carried out irrespective of gender, religious belief or marital status, and neither has any apparent political implications. Social anthropologists collecting data in the field before 1945 tended to act as if they were conforming to the natural science paradigm, in what I have called a colonial situation (Barnes 1980a: 15), even if in fact there was more feedback from the observing scientist to the observed tribal peoples than he or she recognized at the time. Likewise, some of the modes of inquiry and analysis followed by social scientists are quite simple and can be replicated by any conscientious schoolchild; indeed school projects are often elementary excursions into social science. This is a consequence of the limited
SOCIAL SCIENCE IN PRACTICE
21
diffusion of techniques of inquiry which, as noted earlier, accompanies the more widespread diffusion to the laity of the findings and, above all, the technical vocabulary of social science. Even in economics there are, or at least were until recently, some practitioners who hold that elaborate techniques of analysis are not essential. The late Nicky Kaldor, the eminent economist, reluctantly conceded that an understanding of percentages was necessary for an economist but resisted attempts to require higher mathematical qualifications. Among sociologists and anthropologists, resistance to sophisticated and in particular quantitative techniques of analysis is much more widespread, as demonstrated for example in the resistance to developing network analysis with relatively rigorous mathematical techniques (Barnes and Harary 1983c). Yet those social scientists who resist quantification are likely to insist on the need for other specialized qualitative techniques of inquiry and analysis. Only a few are willing to proclaim, as John Rex once did (Rex 1975: 10), that anybody can do sociology. Indeed, eight years later, when he was challenged on this statement, Rex shifted its meaning to imply that anybody could get a job in sociology, or rather almost anybody. He replied to the challenge by saying that it's actually very difficult nowadays for anyone who's a sociologist to get a job in a sociology department, because on the whole there are all sorts of deviants and criminals and gangsters-to exaggerate somewhat - who have moved in and I suppose it's a case of the lunatics have taken over the asylum (Mullan 1987: 32). Even when we allow for Rex's exaggeration, his comment shows that, in British sociology, professional closure still has a long way to go. The discussion has, I hope, shown that social scientists are indeed scientists rather than non-scientists but of a variety different from that to which natural scientists belong; that some of them are professionals, either because they possess special skills or belong to closed guilds or both; and that most of them are influenced in their professional activities by their attributes as citizens. They are, as it were, scientists sui generis, but none the less scientists for all that. In their discussion of the interstitial status of social science, Fay and Moon (1977: 216) contrast their own view with that of humanists, who treat social science as part of the humanities and therefore not scientific, and with the view of naturalists, who hold that there are no essential differences between social and natural science. They say humanists have failed to appreciate the explanatory task of social science (i.e., they have failed to see in what way these disciplines are scientific), and ... naturalists have misunderstood the crucial role which interpretation plays in the social sciences (i.e., they have given insufficient or misleading analysis of what it means for a phenomenon to be intentional).
22
INTRODUCTION
In 1983, Keith Joseph, while he was the minister responsible for the (British) Social Science Research Council, decreed that social science was an impossibility and forced a reluctant council to remove the word 'science' from its title. But social science persists, even in Britain, despite this attempt to exorcize it by ministerial magic, and so do the scientists who are its practitioners. 10 Outlining the model What then are the main features of the model of social science that emerge from the preceding discussion? First there is the assumption, philosophically enormous yet trivial in terms of common sense, that there is a real world and that we can discover a certain (or, more correctly, an uncertain) amount about it. Reality may properly be said to be socially constructed, or even, as in the proclamations of the ethnomethodologists, individually constructed; but, to adapt a famous phrase from Marx, we do not construct it quite as we please. Some facts are harder than others, though I doubt if many social scientists would follow the philosophers in putting primary sense data at the hard end of the scale; the number of people in the house at midnight on census night and the ability of the police to make arbitrary arrests are stronger candidates. Data can be collected which reflect, however imperfectly and incompletely, 'what actually happens'. Then the fun begins, at least for some. For others, the fun lies only in the process of data collection, particularly if participant observation is the mode of inquiry followed; for them, what follows is a necessary but burdensome chore. For social scientists of both types the stage of analysis is arduous, even though for some it is enjoyable. It is possible, particularly if the data are quantified, to engage in what is usually known as datadredging, the relentless search for correlations between variables, followed by the formulation of explanations for the correlations that have been detected. Most textbooks on methods of data analysis draw attention to the hazards of this procedure, but the dangers are most succinctly identified in a remark made many years ago by Kuczynski (1937: xiii): 'an explanation does not necessarily look less plausible if the event which is explained has not occurred'. Given a corpus of data, whether qualitative or quantitative, careful thought often enables the analyst to discern systematic regularities and variations between similar events that have occurred at different times and different places, and that have involved different communities and societies. The way is then clear for the construction of a model, a drastically simplified version of the diverse actual manifestations of the phenomenon under scrutiny. How well, or how poorly, the model fits the data set can sometimes be expressed quantitatively, as for example when the model is a
SOCIAL SCIENCE IN PRACTICE
23
path diagram; but even if, as with all the models displayed in this book, a precise measure of goodness of fit is not feasible, it is always possible to confront the model with additional data. The confrontation may merely increase the analyst's confidence in the plausibility of the model, or lead to its elaboration to accommodate the new data, or simply to a closer specification of its limited applicability. More importantly, the model suggests likely causal connexions that, at least in principle, can be investigated further. Suppose, for example, that the model implies that, other things being equal, if a man's mother was a university graduate, his salary on entering the paid workforce is likely to be greater than would otherwise be expected. The next step is to seek to discover what it is, if anything at all, that graduate women do that results in their sons getting better-paid jobs. The correlation may be valid or spurious, but if it is shown to be valid, nothing is adequately explained until we have identified the mechanism that establishes the causal chain. Likewise, if the model suggests, as for example in Chapter 6, that frontier contacts between indigenous agricultural populations and colonial administrations led to less social disruption than similar contacts with invading European settlers, the next step is to inquire into the reasons for the difference and to establish how the two types of external impact impinged on the various features of indigenous society and culture. A model that does no more than exemplify connexions between various categories of data, but which says nothing about how these connexions arise and are sustained, is unlikely to be controversial. Some essentially descriptive models do indeed make causal statements, but few social scientists would be prepared to go to the stake to defend, for instance, the direction of the causal arrows in their path diagrams. Serious controversy is more likely to arise when explanations and interpretations are added. Does the rain-making ceremony cause the rain to fall, as the participants in the ceremony maintain, at least in retrospect, or is the ceremony held only because the black clouds indicate that rain is a possibility? Are the explanations offered by actors the only ones that count, since these are the understandings and perceptions in terms of which the actors act, or are these not reasons but rationalizations, designed to conceal rather than reveal the truth? Or should we give priority to the observer's explanation, by granting to social science the same privileged status that at least some writers ascribe to natural science, when it is contrasted with folk perceptions of the physical world? The current literature on the philosophy of the social sciences provides a cafeteria of answers to these questions (cf. Fay and Moon 1977: 209), and this is not the place to run through the menu. I hope that what has been said in earlier sections of this introduction, and what follows in the various chapters, indicate the choice I have made. The earlier positivists tended to
24
INTRODUCTION
give little attention to the dispositions of the actors; the later ethnomethodologists tend to pay little attention to anything else. I have tried to steer a middle course, claiming provisional privilege for the causal connexions detected by the trained outside analyst; this is the only product that the social scientist can offer the world. But social scientists cannot claim to be, as Mannheim (1936: 143) hoped intellectuals would be, 'watchmen in what otherwise would be a pitch-black night', for their candles are feeble and flickering. In any case, the night is not pitch-black; there are other candles held by artists and poets, and most of the laity know their way around, despite the poor lighting. For I assume that actors are at least partially correct in their perception of their own actions; as a methodological principle I assume that they are as muddled as I know myself to be about what I do, but not much more muddled. This might be seen to be close to the 'principle of humanity' proposed by Richard Grandy (1973: 443), a methodological assumption that 'the imputed pattern of relations among beliefs, desires, and the world be as similar to our own as possible'. But I take a less charitable view about the distribution of rationality than I think he holds. I agree with Fay and Moon (1977: 2 2 3 4) when they say that the social scientist cannot confine himself to explicating the way in which the actors' concepts and self-understandings form a coherent system. In order to understand these cases, the social scientist must recognize how the actors' selfunderstandings are incoherent, and he must show what consequences these incoherencies have. They limit this prescription to those cases where 'people systematically misunderstand their own motives, wants, values and actions, as well as the nature of their social order' but I would prefer to assume that unsystematic misunderstanding exists everywhere until empirical inquiry shows otherwise. A somewhat similar stance is taken by Macdonald and Pettit (1981: 54) when they comment on the tendency of anthropologists, by focusing on the contrast between science and religion, to overemphasize the difference between modern and traditional cultures. They say Thus it may be that the anthropologist, or at least the atheistic anthropologist, will find the sharpest challenge to the assumption of human rationality, not in researches among the alien, but in investigations among her own. The procedure of building explanatory models, which I advocate and which I hope is demonstrated in the chapters that follow, is far removed from the search for elusive social laws that for earlier generations of social scientists served as a badge of scientific respectability, without greatly affecting their investigative praxis. It is also unlike the quest for empirical generalizations, as exemplified by Berelson and Steiner's Human behavior with its 1045 'scientific findings', systematically numbered but only
SOCIAL SCIENCE IN PRACTICE
25
weakly connected logically (1964: 659). It is even further from the relentless data-dredging exemplified by Textor's Cross-cultural summary (1967), with its 20,000 statistically significant correlations. Model building aims instead at providing a summary account of how a limited collection of social processes, attributes, perceptions and events are interconnected logically and in time and space. Furthermore the construction of models can become an activity in its own right, independent of any data set, a development we examine at length in Chapter 13. What I have presented in the introduction, and particularly in this last section, may be seen as an attempt at a model of the inquiries and analyses set out in the following chapters. The reader must judge how well the model fits the data.
Parti
MODELS OF THE REAL SOCIAL WORLD
GENETRIX: GENITOR:: NATURE: CULTURE?
The seven chapters in the first part of this collection present models of the real social world outside the academic ivory tower, while in the second part the chapters discuss ideas and processes that belong mainly to the relatively specialized arena of social inquiry. Given this initial focus on reality, we may as well plunge in at the deep end. Of all the varieties of social interaction there is nothing more closely tied to the material world than the relations of kinship. Copulation and birth, the two activities which sustain even the most culturally elaborated or attenuated kinship systems, are the processes above all others where the similarity of human beings to other mammals is unmistakable. There are innumerable accounts of religion and even politics that are cast entirely in transcendental or ideational terms, and which ignore completely human membership of the animal kingdom. Throughout the ages these accounts have enjoyed widespread acceptance. Even in the study of kinship, particularly when this is reduced merely to the study of kinship terminologies, the fact of human physicality is often ignored or silently taken for granted, though it is seldom deliberately excluded on methodological grounds. Meyer Fortes, in whose honour the present paper was written, took the material world well into account in his analysis of tribal societies in West Africa. Yet even he, in the eyes of some of his critics (cf. Worsley 1956), sometimes tended unjustifiably to favour transcendental explanations for human actions. In particular, he tended, so I thought, to posit the notion of fatherhood in absolute terms, on a par with the undisguisable fact of motherhood, without paying sufficient attention to the 29
3O
MODELS OF THE REAL SOCIAL WORLD
differences in the physical processes on which these two types of social relation are culturally constructed. For these reasons a paper about fatherhood seemed to me an appropriate topic for my contribution to a festschrift for Fortes. The form my paper took was influenced by the publication of Schneider's book on American kinship (1968), with its challenging subtitle A cultural account; he provided me with a convenient target to tilt at, and I used his book again in the same way a few years later when writing more generally about kinship studies (Chapter 10). Schneider continues to campaign vigorously for his standpoint (1984). I was somewhat apprehensive about Fortes' reaction to my paper; was I breaking the conventions of festschriften by criticizing the person being honoured, one of my own teachers? I need not have worried, for in a monograph published a few years later he wrote that my discussion about the difference between motherhood and fatherhood fitted well with his own argument (Fortes 1983: 44). Roy Wolfe, a member of the graduate student seminar at which I read my paper, thought that my treatment of the concept of fatherhood would be enhanced by adopting the notion of scientific research programmes advocated by Lakatos. His critique of my paper (Wolfe 1974) led me to defend my old-fashioned empiricism in a paper entitled 'Positivism in the womb' (Barnes 1974a).
We salute Meyer Fortes for his achievements not only in the intensive investigation of the Tallensi and Ashanti of Ghana but also in the comparative analysis of diverse social and cultural systems, notably in his Lewis Henry Morgan lectures published as Kinship and the social order (1969). Although he has long remained a steadfast defender of the strategy of concentrating anthropological field resources on the study of peoples whose material artefacts are simple (Fortes 1958b), he has always been sensitive to the light these studies shed on patterns of living found in industrial societies. It is therefore appropriate in the present context to discuss some of the issues that arise in kinship studies when we endeavour to compare social and cultural patterns in many different societies including our own. My thesis is that this comparison suggests a reformulation of the relation between kinship and nature. I focus on putative physical relations rather than on relations of social parenthood. It can be argued that in anthropology and sociology comparative analysis is impossible without including, either explicitly or by implication, the society to which the analyst himself belongs and the culture whose concepts and categories he uses to think with. This view has been
GENETRIX : GENITOR :: NATURE I CULTURE?
31
expressed by Schneider (1972: 47-8) who, in recent years, has appealed forcefully for the study of kinship as part of a cultural or symbolic system and who has provided the most uncompromising account of such a system in the Western tradition. He says: The next problem ... is the old one of how comparison can be conducted on a cultural level if it is assumed that each and every culture may be uniquely constituted ... our own culture ... always serves as a base-line for cross-cultural comparison. Without some comprehension, however botched, distorted, biased, and infused with value judgments and wishful thinking, both good and bad, our own culture always remains the baseline for all other questions and comparisons. In part, this is because the experience of our own culture is the only experience which is deep and subtle enough to comprehend in cultural terms, for the cultural aspects of action are particularly subtle, sometimes particularly difficult to comprehend partly because they are symbols not treated usually as symbols but as true facts. Schneider implies, so it would seem, that even when we are comparing, say, unilineal systems found in different parts of Africa, as in Fortes' classic paper on 'The structure of unilineal descent groups' (1953), there is an implicit comparison between the various African notions of unilineal descent and filiation and similar notions current in the Western tradition. In the passage cited, however, Schneider is talking about cultural symbols, and it is not clear whether he would argue that in comparing, say, forms of social organization we are similarly forced to begin our analysis with forms prevalent in our own society. Indeed, at other places in his paper he draws a distinction between, on the one hand, 'the scientific facts of biology' and 'biology as a natural process' and, on the other, the cultural symbols that may perhaps (or perhaps not) be derived from these scientific facts. This suggests that he sees science as the study of nature, i.e., as natural rather than social science, and that he makes a distinction between 'science', dealing with facts, and 'culture', dealing with symbols. But if there are facts of nature and biology that can be demonstrated scientifically, as he maintains (he seems to have the processes of human reproduction in mind), then presumably there are other scientifically ascertainable facts about where people live, who they work with, who commands whom, and so on, which can provide a framework for a comparative analysis of social organization that is not linked distinctively with any particular society, not even our own. The distinction Schneider draws is widely used in social science. It is somewhat akin to the contrast between objective and indexial meanings used by the phenomenologists, and to that between 'objective' and 'subjective' social class by students of stratification. The same distinction is presented in another form in what Naroll (1964: 306) calls Goodenough's rule: what we do as ethnographers is, and must be kept, independent of
32
MODELS OF THE REAL SOCIAL WORLD
what we do as comparative ethnologists (Goodenough 1956: 37). The closest analogue to Schneider's contrast is found in the distinction between etic and emic categories, labels which some social anthropologists have taken over from linguistics (see Goodenough 1970b: 98-130). In Pike's (1967: 37-72) formulation the etic-emic contrast is unashamedly positivist. The scientific linguist observer, with his objective categories, is contrasted with the speaking actor who uses subjective categories to produce and decipher meaningful utterances. Inter-language comparison is implicit in Pike's scheme. As used in anthropology the cross-cultural and cross-societal emphasis has been retained but the positivist implication of the contrast has been played down. Instead we have the emic categories of thought of the actors contrasted with the etic categories of analysis of the observer, neither set necessarily more real or true than the other. I have argued that this contrast can be applied without modification only in the 'colonial' or laboratory situation such as Pike had in mind (Chapters 5, 8 and 11). If the actors speak only their own language, think only in their own terms, and draw only upon a locally generated stock of 'knowledge' of their environment, then theflowof information is only oneway. The observer may well modify his etic analytical categories in the light of what the actors do and say, but they do not alter their ways of thinking and acting because of assertions made about their behaviour by the observer. This is the paradigm situation of inquiry in natural science, the principle of indeterminacy notwithstanding. Until a couple of decades ago anthropological fieldwork in distant colonies approximated to it, though even then there were substantial and critical differences from the typical scientific laboratory. These laboratory-like conditions have not persisted and, following the end of colonialism in its classic form, they are probably now gone for ever. Instead there is two-way communication between actors and observers, so that the actors begin to take over not only the material artefacts brought by the observers and their compatriots but also their languages, concepts and social institutions, changing them in the process. In particular they take over and adapt some of the jargon and some of the content of science. Once this happens the observer in the remotest jungle begins to face the same difficulties as his colleagues working in the metropolis have always faced: the facts that the language of science, and of social science in particular, is also to some extent the language of the people and that the findings of science, and even its techniques of inquiry and verification, are continually seeping into popular consciousness. In general, then, though his reasons are different from mine, I accept Schneider's point that the categories and concepts of the observer's own culture are the starting-point for comparative analysis in social science. If this is so, how does it affect the study of kinship? There are many issues
GENETRIX : GENITOR :: NATURE *. CULTURE? we might take up but I want here to consider just two related matters: how valid is the distinction Schneider draws between culture and natural science: and how does kinship, in contrast to other aspects of social and cultural life, relate to nature? I use the standard triple distinctions between genetic or carnal father, genitor and pater, and between genetic or carnal mother, genetrix and mater, stressing that the statuses of genitor and genetrix are defined, if at all, in terms of local doctrines about the process of human reproduction (Barnes 196id: 297-8, 1964b: 294; Goodenough 1970b: 27). Fatherhood and motherhood are used as cover terms. At first glance Schneider's position seems to be paradoxical. He seeks to establish science as distinct from culture and yet to insist that a comparative science of cultures has to be rooted in a particular culture, the culture of the investigator. He appears to make natural science free of culture but to query the possibility of metacategories for analysing cultures. But this apparent paradox can be quickly disposed of by referring to his book on American kinship, where he makes a fourfold contrast between (1) what he calls biological facts, (2) formal science, (3) informal ethnoscience and (4) 'certain cultural notions which are put, phrased, expressed, symbolized by cultural notions depicting biological facts, or what purport to be biological facts' (Schneider 1968: 114-15). I find it confusing to use 'biology', the name of a science, for phenomena that exist independently of efforts to study them, and therefore re-label (1) nature. Category (4), of which 'broken heart' and 'heartache' are examples from American culture, need not detain us. This category contains what in more traditional language might be called extensions of kin usages that are perceived by the actors as being metaphorical, figurative, symbolic; all the parishioners know that the village priest is not 'really' their father. We can concentrate on categories (1), (2) and (3). Formal science, category (2), is part of American culture as much as categories (3) and (4). Indeed Habermas (1972) argues that the salient diagnostic feature of contemporary culture in industrialized societies is the belief that science is the only authenticated form of knowledge. Though both are part of Western culture it is possible, at least for classical times and since the Renaissance, to draw a fairly clear distinction between professional scientific assertions and lay beliefs that, rightly or wrongly, are perceived as based on formal scientific inquiry (see S. B. Barnes 1969). Informal ethnoscience embraces more than the latter category but it certainly includes it. Weber (1946: 139) notes this distinction, and it is well put by Evans-Pritchard, in a discussion of the views of Levy-Bruhl. The fact that we attribute rain to meteorological causes alone while savages believe that Gods or ghosts or magic can influence the rainfall is no evidence that our brains function differently from their brains ... It is no sign of superior intelligence on my part that I attribute rain to physical causes. I did not come to this conclusion
33
34
MODELS OF THE REAL SOCIAL WORLD
myself by observation and inference and have, in fact, little knowledge of the meteorological processes that lead to rain. I merely accept what everbody else in my society accepts, namely that rain is due to natural causes. (Evans-Pritchard 1934: 21) Our yardstick, then, is our own culture, which contains a vast number of propositions perceived as science. Against it we compare other cultures, noting in what respects they resemble one another and how they differ, and endeavour to discover why this is so. How does a comparison of this kind work in the field of kinship? It is reasonable to expect that data from category (1) will impinge in special fashion on kinship data from categories (2) and (3). Despite the recent efforts of some ethologists to postulate a pan-primate basis for political order, and for much else as well, kinship remains the aspect of human culture with the closest links to the natural world. Indeed, in American culture, we are told, 'kinship is biology' (Schneider 1968: 116). Apes and monkeys may have dominance hierarchies and territories but, unlike men, they do not have representative government nor, as far as we know, do they believe in God. Like us, however, they copulate, conceive and give birth, activities with which kinship has a close connexion, however problematic the qualities of the connexion may be. These activities, when performed by humans, are perceived as natural rather than cultural. Part of the basis for a comparison of ideas of kinship has then to be our own cultural notions about the reproductive process, some of which are derived directly from formal science but which include others that belong solely to ethnoscience. The inevitability of beginning cross-cultural comparison by matching alien cultures against our own is well shown by the discussion in Man a few years ago on virgin birth (Leach 1967a; Spiro 1968; Douglas 1969 and references therein), and by earlier controversies about the ignorance of physiological paternity. The diverse beliefs about non-miraculous human reproduction found in pre-scientific cultures have been described many times and need not be repeated here (Ashley-Montagu 1937, 1949; Ford 1945; Leach 1961a: 8-26, 1961b, 1967a; Malinowski 1963; Meyer !939- 1-16; Spencer 1949-50). The point I emphasize is that when these beliefs are compared, the yardstick used is falsely presented, for we tend to assume that for ourselves no distinction between formal science and informal ethno-science is needed. We present our own view of conception as a single event, in which only one man and one woman are involved, and which triggers the whole sequence of gestation, as scientifically validated. We contrast this view with theories that the foetus forms and grows in the womb by receiving contributions via many acts of coitus not necessarily all performed by the same man, a view held, for example, by the Azande (Evans-Pritchard 1932: 410); or with other theories, found for
GENETRIX I GENITOR II NATURE I CULTURE?
35
example in Aboriginal Australia, whereby the process of gestation is neither initiated nor sustained by coitus; or with intermediate theories. These indigenous ideas are recorded in the ethnographic literature, but where do ours come from? From formal science, or informal ethnoscience, or from a cultural heritage in which natural substances like blood and semen serve as symbols in statements that have nothing at all to do with natural science? Even if we prune away metaphorical ideas in Schneider's category (4), a moment's reflection shows that ideas in categories (2) and (3) are not as easy to pin down as may seem at first sight. At this point we can come to grips with a distinction between fatherhood and motherhood. Consider first fatherhood. Nowadays most educated people in the West have heard of genes and chromosomes and know that the embryo draws its stock of chromosomes equally from its genetic father and mother. I guess that, in the sex-conscious culture of contemporary Britain, almost all adults believe that conception occurs when a spermatazoon penetrates an ovum. But what sort of knowledge is this? Surely most of us know as little about the physiology of human reproduction as Evans-Pritchard knows about meteorology. We believe these processes to occur because we believe also that at some point in the past long-forgotten scientists discovered that this is what really happens. We assume that though the discovery of genes and chromosomes is postDarwin, the fact that conception is a unique event and not a prolonged process has been scientifically established for a long time. The view that conception and gestation can follow a single act of coitus is indeed consistent with Aristotle's account of reproduction in Generation of animals, Book 2, and is implied in his statement in the History of animals that 'if the second conception take place at a short interval, then the mother bears that which was later conceived and brings forth the two children like actual twins ... The following also is a striking example: a certain woman, having committed adultery, brought forth the one child resembling her husband and the other resembling the adulterous lover' (585a). Thus the doctrine of 'one child, one genitor' has been part of the Western tradition for more than two thousand years. Yet although the presence of physical resemblances between some, though not all, children and their mother's husbands calls for an explanation, it does not necessarily demand a theory of universal monopaternity. The dominance of a monopaternal theory cannot have been determined by the weight of evidence, for apart from resemblances there was little material evidence available until the seventeenth century. Spermatozoa were discovered accidentally in 1677 by Ham, though their connexion with fertilization remained unknown. Mammalian ova were discovered, also accidentally, in a pregnant bitch by van Baer in 1828 and in 1853 Newport claimed to observe spermatozoa entering an ovum. Not until 1875 were the male and female pronuclei in
36
MODELS OF THE REAL SOCIAL WORLD
spermatozoa and ova identified by Oscar Hertwig, who described how they combine (Meyer 1939:123, 137-8,189-92). Thus for most of the historic period in the West, the uniqueness of physical paternity was a cultural construct for which there was very little conclusive evidence. Even so, this doctrine was modified by a belief in 'maternal influences', the idea that events experienced by a pregnant woman are reflected in the constitution of her child. The belief forms part of several indigenous theories of procreation (e.g. Levi-Strauss 1966b: 76) and is certainly still present in contemporary Britain. It is exemplified for animals in the story told in Genesis, chapter 30, verses 25-43, about Jacob changing the colour of the lambs borne by Laban's ewes. 'Maternal influences' may always have been restricted to ethnoscience, old wives' tales, but orthodox formal science long entertained the related idea that Weismann (1893: 383) calls 'telegony', the notion that the physical characters inherited by an individual are influenced not only by his (or her) own father but also by other men by whom his mother may previously have had children. Dobzhansky (1970: 420A) attributes this belief to Aristotle and it was supported, for plants and animals at least if not for humans, by Darwin (1875: 435-7; see Morton i82i;Zirkle 1935: 117 and 1946: H9;Parkes i960: 242) in conformity with his thesis of pangenesis. Thus whereas most pre-scientific beliefs about multiple physical fatherhood identify as genitors men with whom a woman has had intercourse during a given pregnancy, telegony ascribes physical paternity to her earlier mates as well as to the man who initiates the pregnancy. The doctrine of telegony lives on among animal breeders but has been abandoned by orthodox science, as has a later suggestion of a naturally occurring polypaternal process called 'somatic fertilization'. According to this hypothesis, substances may be absorbed in the female genital tract after copulation; these evoke the production of factors which may exert an influence on the embryos of subsequent matings (Austin and Walton i960: 393; Parkes i960: 242). In the laboratory, however, the fusion of two embryos at the eight-cell stage has been achieved, producing tetra-parental mice. Chimeric mice with even more complex constitutions have been bred and studied (Tarkowski 1961; see Wegmann 1970; Mullen and Whitten 19 71 and references therein). Indigenous assertions of human polypaternalism in nature have thus been vindicated for some mammals in the laboratory. Indeed there is evidence that double fertilization sometimes occurs naturally in humans (Benirschke 1970: 40-5). Human polypaternalism seems therefore to be compatible with the available scientific evidence. Tetraparental mice and other chimeras produced in the laboratory receive their diverse constituents before the implantation stage, long before birth. A belief in the post-natal physical transmission of information and attitudes is implied in the expression 'He took that in with his mother's
GENETRIX : GENITOR '.'. NATURE : CULTURE?
milk'. An earlier belief in a more specific and selective form of located transmission is suggested by Dobzhansky's (1970: 420A) statement, made in the context of an article on heredity, that 'An ancient English law holds a man who seduces the wet nurse of the heir to the throne guilty of polluting the "blood" of the royal family'. I have been unable to trace this law. The closest comparable laws seem to be those listed during the reign of King /Ethelberht of Kent about A.D. 600, whereby a man who seduced a maiden of the king's household had to pay fifty shillings in compensation, compared with only twelve shillings for the seduction of a girl occupied on menial tasks (Attenborough 1922: 5; Leibermann 1903: 3 and 1916: 7). These laws give special recognition to the king's entourage but make no reference to suckling or pollution. It may well be that the ultimate source for the alleged ancient law is merely the Mirror of justices, where it is said that one of the ways in which an adulterer may commit the crime of lesemajesty, 'a horrible sin', is by seducing the nurse suckling the heir of the king (Whittaker 1895: 15). The Mirror was at one time regarded as a true account of the laws of England before the Norman Conquest but in Maitland's view was largely fabricated by Andrew Horn, fishmonger and Chamberlain of the City of London, in about 1289; it contains many wilful falsehoods and misstatements of law (Maitland 1895). The Anglo-Saxons may never have held the doctrine that some kind of malign influence can be transmitted from a man by adulterous copulation to a lactating woman and thence through her milk to her royal foster-child. But if the law never existed, at least the doctrine formed part of the imagination of a thirteenthcentury fishmonger. Despite these contrary notions, the main stream of Western popular belief has clearly been 'one child, one genitor'. If there was no compelling scientific evidence for this belief the reasons for its persistence must be sought elsewhere, in the organization of social life and in other parts of Western culture, rather than in nature. As far back as we have knowledge, Western society, like most other human societies, has been organized on the premise of one child, one pater. Likewise the Christian faith of the West stresses the uniqueness of God the Father. The Holy Ghost impregnated Mary through her ear and was manifest in, or symbolized by, a dove at Christ's baptism, but neither act makes the third person of the Trinity copater with the first (see Jones 1951; Swete 1909: 28-9, 45, 365-6; Gudeman 1972: 54). If we encountered this constellation of facts in a tribal society, surely we would have no hesitation in saying that the organization of society and the major premisses of religion are reflected in myths about unique physical parenthood. Motherhood is different. Conception is an internal and microscopic event that we laymen believe scientists have investigated, whereas gestation and birth, and with them the relation of physical motherhood, are
37
38
MODELS OF THE REAL SOCIAL WORLD
macroscopic processes that, in principle, anyone can see for himself. Hence the descriptions of physical motherhood in diverse cultures do not vary as greatly as with fatherhood. The so-called denial of physical maternity is not homologous with the denial of paternity, except when applied to special myths for uninitiates, as for example in our own tale for children about storks bringing babies (cf. Spiro 1968: 260 n. 11). The denial of physical maternity usually means merely that the mother is thought to contribute nothing of importance to the foetus during pregnancy, as for example was believed in ancient Egypt (Needham 1959: 43) and is stated by Apollo in Aeschylus' Eumenides (lines 657-61), when defending Orestes against the charge of matricide. This lack of symmetry between the notions of genitor and genetrix is emphasized by Goodenough (1970a: 392), who says that 'procreation associates children directly with women but only indirectly with men' and that 'Motherhood and fatherhood cannot be defined in the same way for comparative anthropological purposes'. Fathers are not self-evident as mothers are. 'Genitor' is a social status, and societies vary greatly in the rights and duties, privileges and obligations, if any, that they associate with this status. If the status exists, there must be a rule for identifying genitors. But for the status to exist at all there must be a theory of procreation that calls for one, or for several, and, for all cultures prior to the physiological discoveries of the late nineteenth century, this theory cannot be supported by scientific evidence. Even though Aristotle wrote his Generation of animals in terms of a unique and necessary genitor, who might be the wind rather than an animal, he misunderstood the significance of menstrual blood, which he thought was coagulated by semen just as milk is coagulated by rennet, thus forming the foetus (739b). It is scarcely surprising that Australian Aborigines and many other prescientific peoples should have developed theories of human reproduction which do not require a genitor or which allow for the possibility of several. What calls for explanation is why in the pre-scientific West the dominant folk theory happened to be in one particular, though not in many others, more or less in accord with evidence from nature later to be disclosed. In this light, the debate between Leach, Spiro and others about ignorance of physiological paternity is cast in the wrong mould, for their arguments are all about how to interpret correctly apparent ignorance of a fact that everyone should know. Against the participants in this debate, on both sides, I contend that physical paternity is a fact that, until recently, nobody can have known scientifically. Our proper task is to explain the fabrication of flimsy hypotheses as well as the denial of material evidence. Schneider (1972: 62 n. 9) queries the assumption that American cultural symbols like blood and shared biogenetic substance, and perhaps even coitus, derive from the facts of nature. Beliefs centring on these
GENETRIX : GENITOR :: NATURE I CULTURE?
symbols presumably belong to informal ethnoscience, where the predominance of cultural rather than natural influence is not surprising. For example, in the fourth century B.C. Anaxagoras stated that sperm coming from the right testis produced males, and from the left testis females, an assertion repeated in the sixteenth century by Melanchthon, Luther's supporter, with the rider that males were born from the right side of the womb. What is more impressive is the effect that cultural influences, usually in the form of adherence to unproved theories, have had on formal science as well as on ethnoscience in blotting out the evidence provided by nature. The most striking examples are given at the end of the seventeenth century by the homunculi, minute but fully formed human beings, which Plantade and Hartsoeker separately asserted they had seen through their magnifying lenses inside human spermatozoa (Meyer 1939: 69-70, 133, 152, and figures 16 and 17). Even Leeuwenhoek, who reported Ham's discovery of spermatozoa to the Royal Society of London, wrote about the 'nerves, arteries, and veins' he saw inside his own spermatozoa: 'I felt convinced that, in no full-grown human body, are there any vessels that may not be found likewise in sound semen' (Cole 1930: 12). The history of popular and professional scientific beliefs about monsters, malformed foetuses, provides further proof of the difficulty we encounter in recognizing the evidence of nature when this challenges doctrines we cherish (Meyer 1939: 212-17). Kuhn (1970) and many other writers have drawn attention to the way in which fresh evidence from nature is moulded as much as possible to fit existing scientific theories. Without necessarily accepting Kuhn's notion of a paradigm, we can apply to the scientific quest for physical fatherhood in general Needham's comment on Aristotle's account of human reproduction: 'The whole matter affords an excellent illustration of the way in which an apparently academic theory may have the most intimate connections with social and political behaviour' (Needham 1959: 14). From this standpoint we can easily resolve the paradox that Aboriginal Australia, the major locus of so-called ignorance of physiological paternity, is also the home of what Fortes (1969: 101) calls a kinship polity (Barnes 1963^ xxiii-xxvii). For if of necessity physical paternity is prescribed or denied culturally without the constraint of the natural order, the way is open for the elaboration of rules of fatherhood for any social or cultural purpose whatever. Aboriginal cultures seem generally to have managed without human genitors, while ascribing a relation of social fatherhood to the mother's husband (see Fortes 1969: 106 n. 10). Indeed, Hiatt (1971) analyses secret pseudo-procreative rituals performed by Aboriginal men in terms of the contrast between the uncertainty of the male contribution to reproduction and the certainty of the contribution made by women. In this
39
4O
MODELS OF THE REAL SOCIAL WORLD
perspective we might see all assertions of physical paternity as examples of what is fashionably called male chauvinism. In some Aboriginal societies where many marriages are unorthodox, and also among the very orthodox Walbiri, the required relations between sections, lines and generations are maintained by applying rules of indirect matrifiliation rather than patrifiliation. The unorthodox affiliations of an individual's father are ignored and he acquires the category and group memberships he would have had if his mother had made an orthodox marriage. A rule of indirect matrifiliation in a 'kinship polity' of Australian type reduces the range of contexts in which an individual needs a specified social father; a dependable prospective mother-in-law may be a more important requirement (see Warner 1958: 119-20; Meggitt 1962a, 1972: 74; Shapiro 1971, 1981: 46-74). He needs a single human genitor even less. It is perhaps possible that the lack of interest in nominating a physical father may have been facilitated for Australian Aboriginals by the predominance of marsupials in the fauna. The process of marsupial gestation remained a mystery long after the beginning of White settlement in 1788. Although the unaided passage of a kangaroo embryo from the vagina towards the pouch was recorded in 1830, more than a hundred years later many Australians firmly believed that marsupial young develop on the teats 'like apples on twigs' (Collie 1830: 240; Troughton 1965: 13-21). Only recently has there been a satisfactory explanation of the phenomenon of embryonic diapause, whereby the interval between copulation and birth may increase up to ten or more times its normal value (Sharman 1955; Sharman and Berger 1969). Thus the evidence available from nature for Aboriginal would-be scientists was confusing. There is no reason why Aboriginals should have based their ethnoscience of human reproduction on the eutherial dingo or bat any more than on the kangaroo or other ubiquitous marsupials. Seligmann (1902: 300-1) mentions that in Papua a community he visited knew little about the reproductive organs of a wallaby he had dissected. Where the local theory of reproduction does call for one or more genitors, another problem arises. Copulation may be thought to be a necessary prerequisite for conception or foetal growth, but it is a compelling fact of nature that it is not a sufficient condition. We do not need to be scientists to discover this. To be complete, a theory must specify sufficient as well as necessary conditions, and in the absence of clues from nature these must be generated by the culture rather than derived from observation. Even in the scientific West not all the causes of infertility are known. To fit the facts the actors' causal model has to contain a substantial error term, and it is scarcely surprising that this is labelled God or spirit, beyond human control. Thus for example Evans-Pritchard (1932: 400, 402, 408) reports that the Azande believe that conception results
GENETRIX : GENITOR :: NATURE : CULTURE?
41
from copulation, and that subsequent acts of intercourse are beneficial in that semen assists foetal growth. But they stress that conception cannot occur unless it is the will of the Supreme Being, Mboli. Likewise, according to the Talmud, there are three partners in every human birth: God, father and mother (Abrahams 1924: 150, 176). In the exchange of views on virgin birth in Man, the contributors seem to have forgotten how recently this tripartite doctrine has ceased to be current in Britain. In the days before our present fertility clinics, the only advice available to barren couples seeking a child was: prayer and perseverance. One last point. Whatever may be their ideas about physical parenthood, virtually all cultures attach symbolic value to both fatherhood and motherhood. I suggest that fatherhood is the freer symbol, able to take on a wider range of culturally assigned meanings, because it has a more exiguous link with the natural world. One striking instance of the use of the symbol of fatherhood is in the charters of organization of polysegmentary societies. There are certainly good social reasons why matrilineal societies never achieve segmental hierarchies with as many levels as are found in patrilineal systems of widest span (Schneider 1961). But it can also be argued that the pedigrees that describe the relations between the major components of polysegmentary societies have nothing whatever to do with domestic kinship, whether patrilineal or matrilineal. I have suggested for the Mae Enga that the idiom of agnation is used to describe simply relations of inclusion; that the statement that A, the apical ancestor of one group, is father of X, Y and Z, the apical ancestors of other groups, means in the higher levels of the hierarchy merely that the groups associated with X, Y and Z form part of the larger group associated with A. In describing the structure of the United States of America to a Mae Enga I might say that Uncle Sam is father of California, who is father of San Francisco; but this statement would imply neither that my mother's brother founded the United States nor that St Francis of Assisi is his grandson (cf. pp. 62-3). In other words, the kin-like relations postulated between high-level taxa in segmentary hierarchies belong to Schneider's category (4) rather than (3). But why in these cases are A, X, Y and Z all taken to be men rather than women, so that A is father and not mother of the others? The organizational arguments about the limited possibilities for polysegmentation in matrilineal systems are irrelevant, for at this level explanations of presentday group dispositions in terms of some historically remote differentiation between brothers and between sisters are equally implausible. We can appeal to Fortes' notion of organic societies, in which 'social organization is governed by the same principles at all levels' (Fortes 1949a: 341; see Gluckman 1963: 73-83), though I would recast this to assert that in these societies social organization is described by the same symbols at all levels.
42
MODELS OF THE REAL SOCIAL WORLD
In this case, the Mae Enga organizational plan is written in agnatic symbols at the top because agnatic principles of organization, even if in modified form, are actually at work at the bottom (Barnes 1967a). Fatherhood is certainly not the only kin term that can be used to indicate relations in set theory; in our own culture we speak sometimes of daughter churches and of sister Oxbridge colleges (consisting originally of celibate male dons), while second and third generations of computers seem to be born asexually. In pre-scientific cultures agnatic idioms appear to be more widely used, and as we move up an agnatic pedigree the symbol of fatherhood is switched imperceptibly from referring to the connexion between individual men and their wives' sons to the connexion between taxa in adjacent levels in a segmentary hierarchy. I suggest that this switching occurs partly because of cultural and social parsimony but also because the symbol is largely a cultural construct, unfettered by evidence from nature. My argument can now be summarized. The relations of nature to fatherhood and motherhood are different. The difference is expressed in the title of this paper: physical motherhood is to physical fatherhood as nature is to culture. Some writers have argued that kinship is based on the cultural and social recognition of physical relations, while others have stressed that kinship, as a social and cultural system, has nothing directly to do with genetic linkages (Beattie 1964; Levy 1965; Schneider 1965). I take an intermediate view that will please neither camp (see Gellner 1963; Barnes 1964b). I argue that the mother-child relation in nature is plain to see and necessary for individual survival. An infant may be free to form attachments to mother-surrogates, but most scientists would agree that a woman's response to an infant after she has given birth is at least in some degree innate or genetically determined. Hence a relation of physical as well as social motherhood is always recognized culturally and institutionalized socially. On the other hand the evidence for the human father-child relation in nature has been, until the last hundred years in the West, slight and inconclusive. There seems to be no evidence that a man is programmed genetically to act differentially towards an infant merely because he has sired it. The processes, necessary for collective survival, of socialization, economic and political mobilization, transmission of offices, power and resources, have facilitated, though they may perhaps not have determined, the institutionalization of social fatherhood in some form or other. Combined with the institution of marriage, this role of social father has provided a basis for the possible development of ideas about physical fatherhood. Thus cultural motherhood is a necessary interpretation in moral terms of a natural relation, whereas the relation of genitor is an optional interpretation, in the idiom of nature, of an essentially moral relation.
GENETRIX : GENITOR :: NATURE : CULTURE?
43
Speaking more generally we may say that there is a real world we call nature which exists independently of whatever social construction of reality we adopt. The relation between nature and culture is contingent; some aspects of nature impinge more obviously and insistently on the human imagination than others. The constraints on the construction we make of fatherhood arise from our social lives as adolescents and adults; our concept of motherhood is more closely constrained by our lives in the womb and as young children while we are still largely creatures of nature.
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
The next two chapters begin the move away from the close connexion with physical facts characteristic of kinship. Both deal with models of one type of social organization, agnatic lineages. My apprenticeship as a field ethnographer was served in Africa, and my anthropological teachers were all Africanists. Then in 1956 I moved to a milieu where the focus of interest for most of my colleagues was the highlands of New Guinea. I was surprised to discover that their yardstick for interpreting their field data on social organization was a highly simplified lineage model allegedly realized in certain societies in Africa, but quite unlike the complicated and diversified African societies I had studied. Part of the reason for this misunderstanding, I thought, was the failure of some Africanists to emphasize sufficiently the difference between muddled configurations of actual alliance, cooperation and conflict observed in the field and the elegant structural models which supposedly underlay the muddle. When I was asked to join a symposium on Highland kinship systems at the 10th Pacific Science Congress in 1961,1 therefore wrote an article aimed at highlighting the differences between observed complex social configurations in New Guinea and simplistic African models. I also suggested, albeit rather timidly, that some Africanists had fallen into the same trap as their colleagues working in Melanesia. The paper was rejected by three journals before being accepted for publication in Man. It has been variously described as equivalent to the refutation of Lysenko's theories (Lawrence 1972: 263) and as probably an enormous red herring (Riviere 1972). I wrote the paper
44
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
45
on board ship, and cited no sources for my statements, but in a footnote envisaged providing later a fuller analysis in which my assertions would be substantiated and qualified. I never wrote this, but discussion of the issues touched on here has been continued by de Lepervanche (1967-8), Scheffler (1985) and many others (cf. Feil 1988). My own contribution to the debate was limited to the discussion of Meggitt's book on the Enga, reproduced here as Chapter 3.
1 Introduction The peoples of the New Guinea highlands first became accessible for study at a time when anthropological discussion was dominated by the analyses of political and kinship systems that had recently been made in Africa. Ethnographers working in New Guinea were able to present interim accounts of the poly-segmentary stateless systems of the Highlands with less effort and greater speed by making use of the advances in understanding already achieved by their colleagues who had studied similar social systems in Africa. Yet it has become clear that Highland societies fit awkwardly into African moulds. When first tackling the New Guinea societies it was a decided advantage to be able to refer to the analytical work available on Nuer, Tallensi, Tiv and other peoples, but it may be disadvantageous if this African orientation now prevents us from seeing the distinctively non-African characteristics of the Highlands. The central highland valleys of New Guinea have become accessible to travellers only during the last fifteen years and early ethnographical research was necessarily undertaken on the coast and in the coastal mountains. These inquiries were made before the work of Evans-Pritchard and Fortes on the Nuer and Tallensi had made its full impact on social anthropology and were carried out among peoples living mainly in politically independent villages whose social organization appeared not to offer any striking parallels with Africa. After 1945 the New Guinea highlands were opened to a new generation of ethnographers strongly influenced by structural thinking, who found here larger societies, apparently patrilineal and lacking hereditary leadership, whose structures invited comparison with Africa. When in several respects these societies were discovered not to operate as an Africanist might have expected, these deviations from the African model were often regarded as anomalies requiring special explanation. Yet in the last year or so a closer examination of the ethnographical facts, the presentation of data from a wider range of Highland societies and, more recently, the discussions about non-
46
MODELS OF THE REAL SOCIAL WORLD
unilineal systems in Malayo-Polynesia have considerably weakened what we might call the African mirage in New Guinea. The Tiv, Nuer, Tallensi and others differ considerably from one another but in making inter-continental comparisons the substantial differences between them have often been overlooked. The possible existence of lineage systems in New Guinea has even been discussed without stating precisely which African lineage systems have been used as type specimens. Comparisons have often been drawn with the more abstract accounts of African societies, as for example Evans-Pritchard's essay in African political systems, rather than with the detailed descriptions of actual African situations given, for instance, in his paper Marriage and the family among the Nuer. It has been easy to make the mistake of comparing the de facto situation in a Highland community, as shown by an ethnographical census, with a nonexistent and idealized set of conditions among the Nuer, wrongly inferred from Evans-Pritchard's discussion of the principles of Nuer social structure. The New Guinea hamlet is found to be full of matrilateral kin, affines, refugees and casual visitors, quite unlike the hypothetic entirely virilocal and agnatic Nuer village (though similar to real Nuer villages). This procedure gives an exaggerated picture of the differences between the Highlands and Africa, and although most ethnographers have avoided this error in print, it persists in many oral discussions. Yet, despite this caveat, major differences in social structure remain between, say, Nuer, Tiv, Tallensi, Dinka and Bedouin on the one hand and, on the other, Chimbu, Enga, Fore, Huli, Kuma, Kyaka, Mbowamb, Mendi and Siane. This is not the place to compare all these systems but rather to suggest topics that should form part of any detailed comprehensive comparison. 2 Descent In the Highlands usually a majority, though rarely all, of the adult males in any local community are agnatically related to one another. Most married men live patrivirilocally. Many a large social group is divided into segments, each associated with a son of its founder. It is argued that these groups are patrilineal descent groups. Yet several other characteristics of Highland societies make this categorization less certain. These may be summarized as follows: (a) In many instances non-agnates are numerous in the local community and some of them are powerful. (b) It is often hard to detect any difference in status between agnates and non-agnates. If a distinction is drawn it may be made in such a way that the patrilineal descendants of non-agnates after one or two generations are assimilated to the local agnatic group.
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
(c)
(d)
(c)
if)
(g)
(h)
47
An adolescent boy, and even an adult man, has some choice in deciding whether he will adhere to the local group in which his father is an agnate or to some other group to which he can trace nonagnatic connexion. He may be able to maintain multiple allegiance or to shift his affiliation. A married woman neither remains fully affiliated to her natal group nor is completely transferred to her husband's group but rather sustains an interest in both. Yet the division of rights in and responsibilities towards her is not exclusive. Many individuals who assert a mutual agnatic relationship are unable to trace out their connexions step by step and are uninterested in trying to do so. The names of remoter patrilineal ancestors are forgotten; or alternatively the genealogical structure of the group is stated to be a single (or sometimes a double) descending line of males with no remembered siblings leading to a large band of brothers about three generations above living adults; or else there is a gap of unspecified magnitude between the putative remote ancestors who give their names to contemporary segments and the father's fathers or father's father's fathers of the living. Even if the agnates form a recognizable core to the local community there may be no context in which all potential members of this core, including non-residents, act as a unity distinguished from their nonagnatic neighbours. An agnatic ancestor cult either does not exist or else does not provide contexts in which non-resident agnates, or agnates from co-ordinate segments, are brought together.
Hence it seems prudent to think twice before cataloguing the New Guinea highlands as characterized by patrilineal descent. Clearly, genealogical connexion of some sort is one criterion for membership of many social groups. But it may not be the only criterion; birth, or residence, or a parent's former residence, or utilization of garden land, or participation in exchange and feasting activities, or in house-building or raiding, may be other relevant criteria for group membership. If, as Fortes advocates, we continue to restrict the category 'descent group' to groups in which descent is the only criterion for membership, then in many Highland societies it is hard to discover descent groups. Furthermore the genealogical connexion required for membership may not necessarily be agnatic. Other connexions can be invoked, and this appeal to other cognatic, and sometimes to affinal, ties does not have to be justified by some elaboration of, or dispensation from, an agnatic dogma. In the Highlands the patrilineal ancestors do not act as guardians of the agnatic principle.
48
MODELS OF THE REAL SOCIAL WORLD
These remarks apply unequally to different Highland societies. In some, long lines of agnatic ancestors are remembered while in others genealogical knowledge is poor and not agnatically biased; in some the local incidence of agnates is high, in others it is low; in some there is strong pressure on a man to affiliate himself exclusively with his agnates while in others he can divide his allegiance between two or more kin groups; and there are other dimensions of variation. The Mae Enga, for instance, fit well into an agnatic model whereas the Chimbu and some other peoples can be treated as agnatic societies only with increasing difficulty as we come to know more about them. Thus although some Highland societies are appropriately classified as agnatic, the area as a whole appears to be characterized by cumulative patrifiliation rather than by agnatic descent. Here I am making a distinction between filiation as a mechanism of recruitment to social groups and to ascribed relationships and descent as a sanctioned and morally evaluated principle of belief. The Tallensi, for example, have both these characteristics. But in most, though not in all, Highland societies the dogma of descent is absent or is held only weakly; the principle of recruitment to a man's father's group operates, but only concurrently with other principles, and is sanctioned not by an appeal to the notion of descent as such but by reference to the obligations of kinsfolk, differentiated according to relationship and encompassed within a span of only two or three generations. In each generation a substantial majority of men affiliate themselves with their father's group and in this way it acquires some agnatic continuity over the generations. It may be similar in demographic appearance and de facto kinship ties to a patrilineal group in which accessory segments are continually being assimilated to the authentic core, but its structure and ideology are quite different. A genealogy in a pre-literate society is in general a charter, in Malinowski's sense, for a given configuration of contemporary social relations. Where there is a dogma of descent, and in particular a dogma of agnatic solidarity, the genealogy must reflect the contemporary situation, or some desired modification of it, in terms of the dogma. But if the dogma is absent, appeal to a genealogy to validate present action is of no avail. Hence it is not surprising that several Highland societies, though again not all of them, neglect their genealogies, either by not revising them or by simply forgetting them. Where revision does take place, it may be simplification rather than the manipulation characteristic of Tiv and Nuer. 3 Bounded and unbounded affiliation In a poly-segmentary society like Tallensi the main affiliations that govern an individual's status and activities are determined by birth. He has a
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
49
specified and unique position in the lineage system and cannot escape from it, though within the minimal lineage he can exercise some initiative, as well as in the affinal ties which he chooses to establish, and in the relationships which he enters into outside the lineage system. In Firth's terminology there is little or no optation in the descent system itself. New Guinea societies, on the other hand, seem to be characterized by a considerable degree of optation. The absence or weakness of a dogma of agnatic descent is one aspect of this and the possibility of affiliation with some local group other than one's father's follows from it. In some societies, Mae Enga for example, sooner or later a man must declare his allegiance one way or the other but in other societies he can, and indeed, if he is ambitious, he will, keep open until late in life the possibility of shifting from one group to another. In the southern Highlands, and possibly elsewhere, a man can successfully continue as a member of two or more groups at the same time. In a unilineal descent system multiple memberships or affiliation of this kind is obviously impossible; one of the arguments used against the alleged feasibility of non-unilineal descent systems is precisely this potential or actual plurality of membership. There are three separate issues involved: the distinction between membership of a group and residence on its territory; the feasibility of multiple affiliation in a system of competing groups; and the notion that a man must have a single home with which he is principally identified. Co-residence implies the possibility, but not the necessity, of continual day-to-day face-to-face interaction and in a nonliterate society, however clearly their rights are recognized, absent members cannot play as full a part in the activities of the group as do those who are present. But just as co-residence does not necessarily imply coactivity, so some form of co-activity is possible without continuous coresidence. This is particularly relevant to those Highland societies where there is no nuclear family residence and where a man sleeps with his fellows while his wives sleep with their young children and pigs in their own houses. Under these conditions, where a man spends the night is only one indication among many of where his principal allegiances and interests lie. His gardens may be scattered, not only in the sense of being located on various ridges and in various valleys but also by being on land under the control of several local groups. In effect, even in those societies where a man's main allegiance is always to one and only one local group, he may have substantial interests in a number of others. There is no great difference between unilocal residence in these circumstances and the manifest poly-local residence reported from some of the southern Highland societies. Multiple affiliation may give individuals greater security and room to manoeuvre but may be detrimental to group solidarity. A group can either
5O
MODELS OF THE REAL SOCIAL WORLD
be jealous of its resources and discourage immigrants or it can seek to build up its strength by recruiting new members. The choice it makes will depend at least in part on the availability of garden land and other natural assets under its control and on its strength as a fighting unit vis-d-vis its likely or actual enemies. Either it can restrict membership by insisting on agnatic purity or in some other way or it can build up its numbers by recruiting non-agnates and by bringing back agnates who have strayed. Highland societies vary in the choice made; probably enough has been published to make a preliminary comparative survey worth while. No simple answer is likely, for it should be remembered that restrictive policies act both ways. A man whose agnatic group is short of land may support a policy restricting use of the land to agnates, but if he is short of land himself he may be relying on exercising his claims as a non-agnate in the territories of neighbouring groups. In the Highlands an individual often has allegiances, of the same kind if varying in degree, to several groups which may be either at enmity or amity with one another. This multiple allegiance is quite distinct from the allegiances of different kinds to different groups which occur in even the most determinate unilineal societies. This multiplicity in New Guinea is largely a result of individual initiative and is not due to the automatic operation of rules. A 'rubbish man' is typically a man who is a member of one local group but who has no ties that lead him outside it, whereas a 'big man' is likely to have a great variety of individual and group ties, along with a clear primary identification with one specific group. Moreover it is proliferation of ties at the individual rather than at the group level that seems to distinguish New Guinea from Africa. As we would expect, both kinds of bond occur in both areas. In most parts of the Highlands there are fairly stable alliances between large groups such as clans and phratries, and sometimes enduring relationships of hostility as well, and these are often expressed in an affinal or fraternal idiom. It is also true that in all the poly-segmentary African societies that we are considering, explicit recognition is given to the rights and obligations which a man has with respect to the groups to which he or his agnates are linked matrilaterally. Yet the relative importance of what we might call highlevel and low-level non-agnatic (and also pseudo-agnatic) ties seems to differ in the two areas. Complementaryfiliationplays a greater part in the lives of New Guinea highlanders and traditional inter-group ties seem less important. It may be argued that this is due to the imposition of colonial peace, for when warfare was endemic inter-group affiliation was presumably more significant than it is now. But the accounts of pre-contact fighting, of the military alliances arranged and the refuges sought after defeat, do not bear this out. In any case pre-colonial fighting is at least as close to the present in New Guinea as it is in the relevant regions of Africa.
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
51
The emphasis on low-level rather than high-level affiliation is clearly associated with the greater range of choice in the New Guinea systems, and in particular with the widespread cultural emphasis on ceremonial exchange. Although exchanges and prestations may be spoken of as arranged by the clan or sub-clan and may even be timed on a regional basis, the great majority of these ceremonial transactions are undisguisedly transactions between individuals. In establishing a position of dominance in these transactions a man is seriously handicapped if he lacks the support of his agnates, but he cannot hope to succeed without utilizing in addition a wide range of other connexions, some matrilateral, others affinal and yet others lacking a genealogical basis. If he is successful it is his local group, usually but not invariably consisting of his close agnates, who more than others enjoy his reflected glory. Among Tiv and Tallensi, and less certainly among Nuer, it seems that a man acquires dominance primarily because he belongs to the dominant local group, whereas in the New Guinea highlands it might be said that a local group becomes dominant because of the big men who belong to it. The contrast is greatest between the Highlands and those African societies where leadership within lineage segments is determined more by rules of seniority than by individual effort. Two aspects of this contrast require special mention. Fortes, in his discussion of what he calls the 'field principle', draws attention to the fact that Tallensi matrimonial alliances are established not at random but in accordance with social interests. The pattern of marriages is determined partly by the choices made by individuals within the range of potential spouses permitted by the rules, and partly by the configuration of rules themselves. Prohibition of marriage within one's own clan, or mother's subclan, or preference for marriage with a specified kind of cousin, indicate the variety of interest involved. Two alternative trends can be seen. Either marriages are restricted to a certain group, so that enduring connubial alliances, either symmetrical or one-way, are maintained and renewed down the generations, or else every marriage between two groups is an impediment to further marriages between them. In other words, matrimonial alliances are either concentrated or deliberately dispersed. The latter alternative is more common in the Highlands and accords well with the emphasis on a multiplicity of freshly established interpersonal connexions rather than on group and intergroup solidarity. The other aspect that should be mentioned is the availability of natural resources. Some of the differences between New Guinea and Africa may be due simply to the differences between pigs and cattle, but obviously this is only part of the story. In the African societies which we are considering a man is largely dependent on his agnatic kin for economic support, but this is less true of the New Guinea highlands. Inheritance and the provision
52
MODELS OF THE REAL SOCIAL WORLD
and distribution of bridewealth play a major part in African societies in determining the structure of small lineage segments and in establishing their corporate qualities. In New Guinea a man depends less on what he can hope to inherit from his father and pays less attention to the ill defined reversionary rights which he may perhaps have in the property of his agnatic cousins. In both areas a man looks first to his agnatic group for garden land, but it seems that in New Guinea he can turn with greater confidence to other groups as well. Before the coming of commercial crops there were in the Highlands, apart from groves of nut pandanus, comparatively few long-lived tree crops or sites of particularly high fertility such as in Africa often form a substantial part of the collective capital of a lineage segment. In New Guinea a man's capital resources consist largely in the obligations which he has imposed on his exchange partners and on his death these resources may be dissipated or disappear entirely. Hence to a greater extent than in Africa every man in the New Guinea highlands starts from scratch and has to build up his own social position. Once again, we must not carry the contrast too far. Clearly even in New Guinea it is generally an advantage to be the son of a big man, just as in Africa the eldest son of an eldest son does not attain leadership without some personal ability; but the contrast remains. In general terms this contrast might be phrased as between bounded affiliation in Africa and unbounded affiliation in the Highlands; or between African group solidarity and New Guinea network cohesion. 4 Social division as condition or process Concentration on the network of alliances between individuals and between small groups may perhaps explain why comparatively little attention has been paid in New Guinea studies to the processes whereby groups such as clan and sub-clan segment and divide. In the analysis of segmentary societies there are always two points of view. On the one hand poly-segmentation is seen as an enduring condition whereby there are in existence, and perhaps have been for a long time, a fixed hierarchy of segments, each segment of higher order containing several segments of lower order. Evans-Pritchard and Fortes' earlier work discusses how in different contexts segments variously oppose and support one another without changing their status in the segmentary hierarchy. The terms 'fission' and 'fusion' were applied to these shifts of opposition and alliance in different contexts. On the other hand we may turn our attention to the ways in which new segments are formed and how existing segments are upgraded, downgraded and eliminated. Many recent writers have followed Forde in using 'fission', 'fusion' and other terms to refer to these processes of status alteration of segments rather than to the
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
53
contextual shifts with which Evans-Pritchard and Fortes were initially concerned. In New Guinea the contemporary pattern of poly-segmentation has been documented for many societies. There has been some discussion of how the fortunes of war have led in the past to changes in this pattern, and a little has been said about contextual shifts of opposition and alliance. There has been less analysis of how, for example, increasing population over the years may result in a segment of one order converting itself, gradually or suddenly, to one of higher order. This omission arises partly because it is hard to get any reliable time depth from thefieldmaterial; Highlanders are poor oral historians. But it is due also, I suggest, to a basic difference between New Guinea and Africa in the way in which over-large groups split up. In Nuer, Tiv and Tallensi we have a clear picture of how, given adequate fertility, two brothers from their childhood gradually grow apart until, after several generations, their agnatic descendants come to form two distinct co-ordinate segments within a major segment. Even if some analytical queries remain the process over at least the first three generations is well understood. This kind of segmentation we may well call chronic, for in a sense the division of the lineage into two branches is already present when the brothers are still lying in the cradle. The details of the process may be unpredictable but the line of cleavage is already determined. Segmentation orfissionin New Guinea appears not to take this inexorable form; one cannot predict two generations in advance how a group will split. Instead it seems that within the group of agnates and others there is a multiplicity of cleavages or potential cleavages. In a crisis these are polarized, two men emerge as obvious rivals and each with his followers forms either a new unit or a distinct segment of the existing unit. Segmentation, as it were, is not chronic but catastrophic. The regularities, if any, in catastrophic segmentation are obviously harder to determine than in chronic segmentation. In Africa the dogma of descent acts as a continuously operating principle, providing each individual with an ordered set of affiliations, so that in any crisis he knows his rightful place, even if he is not always there. In New Guinea affiliations are not automatically arranged in order in this way; what might be called the principle of social mitosis, whereby potential recruits to rival co-ordinate segments sort themselves before an impending crisis, is absent; the break, when it comes, appears to come arbitrarily. In addition, changes in the poly-segmentary pattern in New Guinea seem to come about more often than in Africa as the result of defeat in war. The cases of war may be predictable but who is killed and who lives, which group wins and which loses, is in New Guinea as much as anywhere else a matter of luck. Here again we have to deal with an apparently arbitrary process.
54
MODELS OF THE REAL SOCIAL WORLD
This lack of predictability or regularity in changes in the segmentary pattern is, of course, another aspect of the basic contrast between group solidarity and individual enterprise. The sanctions that maintain the segmentary status quo, whether derived from economic or physical pressures, or from cult or dogma, are weaker in the Highlands than in Africa and the incentives for change are stronger. A characteristic of Highland cultures, and perhaps of Melanesia as a whole, is the high value placed on violence. The primitive states of Africa, and even the African stateless societies which we have been considering, are readily likened to the kingdoms and princedoms of mediaeval Europe, valuing peace but ready to go to war to defend their interests or to achieve likely economic rewards. Prowess in battle is highly rewarded but warfare is usually not undertaken lightly and most of the people most of the time want peace. In New Guinea a greater emphasis appears to be placed on killing for its own sake rather than as a continuation of group policy aimed at material ends. In these circumstances we might expect to find a less developed system of alliances and countervailing forces, and less developed arrangements for maintaining peace, than we would have in a polity directed to peace and prosperity. Secondly, we would expect that leaders, whatever their other qualities, were moved to violence at least as much as their fellows and possibly more. The highlands of New Guinea cannot have been the scene of a war of all against all, for the pre-contact population was large and often densely settled; indigenous social institutions preventing excess violence and destruction must necessarily have been effective, for otherwise the population would not have survived. Likewise other qualities than prowess in violence were required for leadership, in particular the ability to engage and co-ordinate the efforts of others in ceremonial exchanges. Yet despite these qualifications I think that it may still be hypothesized that the disorder and irregularity of social life in the Highlands, as compared with, say, Tiv, is due in part to the high value placed on killing. 5 Conclusion I have sketched some of the difficulties that follow from assuming that the societies of the New Guinea highlands can be regarded as variants on a pattern established by the Nuer, Tallensi, Tiv and similar African societies, and I have tried to indicate ways out of these difficulties. There are major ecological differences between the two groups of societies and any full commentary would have to take account of these, in particular the lack of storable food in New Guinea. Despite the great difference in structure, culture and environment, one route to a better understanding of the Highlands lies, I think, through a closer examination of the detailed
AFRICAN MODELS IN THE NEW GUINEA HIGHLANDS
55
information available on the stateless societies of Africa. Perhaps this examination may lead incidentally to a clearer formulation of the salient characteristics of these African systems. It so happens that stateless societies were studied and described in Africa before ethonographical research really got under way in the Highlands. It would be interesting to work out how, say, the Nuer might have been described if the only analytical models available had been those developed to describe, say, Chimbu and Mbowamb. At the same time, if the differences between the patrilineal poly-segmentary stateless societies of Africa and the societies of the New Guinea highlands are as great as I have suggested, it might be worth while looking for other societies in Africa that could provide closer parallels.
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
This chapter was written as the introduction to a French edition of Meggitt's Lineage system
of the Mae~Enga
(1965a), an edition that never eventuated. I had become concerned with the lack of attention paid to demographic factors in the classic models of African lineage systems proposed by Evans-Pritchard and Fortes and their pupils (e.g. Fortes 1953). The continually segmenting lineages they postulated seem to me to entail a population that doubled or trebled in every generation, yet most of the other aspects of their models of African societies appeared to be based on the assumption of a steady state. I thought that a model of lineage structure was required that was more plausibly diachronic, that allowed for the disappearance of segments as well as their formation, that took account of the availability of land, and that was linked to whatever historical evidence was available. Peters' analysis of the genealogies of Cyrenaican Bedouin and Lebanese Muslims (Peters i960, 1963) provided empirical evidence for the kind of model I had in mind. I made a preliminary sally in that direction in a review of Meggitt's book (Barnes 1967a), for both the ethnographic data presented by Meggitt and his analysis of them seemed to fit well with Peters' work. I saw the invitation to write an introduction to the book as an opportunity to develop a more fully specified model, taking account of a very perceptive paper by Kelly (1968) which placed the Mae Enga in the broader context of other Highland societies.
When a trained social anthropologist publishes an account of the customs, beliefs and behaviour of a community he usually addresses two audiences, whose requirements may conflict. On the one hand he tries to make a 56
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
$J
contribution to our knowledge of what happens, what is said, even what is thought and dreamt, at a particular time and in a particular place. He supplies ethnographic data, a contemporary record that can be added to the historical archives of mankind. For this purpose almost all details are relevant, particularly when he is describing a people who, like the Enga, lack writing and other means of making their own enduring records, and whose culture is complex, lively, exotic, previously unknown to outsiders, irreplaceable but rapidly changing. In these circumstances we want to learn and record as much as we can about the contemporary facts of traditional human cultural diversity before we lose the chance for ever. For this reason, and with this audience in mind, some ethnographic monographs are lengthy, comprehensive, discursive, profusely illustrated and packed with details of the ephemera of social life. These books are necessary and valuable; but the monograph I consider here is not of this kind. The second audience calls for a different kind of book. Instead of writing primarily for the ethnographic record, describing what men do, the ethnographer tries to make a contribution to our understanding of how men are constrained to think and act in one way rather than another, how one custom or social institution influences another, and how the structure of society is maintained through time. For this purpose the anthropologist takes the ethnographic record for granted, and extracts from it only those items which are relevant to his argument. His aim is to say something about human society in general, rather than about some particular society. Indeed, he is likely to pursue his argument by comparing the features of several societies that belong to the same type. Even if, as is often the case, his theoretical interests are delimited in space and time, he will try to replace the de facto particularistic limits of, say, being in the Great Plains of North America between the sixteenth and eighteenth centuries, or in the highlands of New Guinea in mid-twentieth century, by the universalistic limits of living in a prairie grassland environment or in a tropical cordillera, having or not having sweet potato or other staple crops, using or not using stone or iron tools, and so on. Pushed to the limit, his analysis becomes independent of where in the world his societies are located or whether they died out long ago. Meggitt's book (1965a) is an outstanding, and fairly uncompromising, example of the second kind of writing. In several publications (1957, 1962b, 1962c, 1964, 1965b, 1967a, 1967b, 1969), some published subsequent to the English edition of this book, he has described various aspects of the culture and social structure of the Mae Enga people of the Western Highlands District of the Trusteeship Territory of New Guinea, among whom he began working in 1955. Taken together, these publications constitute a quite full account of how the Enga live and how they
58
MODELS OF THE REAL SOCIAL WORLD
differ from their neighbours. But in this book, based on Meggitt's doctoral thesis presented in the University of Sydney and his major contribution to the corpus of Enga ethnography, we are presented above all with a model and an argument. He uses his field observations not to describe the Enga but to provide evidence for the validity of a model of a segmentary society exploiting a limited amount of land. Like almost all segmentary models, it consists of several similar hierarchies, each comprising a series of nesting segments ranging in size from great phratry down to family. He compares the Enga with various other Highland societies which have a segmentary structure and subsistence economy broadly similar to the Enga. He classifies these societies along two dimensions, the amount of 'coherence and elaboration of the agnatic system' and the amount of flexibility in the pattern of residence. He argues that differences in these two dimensions are to be explained by differences in the availability of arable land. The model has four important characteristics which marked it off from most of the earlier attempts to depict the salient features of the societies of the New Guinea highlands, but which have been echoed in many of the analyses of other societies in the region which have appeared after the publication of Meggitt's book in 1965. The model is stochastic rather than determinist, specifying mean values rather than the values of individual components. Although Meggitt makes extensive explanatory use of the values and institutional inventory of the Enga, which are taken as invariant throughout the system, the model is not made up of cultural rules. Instead it is a model of social action, of behaviour, of what actually happens, of the decisions taken by individuals and groups, in the light of Enga culture, to marry, fight, break away, attend ceremonies, claim compensation, borrow land and so on. Thirdly, the model is calibrated, though only partially. Instead of limiting himself to purely qualitative statements about the shape of the lineage system and how its parts change, Meggit endeavours wherever possible to measure frequency and size. He provides tables showing the mean size of segments of different orders, the rate at which they shift their position in the hierarchy, the frequency with which ghosts attack kinsfolk of different kinds, the number of pigs transferred in various contexts, the number of relatives who attend funerals, and so on. This quantification has considerable importance for the development of Meggitt's own analysis, but also it opens up two possibilities of great significance. If similar quantified information becomes available for other Highlands societies, then we shall be able to make much more reliable and precise comparisons between one society and another. Instead of being limited to saying, for instance, that, as compared with the Enga, the Kuma or the Chimbu have more flexible residence patterns, we shall be able to quantify the differences between them. Meggitt's book has already stimulated many subsequent writers on the Highlands to provide
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
59
the information needed for this kind of precise comparison. The number of variable features which differentiate one Highlands society from another is so great, despite their substantial overall structural similarity, that a necessary first step toward discovering to what extent there is order in this diversity is to reduce the differences to quantitative terms. The other possibility lies not in cross-society comparison but in plotting the trend of change within Enga society itself. It is clear from Meggitt's book, even though it is peripheral to his main theme, that the intrusion of Australians and other outsiders into the New Guinea highlands from 1933 onward has already led to major irreversible changes of quality in the Enga way of life. The suppression of warfare, the propagation of Christianity, the introduction of wage labour, cash crops, schools, and the development of a literate territory-wide administrative and political system have all had a substantial qualitative impact on traditional institutions, customs and beliefs. Yet for some time to come there will also be changes in quantity: the number of pigs paid in bridewealth may rise, ghosts may shift their targets, more men may borrow land from matrilateral kin, and so on. In order to understand these processes of change, it is necessary to chart them quantitatively, to see in what order changes occur, and to discover where old patterns persist longest. Only then can we begin to look at the changes, not as social disintegration, but as a series of would-be constructive adaptations to the massive pressure of external circumstances. The fourth characteristic of the model arises from Meggitt's definition of his task in a situation of rapid social change. Most of his data relate to events which occurred while he was in the field or during the lifetime of Enga men and women whom he interviewed. Because the arrival of alien administrators, missionaries and others was so recent, and the effect they had had so far on Enga society so limited, it was possible for Meggitt to treat most of his observations as if they occurred within a traditional indigenous social order. He constructs a model of this traditional society, and at those points where administrative or mission influences cannot be ignored, as in the suppression of warfare, and the appointment of minor officials, luluais and tultuls, he deals with the observed facts in, as it were, footnotes to the model, which remain entirely extraneous to it. The Australian impact on the Highlands is treated, quite properly in this context, as an enormous discontinuity whose rationale lies outside Meggitt's terms of reference. Another limitation which he imposes lies at the other end of the time scale. The Enga cannot have been living in these highland valleys for ever; like every other human group, their prehistory is limited. Recent research in the Highlands suggests that there have been agricultural peoples living there for several thousand years, long before the introduction of the sweet potato. But Meggitt deliberately ignores these facts of prehistory in this book. He is not concerned with the development
6O
MODELS OF THE REAL SOCIAL WORLD
of Enga society, either through hundreds or through thousands of years; for purposes of analysis he treats Enga society as if it had been there for ever in essentially the same form. In the absence of reliable historical traditions, and of an established archaeological record, he has no option. Hence his model is one in which hardly any cumulative or irreversible changes occur. I think the only exception lies in his reference to the colonization of empty valleys outside the crowded central Enga area 'a few generations ago', but even then the outer valleys are treated as if they constituted an indefinitely large sink which never fills up and which is not fully utilized in solving the problems of land pressure in the central valleys. Yet despite this very limited inclusion of cumulative diachronic processes, Meggitt's model of Enga society is not synchronic or static. On the contrary, as we have seen, he is concerned with processes of change in the segmentary hierarchies and endeavours to measure the rate at which changes occur. Thus we have a non-cumulative diachronic model, rather than a synchronic model, as might have been offered by an ethnographer writing thirty years ago, or a cumulative diachronic model, which would require much more evidence about social evolution than could possibly become available for the Enga. Again, Meggitt has established a style in models which has been followed by many subsequent writers on the Highlands. In a subsequent paper (1967b) he has elaborated his model to include a two-phase alternation of state in each segment up to the level of clan. In periods of diastole, when peace prevails, leaders jockey for position and compete for followers. This 'Melanesian' phase is followed by an 'African' period of systole when agnatic distinctions are stressed, leadership is stable, and conflict becomes warlike. This in turn gives way to diastole and so on for ever. The model remains diachronic and non-cumulative. A segmentary model of a society, whether synchronic or diachronic, cumulative or non-cumulative, repetitive or non-repetitive, is not necessarily tied to any particular ideology or rule of recruitment. All that is needed is a rule for assigning newly born or newly arrived members to terminal segments, and a language for describing how the segments are related to one another and how they shift. There is no a priori connection between poly-segmentation and kinship. In the organizational charts of industrial enterprises and administrative systems in modern states we find innumerable examples of segmentary models, usually synchronic, with rules of recruitment based on other criteria such as locality, specific need or technical skill, and with relations between segments described in terms of the allocation of power and responsibility. Yet in the tribal world almost all segmentary models of society are described in the language of kinship, usually of agnation, and recruitment of the newly born is usually by a rule of descent or filiation. We refer to them as lineage systems. We might be
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
6l
tempted to think that the language and processes of kinship are obvious choices for the rationale of segmentary organizational charts in preindustrial societies, and that therefore their use does not raise any structural difficulties. But this is not so. The choice may be obvious, but the difficulties are inescapable. If tribal segmentary models were cumulative, capable of indefinite growth to accommodate an indefinitely large population and explicated in indefinitely long pedigrees, there would, I think, be no great difficulty in linking the model to a descent rule of recruitment and a kinship structural language. Perhaps there might be problems in getting the system started, as in any creation myth. But in fact all ethnographers, and in many cases the members of the societies themselves, recognize that the population of a single tribal society cannot be thought of as expanding indefinitely, and that people without writing cannot remember indefinitely long pedigrees. Furthermore, most ethnographers, whatever the people themselves may do, prefer to ignore or to treat as a separate question how the system became established. They aim to present the system as if it has been going for an indefinitely long period and is capable of continuing indefinitely. Hence we get a non-cumulative model of the society. Unfortunately, fitting a descent or filiative rule of recruitment and a kinship idiom for structure to a non-cumulative segmentary model presents certain difficulties. These difficulties were certainly implicit even in the pioneering studies of the Nuer and Tallensi in Africa by EvansPritchard and Fortes but, for a variety of reasons, they were not seen at that time as clearly as they might have been. The operation of processes of absorption, accretion, telescoping, and the like was, of course, well recognized and documented in these and other African studies, but their full structural significance seems to have been overlooked when methods of analysis developed in Africa were first applied to the societies of the New Guinea highlands (cf. Chapter 2; Barnes 1967a; Langness 1964). In Meggitt's volume he looks quantitatively at these processes and repeats his set of fifteen propositions about the frequency of their occurrence in specified contexts (cf. Meggitt 1962b: 162-4). Nevertheless, I think there is something more to be said about the connexion between the segmentary model and the language and processes of kinship, about what is entailed when we say that a certain segmentary model is a lineage system. The principal difficulty can be stated quite simply. If the model is noncumulative, then the number of levels of segmentation must remain approximately constant, for this is one of the salient parameters of the model. Yet if the segmentary system is also a lineage system, maybe an agnatic lineage system as with the Enga, then relations between the apical ancestors of superior and inferior segments are said to be those of father and son. Each generation adds an extra step between the living and the eponymous ancestors of the segments to which they belong. The steady
62
MODELS OF THE REAL SOCIAL WORLD
passage of time and the replacement of one generation by another calls for a cumulative model of a segmentary system which will grow indefinitely at the rate of one level every generation. Nothing like this cumulative model has been found anywhere in the tribal world, yet the use of the language and ideology of agnation to describe and explain segmentary systems is widespread. Clearly there must be other processes at work which enable a cumulative ideology to be applied without contradiction to a non-cumulative model. The best account of what seems to happen has been given by Peters (i960, 1963), based on observations made among Cyrenaican Bedouin and Lebanese Muslims. The configuration of segments at the uppermost levels is fairly constant, and represents what Peters calls a description of the 'cultural past'. Among the Enga, presumably the sun and moon, sky people and phratry founders belong to this stable elite. The lower levels of the segmentary system record, ostensibly correctly, who is the father of whom. Among the Enga, genealogical demographic accuracy does not extend very far. The son's son's son of a non-agnatic immigrant who lived uxorilocally is classified as an agnate, and the father's father of a living man with many close agnates may be raised a generation in the pedigree. Therefore Enga segmentary pedigrees cannot be relied upon as accurate demographic records at levels more than two, or even more than one, steps upward from living adults. In between the top and bottom we have in general what Peters calls an 'area of ambiguity', where adjustments are made in the relative position of segments so as to reflect their contemporary size and political status. This is an area where segments amalgamate and redundant apical ancestors get forgotten. Although evidence for these shifts and deletions is hard to obtain when there are no written records and when the ethnographer's observations extend over only a couple of years or so, Meggitt manages, with considerable ingenuity, to provide good evidence that these changes do occur in the Enga system. As he puts it, 'The amendments reflect the Mae rule of thumb that the status of an agnatic group is to some extent defined by what it does'. Since this is so, we cannot take at face value statements about the agnatic connections between the eponymous apical ancestors of any of the higher segments. We may be told that Al, the founder of the subclan of that name, is the son of A, the founder of the clan of which subclan Al is a part. We can accept this statement as true about the relation between the clan and the subclan, but we cannot accept it as necessarily true about the relation between the individuals A and Al. I would go further and suggest that there is no reason for assuming that these individuals ever existed at all. Indeed, Meggitt gives one instance where the names of two co-ordinate subclans are words which translate as 'hill-dwellers' and 'mountaindwellers', though the Enga insist that these are the names of the two sons
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
63
of the clan founder. I argue that the simplest, and perhaps in some languages the only, way of describing the relative positions of segments in a hierarchy, the relations of inclusion, co-ordinate status and superior or inferior level, is in the idiom of kinship. Suppose I say: Uncle Sam had several sons, two of whom were called California and Texas. San Francisco and Los Angeles were sons of California, while Dallas and Houston were sons of Texas. The metaphor may be unfamiliar, but there is no doubt about the relative position of the various taxa. If I had to describe the administrative structure of the United States to an Enga who was unfamiliar with the outside world, this might be the best way to start. In doing so, I would not be asserting that St Francis was the grandson of my uncle Samuel. In other words, there is a distinction between the use of an agnatic idiom to describe structure and an agnatic rule to determine recruitment. The relative position of the larger Enga segments is described in the language of agnation. If they shift their relative position at all, they do so very slowly, so that the validity of the idiom is not challenged. The eponymous founders of segments are dead before the scene opens and remain dead for ever. On the other hand, the smaller segments cannot be treated as if they were unchanging, for any genealogical calculus, whether agnatic or not, has at the lower levels to take some account of the simple facts that new individuals continually enter the system through birth, in somewhat random fashion, and that they steadily get older. The distinctive dynamic features of any segmentary system are therefore likely to be discovered by examining the lower levels, where the passage of the generations cannot be ignored, rather than by looking at the relatively static pattern of highlevel segments, whatever may be the language in which this is described indigenously. Meggitt refers to the Enga segmentary system as an agnatic lineage system. It certainly looks like an agnatic system at the top. But what about the lower levels of the system? There are three ways in which men can be located in a segmentary system: by birth, by transfer later in life as individuals, and by transfer as members of an existing segment. Most men at birth join the segments to which their father belongs; the Enga ideology of agnation is matched by frequent patrifiliation in practice. Later in life, a man may reject the preferred pattern of patrivirilocal residence in favour of living with his wife's group, but he remains a stranger in their midst. However, the sons of a man who has lived uxorilocally are known to his wife's group as 'sons of female agnates'. If they follow the standard virilocal rule, their children are known as 'clan agnates', though Meggitt distinguishes them as 'quasi-agnates'. In the next generation, as noted above, assimilation to agnatic status is complete. Thus the segment
64
MODELS OF THE REAL SOCIAL WORLD
founded by an immigrant husband living uxorilocally is reclassified, after the passage of three generations, as part of the clan, or perhaps subclan, of his wife. There is some doubt about the frequency with which this process of reclassification occurs, and the reader is referred to a paper by McArthur (1967) for details. However it is clear that, even at the middle levels of the system, the agnatic idiom is used in part to express relations of continued residential association over the generations as well as relations of patrifiliation. There are one or two other ways in which quasi-agnatic, and later fully agnatic, status may be acquired, and some of the ostensibly agnatic links between the founders of lower-level segments are recognized as fictitious. In a few instances the founder of a segment is said to have been a sister's son of the founder of the superior segment, and not his own son. Meggitt also suggests that if initially A is the son of B, and A's segment expands while the rest of B's segment contracts, A may be reclassified as the father of B. Despite these deviations from the untarnished paradigm of an agnatic lineage system, Meggitt insists that the Enga system is strongly agnatic, in contrast to a system based merely on patrifiliation. Here he is making use of a distinction between agnation and patrifiliation by Fortes (1970: Chapters 3 and 4), a distinction of considerable subtlety that cannot be elaborated here. Meggitt's contention can, I think, be challenged in terms of the assimilation of agnates, the definition of rights to land, beliefs about the transmission of 'blood' from parent to child, and other cultural features (cf. Barnes 1967a). Whether it be agnatic or not, the Enga segmentary system is certainly comparatively sharply defined, and most couples conform to the dominant pattern of patrivirolocal residence. Meggitt attributes this to the limited availability of arable land and the high demand for it. In his final chapter he compares the central Enga with fringe Enga groups whose territories are not so crowded, and with other societies in the Highlands for whom comparable information on land and social structure is available. He endeavours to reduce his comparison to two variables, population density and a Variable of agnation', i.e., the extent to which social groups are structured in terms of agnatic descent and patrilocality. His classification of Highland societies in three main groups makes good sense, but it is hard to clinch the argument, for two reasons. Reliable quantified information about pressure on land is not available for most of the societies he compares. Even where we do have full information about land use (cf. Glasse and Meggitt 1969: 235-42 for references) it is not obvious how to construct a single index of land pressure. The highlands of New Guinea constitute a fully three-dimensional environment, with a variety of soils and climates at different altitudes, with complex substitution possibilities. Secondly, the other variable, agnation, may be interpreted in more than
AGNATIC TAXONOMIES AND STOCHASTIC VARIATION
65
one way, as I have indicated. Yet despite these limitations, Meggitt's pioneer classification has stimulated others to try their hand, and de Lepervanche (1967-8) in particular has discussed at length the variations in residence patterns, leadership and segmentary formation found in the area. Meggitt's thesis about agnation and land pressure has been examined from a novel viewpoint by Kelly (1968). If we ignore the somewhat controversial issue of agnation, Meggitt's thesis reduces to saying that when land is short, land-holding groups will tend to refuse loans to outsiders. Against this, I have suggested (Chapter 2) that when men are short of land in their own group, they will tend to borrow from groups other than their own. Kelly breaks out of this impasse by looking in various Highland societies at the type of segment that has land to allocate to its members and to lend to outsiders. The smaller the group, the greater the probability that in the next generation there will be a serious imbalance in the number of male and female children, and consequently an imbalance in the availability of land to the men of the group. The amount of imbalance may also be affected by the amount of polygyny, differences in wealth between clans, and the mode of distributing women following on warfare. While recognizing substantial differences between them, Meggitt puts both Enga and Chimbu at one end of his continuum. Kelly notes that, although land is scarce in both these societies, Enga patrilineages which allocate land are, because of their size, particularly susceptible to imbalance due to demographic causes; people are redeployed over the available land primarily through loans within the clan, as well as by warfare between clans in the past. Chimbu warfare takes a different form and does not constitute a means of reallocating land; this is done by transfers to affines and matrilateral kin more often than to fellow clansmen. Clan territories among the Enga usually range from ridge top to valley bottom, providing the full range of possibilities for planting, whereas Chimbu clan territories tend to range less widely in altitude. Thus Enga gardeners can usually find a clansman who controls land they want to borrow, while Chimbu gardeners often have to seek the land they need in the territories of other clans. Among the Siane, where there is little pressure on the use of land, redistribution of land occurs mainly within the clan because the territories of other clans are so far away. Kelly's analysis, like Meggitt's suffers from a lack of reliable quantified data, and the validity of his index of demographic dislocation needs testing. The importance of Kelly's analysis lies in its use of the facts of demography and environment, and in its examination of the structural consequences of differences in the extent of stochastic variability. He points the way toward a firmer understanding of the complexities of social structure in the Highlands. The best evidence for the good qualities of the ethnographic enquiries made by Meggitt
66
MODELS OF THE REAL SOCIAL WORLD
and other social anthropologists who have worked in the region, and for the intellectual stimulus provided by their writings, is that their data and findings prompt others to rework the material and to look for new answers.
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
As an Africanist trained in Cape Town and Oxford, I had been taught that in the field my initial task should be to search for corporate social groups, the agnatic lineage segment being the prime exemplar. I had found this advice of only limited value when working among the Ngoni of what is now Zambia, for their kinship system, though heavily coloured by agnatic ideology, was in practice what I called 'omnidirectional' (Barnes 195if: 125-6) and lacked any corporate groupings above the level of the domestic family. When a few years later I began working in Norway, I thought that the lesser salience of a kinship system in an industrialized society, and the greater salience of an institutionalized legal system would restore importance to my teachers' advice. I was mistaken, for I had overlooked the large place occupied in the social life of western Norway, as in all industrialized societies, by the relations of friendship and the categories of social class, neither of which generate corporate groups. In the paper I proposed the use of the notion of social network to describe and analyse systems of relations which do not necessarily cluster into clearly delimited groups. I also drew attention to a distinctive pattern of leadership and decision-taking found on the island where I worked, reminiscent of the Duke of Plaza-Toro, who 'led his army from behind, he found it less exciting'. The study of social networks is now well established as a specialization within sociology and social anthropology (cf. Wellman 1983) but 'leadership from behind' still awaits more professional scrutiny (Barnes 1987b). My paper attracted the attention of the American
68
MODELS OF THE REAL SOCIAL WORLD
anthropologist Robert Redfield, who then used my findings as an illustration of a community lying, morphologically and developmentally, somewhere between a peasant village and a town (Redfield 1956). I queried some of the implications Redfield had drawn from my data, and later (i978d) published my critique of Redfield's folkurban continuum.
1 Introduction When we study the social organization of a simple society, we aim at comprehending all the various ways in which members of the society systematically interact with one another. For purposes of analysis we treat the political system, the pattern of village life, the system of kinship and affinity, and other similar areas of interaction as parts of the same universe of discourse, as though they were of equal analytical status, and we strive to show how the same external factors, principles of organization, and common values influence these different divisions of social life. This task, though always difficult, has been accomplished for a growing number of technologically simple societies, about which we can feel confident that we have an appreciation of what the society as a whole is like. When we turn to the enormously complex societies of Western civilization our task becomes much more difficult. Fieldwork in a Western community can lead directly to knowledge of only a very small sector of the social life of a largescale society. This limited area of detailed knowledge has then to be related, as best we can, to experience and information derived from other parts of the society. In 1952 I began studying in western Norway, in what was then the civil parish (kommune) of Bremnes. During my fieldwork I did not try to gain first-hand knowledge of Norwegian society as a whole. Any such attempt would, I am sure, have been entirely unsuccessful. Instead I attempted to isolate for study certain aspects of social life in which I was interested, which were relatively unknown, and relatively easy to grapple with. Many writers have discussed the political history of Norway, the development of its economic institutions, the personality of its inhabitants, their forms of religious belief, and similar topics. Some sociological fieldwork had been carried out in the country before I began my inquiries but very little was then known about the operation of the social class system in a land which prided itself on its affirmation of social equality. Therefore I decided to concentrate my attention on those kinds of face-to-face relations through which a class system, if there were one, might operate. I was also interested in the way in which collective action was organized in a society of this kind, and was therefore led to consider the working of committees. I
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
69
chose to study an island parish partly dependent on industry since I believed that there the field data I sought would be easier both to obtain and to understand than in a town. In 1952 Bremnes had a domiciled population of some 4,600. Of every ten men over fifteen years old, three were engaged mainly in fishing, two in agriculture, another two in industry, and one was a merchant seaman. These occupations accounted for 84 per cent of the adult male population. Another 6 per cent were gainfully employed in other occupations, and the final 10 per cent were retired. The majority of adult women were housewives. These percentages indicate only the principal occupation, for many men divided their time between different ways of earning a living. All farms were small, very few employed paid labour, and few households could live off their land alone. Therefore most peasants spent at least the time from Christmas to Easter fishing for herring, while others worked in the local marine-engine factory. A few men were almost full-time administrators, but the bulk of the work of local government was carried out by part-time officials such as the mayor, the parish treasurer, the tax assessors and collectors, and the chairmen of the various standing committees, most of whom had small-holdings as well as their public work to attend to. Bremnes was part of Norway, and its inhabitants shared much of their culture with their fellow countrymen, as well as belonging with them to a single economic, social and administrative system. Here it is sufficient to mention that Norway was, and still is, a democratic monarchy with high taxation and comparatively little extremes of poverty and wealth. Over 95 per cent of Norwegians then belonged to the Lutheran state church. There was no tradition of feudalism, there were no hereditary titles outside the royal family, and virtually all children attended the official elementary schools. An idea strongly stressed in Norwegian thought was that no man should have more privileges than his fellows. 2 Social fields Each person in Bremnes belonged to many social groups. In particular he or she was a member of a household, of a hamlet, of a ward and a member of the parish of Bremnes. At different times and different places membership of one or other of these groups was definitive for what he or she did. An adult went to the prayer-house with his or her household, sat at weddings with other members of his or her hamlet, and paid tax according to his or her parish. There were other series of groups which to some extent cut across these territorially based ones, although they might themselves be based on territory. Thus for example a man might belong to a hamlet missionary working-party, or to a bull-owning cooperative based on a ward. In formal terms these various groups fitted one inside the other, each
7O
MODELS OF THE REAL SOCIAL WORLD
in its own series. Thus there were three territorially defined fishermen's associations in the parish. All three belonged to the provincial fishermen's association, and this in its turn formed part of the national association. There might be conflicts because of the duties and rights a person had in the various groups in any one series, and there might be conflicts because of his interests in different series. This is true of all societies. The territorial arrangement of the Bremnes population was fairly stable. The same fields were cultivated year after year, and new land came into cultivation only slowly. Houses could be moved from site to site, but this was expensive. Land could be bought and sold, but there were several factors tending to discourage frequent sales of land. Thus for the most part the same people went on living in the same houses and cultivating the same land from year to year. This provided, as it were, a stable environment in which social relations were maintained through the decades, and a frame of reference by which individuals could relate themselves to other people. This territorially based arrangement of persons was, however, only a part of the social system of Bremnes, for men utilized the sea as well as the land. Herring were not cultivated, they were hunted. They were taken from the sea, where there were no territorially defined rights in property. They were caught by men organized in groups of from five to twenty, whose composition varied from year to year much more than did the household groups who worked together on the same holding. The fishing vessel, the temporary home of the fishermen, wore out much more quickly than a house or a plot of land, and it could be bought and sold comparatively easily. Even more important, there were no women on board. Wives and children remained behind and stayed in one place while the men moved from one fishing-ground to another, and from one crew to another. Here then we had two distinct kinds of social field, a fluid and a stationary, and we shall presently discuss a third field linking these two. The fluid field was the field of industrial activity, in which men earned money by catching fish; the stable field was the field of domestic, agricultural and administrative activity ashore where they, or their wives, spent the money. The Norwegian fishing industry was efficient; technological change was going on continually, and vessels made use of modern equipment like radio telephones, echo-sounding gear, nylon ropes, radar and asdics. It was a highly competitive industry, each vessel striving against all other vessels. Loyalties to kinsmen, neighbours and friends continued to operate, but only to a limited extent. In their own words, 'herring fishing is war'. Any man could try to get himself included in a crew and each owner sought to engage the crew that would catch most fish. During the herring season, men from Bremnes sailed in vessels belonging to other parishes, and vessels registered in Bremnes sometimes had on board fishermen from as
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
71
much as six hundred miles away. In effect, there was something like a free labour market. Men applied for a place in a particular vessel because of contacts they had made, friends and relatives who had already served on board, or the success of the vessel in previous seasons. An industry of this kind could scarcely have operated if the pattern of social relations had been as fixed and stable as it was in the round of social and economic activities ashore. The greater portion of the herring catch was exported and in order to sell on the world market at a profit the size of the fishing fleet and the amount of capital invested in it had to respond to economic pressures which varied in intensity and point of application. There was a huge marketing organization and various reserve funds which evened out part of the differences in earnings between one vessel and another and from year to year. Even so, the amounts earned by the fleet as a whole varied considerably over the seasons, and in any year some crews did very well while others barely earned any money at all. Most of the tasks in fishing could be carried out by any able-bodied man brought up by the sea, so that men could move fairly easily from fishing into other occupations and back again. From the point of view of the individual fisherman, therefore, the herring-fishing industry was intersected by a social field through which he could move fairly freely along lines of friendship and local knowledge, seeking in the main the achievement of economic goals. Every man was in touch, or could put himself in touch, with a large number of other men, differentiated into shipowners, skippers, net bosses, cooks and others, and into good and bad, and to whom he was linked in a variety of ways. The herring-fishing industry also generated its own social field, which was influenced by ecological factors, such as the disappearance of the fish, by economic factors affecting alternative opportunities for employment and investment, and by many others. It was a social field only partly made up of an arrangement of lasting social groups. Thus in terms of this analysis we can isolate three regions or fields in the social system of Bremnes. Firstly there was the territorially based social field, with a large number of enduring administrative units, arranged hierarchically, one within another. The administration of the parish was carried on through this system, and the same boundaries were used by the voluntary associations. By reason of their physical proximity the smaller territorial units, the hamlets and wards, provided the basis for enduring social relations between neighbours, which found expression in various activities connected with subsistence agriculture, the care of children, religion, entertainment, and the like. The units of the system endured and membership changed only slowly. The second social field was that generated by the industrial system. Here we had a large number of interdependent, yet formally autonomous units such as fishing vessels, marketing cooperatives, and herring-oil factories,
72
MODELS OF THE REAL SOCIAL WORLD
connected with each other functionally rather than hierarchically, yet each organized in a hierarchy of command. These units, which often were true social groupings as well as units of organization, did not necessarily persist through time, nor did their membership remain fixed. The third social field had no units or boundaries; it had no co-ordinating organization. It was made up of the ties of friendship and acquaintance which everyone growing up in Bremnes society partly inherited and largely built up for himself. Some of the ties were between kinsmen. A few of them were between people who were not equals, as between a man and a former employer with whom he had kept in contact. Most of the ties were, however, between persons who accorded approximately equal status to one another, and it was these ties which, I think, may be said to have constituted the class system of Bremnes. The elements of this social field were not fixed, for new ties were continually being formed and old ties broken or put into indefinite cold storage. Let us examine more closely the distinctive features of this third social field. As we well know, cognatic kinship does not of itself give rise to enduring social groups. I have my cousins and sometimes we all act together; but they have their own cousins who are not mine and so on indefinitely. Each individual generates his own set of cognatic kin and in general the set he and his siblings generate is not the same as that generated by anyone else. Each person is, as it were, in touch with a number of other people, some of whom are directly in touch with each other and some of whom are not. Similarly each person has a number of friends, and these friends have their own friends; some of any one person's friends know each other, others do not. I find it convenient to talk of a social field of this kind as a network. Earlier I had used the term web, taken from the title of Fortes' book The web of kinship (1949a). However, it seems that many people think of a web as something like a spider's web, in two dimensions, whereas I am trying to form an image for a multi-dimensional concept. The image is merely a generalization of a pictographic convention which genealogists have used for centuries on their pedigree charts. Other modifications of this convention include the 'tribal sequences' discussed by Armstrong (1928: 37), 'psychological geography' in Moreno (1934: 2 3 8 47) and 'sets' in Chappie and Coon (1942: 284). The image I have is of a set of points, some of which are joined by lines. The points of the image are people, or sometimes groups, and the lines indicate which people interact with each other. We can of course think of the whole of social life as generating a network of this kind. For our present purposes, however, I want to consider, roughly speaking, that part of the total network that is left behind when we remove the groupings and chains of interaction which belong strictly to the territorial and industrial systems. In Bremnes society, what was left was largely, though not
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
73
exclusively, a network of ties of kinship, friendship and neighbourhood. This network ran across the whole of society and did not stop at the parish boundary. It linked Bremnes folk with their kinsmen and friends in other parishes as well as knitting them together within the parish. A network of this kind has no external boundary, nor has it any clear-cut internal divisions, for each person sees himself at the centre of a collection of friends. Clearly there may be clusters of people who are more closely knit together than others, but in general the limits of these clusters are vague. Indeed, one of the ways in which a cluster of people emphasize their exclusiveness is to form a group, to which one definitely either does or does not belong. The social ties linking the members of the group are then no longer merely those of kinship or friendship. In parenthesis, we may note that one of the principal formal differences between simple, primitive, rural or small-scale societies as against modern, civilized, urban or mass societies is that in the former the mesh of the social network is small, whereas in the latter it is large. By mesh I mean simply the distance round a hole in the network. In modern society, I think we may say that in general people do not have as many friends in common as they do in small-scale societies. When two people meet for the first time, it is rare in modern society to discover that they have a large number of common friends, and when this does happen it is regarded as something exceptional and memorable. In small-scale societies this happens more frequently, and strangers sometimes find that they have kinsmen in common. In terms of our network analogy, in primitive society many of the possible paths leading away from any A lead back again to A after a few links; in modern society a smaller proportion lead back to A. In other words, suppose that A interacts with B and that B interacts with C. Then in a primitive society the chances are high that C interacts with A, and in a modern society the chances are small. This fact is of considerable practical importance for the study of societies by the traditional techniques of social anthropology, when we try to become acquainted with a limited number of persons whom we observe interacting one with another in a variety of roles. In a modern society, each individual tends to have a different audience for each of the roles he plays. Bremnes, in these terms, was an intermediate society. In some societies close kinsmen and affines are not necessarily social equals, and in that case the network of kinship ties may have a steep social gradient. Similarly, in our own society, in a street with property ranging gradually from mansions at one end to tenements at the other, we can speak of a network of ties between neighbours who do not regard themselves as equal in social status. However, in Bremnes, as in many other societies, kinsmen, by and large, were approximate social equals. Furthermore, in the 1950s, unlike conditions which prevailed in Bremnes
74
MODELS OF THE REAL SOCIAL WORLD
until about a hundred years earlier, neighbours were approximately equal in social status. In Norwegian thought in the immediate post-war years, the idea of equality was emphasized, so that even between persons of markedly different economic status there was less recognition of social inequality on either side than would, I think, have been the case in Britain during the same period. Thus the social network in Bremnes was largely a system of ties between pairs of persons who regarded each other as approximate social equals. 3 Social class The organization of the population of Norway into social classes, assuming that there was such an organization, may be said to have manifested itself in Bremnes in the social network I have described. The term social class is widely used in general conversation, and naturally it has a great variety of meanings. I think that much of the confusion that grew up around the term was due to our failure to distinguish these different usages (cf. Gross 1948). Thus Marx had in mind definite groups into which the population was divided, which were mutually exclusive, collectively exhaustive, enduring through at least several decades, and which recruited members by reference to their position in the economic system. The study of class through clique membership, on the other hand, implies a view of class closer to a network. For instance, in Deep south (Davis et al. 1941), a series of overlapping cliques is used to define the boundaries of class. Most other approaches to class treat it as a kind of social category, of persons possessing approximately the same size of house, or paid about the same amount, or standing at about the same level on some commonly held scale of social esteem. Lastly there is class as a category of thought, a unit of division used when members of a society mentally divide up the population into status categories. It is clear that the question 'How many classes are there?' is meaningless when applied to class as a social category, for there are as many or as few as we choose; and there is often no consensus within society about class as a category of thought. There may be disagreement about the number of social classes when class is treated as a social group, in the same way as there are disagreements about how many genera and how many species there are in zoological taxonomy; but that is a problem that can be solved. When, however, we look at social class as a kind of network, the question of how many social classes there are falls away completely. I should perhaps emphasize that the concept of network is only one tool for use in the analysis of the phenomenon of social class. The other approaches I have mentioned above are equally valid, and indeed are necessary to any understanding of this complex social fact. As we are well
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
75
aware, there is a fair measure of congruence between the different approaches. In general, most of a man's friends have approximately the same income as he has, live in the same sort of house, are classified together by other members of the society, and fight on the same side in those political and industrial struggles in which, if at all, social classes may perhaps be said to function as groups. For the purposes of this paper I shall nevertheless look at social class from merely the one point of view: as a network of relations between pairs of persons according each other approximately equal status. This choice is not entirely an arbitrary one. It arises from the fact that Bremnes was a fairly small community, with no marked differences of culture and, with a few exceptions, minor differences in standards of dress, housing and the like. In common parlance, there were few class differences in the parish. There were significant differences in income, but these were partly rendered inoperative by the lack of significant differences in patterns of spending. Under these conditions we would not expect to find the emergence of a division of the population into distinct social classes one above the other. It should be clear that such a division is not ruled out by the idea of class as a network. It is only pairs of persons who are directly in contact with one another in a class network who regard themselves as approximately of equal status; each person does not necessarily regard everyone else in the network as his equal. Suppose that A has a friend B. A regards B as his social equal, perhaps a little higher or a little lower in social status. As we have remarked earlier, not all B's friends are friends of A. Suppose that D is a friend of B, but not a friend of A. Then, if A knows D at all, he may or may not regard him as his social equal. If, for instance, A regards B as slightly beneath him in social status, but not so far below as to matter; and if B regards D in the same light; then there is the possibility that A will regard D as too far below him to be treated on a basis of equality. Similarly if A is below B and B is below D, A may think that D is too far above him to be treated as an equal. This process is cumulative with every step taken along the network away from A along any path we choose. Thus for every individual A the whole of the network, or at least that part of it of which he is aware, is divided into three areas or sets of points. One of these sets consists of all those people to whom A is linked by a longer or shorter path, and whom A regards as his social equals. A is similarly linked to each person in the second set, but this set is composed of all the people that A regards as his social superiors. Similarly the people who in A's estimation are his social inferiors form the third set. These sets are like the sets of cousins we mentioned earlier, in that membership of the sets has to be defined afresh for each new individual A that we choose to consider. Thus in the example given above, D belongs to the superior set with reference to
76
MODELS OF THE REAL SOCIAL WORLD
A, but to the socially equal set with reference to B. I think that in some, at least, of the many instances in which people of widely varying economic position say that they belong to the middle class in a system of three (or more) social classes, they are merely stating that they are aware of the existence of these three sets of persons. It does not of itself imply that society can be divided into three groups with agreed membership. I do not wish to digress further with the elaboration of this model of a class system. The idea has been developed sufficiently to deal with the comparatively simple conditions that prevailed in Bremnes. In the first place, Bremnes was small, with a great deal of intermarriage. Hence in the network the number of links along the shortest path joining any two members of the parish was small, probably never more than four. Secondly, because Norwegian culture was egalitarian, everyone was ready to treat as an equal others whose income, upbringing, interests and occupation differed widely from his own. Thirdly, despite the egalitarian dogma, people in Bremnes recognized the existence of differences in social status. They had stereotypes of an upper class who had lived in big houses in the towns, talked a different language and had different religious beliefs. Bremnes folk also spoke of a lower class, people who wandered about unashamed, living on charity and scorning the aspirations of respectable citizens. In between these two classes were 'plain ordinary people like ourselves'. It was in fact the familiar egocentric three-class system, with Ego in the middle class. Class is here a category of thought. In Bremnes, conditions were simplified in that, for the most part, everyone appeared to think of almost everybody else in the parish as belonging to the same class as himself. Most people in the upper and lower classes, as defined by Bremnes folk, lived outside the parish. When they visited the parish, members of these classes were treated by most of the resident population as either social superiors or inferiors, but not as equals. Within the parish community, the range of variation was just sufficient for a few people, perhaps a dozen or so in number, to be regarded by many others as on the upper fringe of their sets of social equals. Yet others treated these dozen persons as social superiors. Similarly, there were a few people who were, in rather oblique fashion, treated as social inferiors by many in the community. However, this recognition of social inferiority was often masked, since it was impolite for anyone to show openly that he considered himself superior to anyone else. At the time of my inquiry, it looked as though the range of status variation was likely to increase in the future. In the community, some men were wealthier than others and although they only occasionally used their wealth to buy socially conspicuous goods and services such as cars, large houses, expensive clothes, pleasure yachts and the like, they did buy more expensive education for their children. Up to the age of fourteen all
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
JJ
children received the same education in the parish schools, but it was the sons and daughters of the wealthier section of the population who, in the main, continued their education for a few years more. For many youngsters, this meant leaving the parish and coming into contact with ideas and values different from those they had known at home. They acquired skills which, when they left school, differentiated them from the majority of the labouring population, and which in a generation or two might well lead to sharper cleavages along class lines, or at least to a recognition of wider differences in social position. Similarly, on evidence from other parts of Norway, I thought that the people lowest in the Bremnes social scale, most of whom were itinerant pedlars and beggars, were at that time more sharply distinguished from the rest of the population than had been the case a hundred years earlier, when many penurious cottars and day labourers lived in the parish; but I could not document this for Bremnes. Although there was a tendency towards greater differentiation in social status, it was slowed down by other social processes. Taxation was high, so that it was difficult for a man to amass a fortune, and since capital was taxed as well as income, it was also difficult for him to retain it. As part of the culturally supported thesis that all should be treated alike, it was universally held that all children should inherit equally. Only a third of a man's wealth could legally be disposed of by will; the remainder had to be distributed according to the laws of inheritance or intestacy. In Bremnes few wills were made and in nearly every case all of a man's chattels, after provision had been made for his widow, were divided equally among his children. Thus in a society of large families, fortunes were dispersed at death. Even death duties operated differentially on the principle of 'he that already hath more shall receive less'. Where land was concerned, one child often took over the whole of his or her father's farm, but even then he (or she) had to buy the land from the father so that the rights of the landless siblings might be protected. In education, inequalities of opportunity were to some extent offset by bursaries and interest-free loans from official sources, and by the custom of allowing adolescents to work for a year or two so that they could save enough to take themselves through the next stage of their education. All these factors hindered the speedy development of wide social differences even though the trend seemed to be in that direction. Thus in general terms we can say that in Bremnes society, apart from the territorial and industrial systems, there was a network of social ties between persons arising from considerations of kinship, friendship and acquaintance. Most, but not all, of these ties were between persons who regarded each other as approximate social equals, and these ties of approximate equality we regard as one manifestation of the social class system, and shall call the class network. Although each link in the class
78
MODELS OF THE REAL SOCIAL WORLD
network was one of approximate social equality, not everyone in the network regarded everyone else as his equal, and there were a few people in the parish who were regarded by many others, but not by all, as belonging to a higher class. The class network was utilized for carrying out social activities, such as mutual help and home entertaining; class ties and also ties between people of recognized unequal status were used by men for a variety of other purposes, for example, to find places for themselves in the fishing industry. 4 Leadership Cooperative activity requires some degree of leadership and consensus, whether carried out by enduring groups or by ephemeral groupings of persons linked by a network of social ties. Let us now consider some of the mechanisms by which leadership and consensus were obtained in Bremnes social life. Characteristically a network has no head and, as I have here used the term, no centre and no boundaries either. It is not a corporate body, but rather a system of social relations through which many individuals carry on certain activities which are only indirectly coordinated with one another. In Bremnes, as we have seen, there was little class distinction, but the social activities which were typically carried on through the system of social class were there carried on in the same way as in a society with a larger range of class variation. People invited their friends to supper, or to a sewing party for the mission, or for a shooting trip, on the same basis of apparent approximate equality of class status which is, I think, definitive of class behaviour. The network of friendship and acquaintance, when men sought out industrial opportunities, was used rather differently. Fish were actually caught, and a large number of distinct activities were brought into close co-ordination with one another. While fishing, men were no longer equals; they were organized in chains of command and differentiated according to function. For as long as the technical process demanded, they were organized in fixed groups standing in a definite relation one to another. At sea the skipper was in charge of his vessel, the coxswain in charge of his boat; they gave orders and their subordinates obeyed. In the same way the marine-engine factory was organized hierarchically for purposes of production, with a board of directors, managers, foremen and workmen. The groups of men who were thrown together on board fishing vessels or in sections of the factory developed and perpetuated other modes of interaction which modified the configurations of the class network besides affecting the productive tasks themselves. In Homans' (1950: 273ff.) terminology, there was a clear-cut hierarchically organized external system, whereas the internal system was the network of friendship and acquaintance.
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
79
Once we leave the field of organized industrial enterprise, the need for quick decisions and for a clear division of responsibility decreases. The achievement of consensus was valued more highly than speed of autocratic command. Decisions had to be made involving collective action: whether the teachers should have salmon or cod at their banquet; whether the electricity supply cooperative should take action against a member who had tampered with his meter; whether a boatload of fishermen should go ashore to the cinema or to the prayer-ship. Such decisions were important, but their importance lay more in the consequences for face-toface relations between members of the society than in their technical merits. Hence it is not surprising that the process of reaching the decision to hold a feast in the prayer-house was more complicated than that by which a command was given to cast a net in the sea. This was true whether the prayer-house was one belonging to a local community or to the hierarchically organized factory. Yet, as is usual in the Western world, most of the formal associations in Bremnes concerned with non-industrial activities had what appeared on paper to be an hierarchical structure suitable for taking quick decisions in an autocratic way. There was, it is true, no one person in Bremnes who was head of local society, who might have been called the chief of the island, but equally Bremnes was not a leaderless society. It was, as we say, a democracy, and there was a common pattern of organization which occurred in nearly every instance of formal social life. Each association had a committee, with powers to act usually for a year but sometimes for longer, elected by an annual general meeting. The committee, if it was big enough, elected a quarter of its members to an executive council to which most of its powers were delegated. The council and the committee each elected one of their members to be treasurer and secretary. There was also a deputy chairman and a number of deputy members who functioned only in the absence of the principal members. This common pattern was followed with only minor variations by sports clubs, missionary societies, producers' cooperatives and by the local government itself. All these bodies employed the same procedure for reaching a decision, by simple majority vote of those present and voting, provided there was a quorum. In practice, whenever possible they avoided taking a vote and the great majority of collective decisions were therefore unanimous. This tendency was most marked at the meetings of missionary societies and least at those of the parish council. Even in the parish council, when there was an irreconcilable division this was sometimes concealed by first taking a trial vote, to decide which view had greatest support; this was followed by a confirming unanimous vote, which alone was recorded in the minutes. Nevertheless there were in fact continual differences of opinion between members of all these different bodies. Why then was the achievement of
80
MODELS OF THE REAL SOCIAL WORLD
formally unanimous decisions considered so important? Here I think we are dealing with a principle of fairly wide application. People living and working together inevitably have conflicting interests but in general they have also a common interest in the maintenance of existing social relations. Individual goals must be attained through socially approved processes, and as far as possible the illusion must be maintained that each individual is acting only in the best interests of the community. As far as possible, that is, the group must appear united, not only vis-d-vis other similar groups, but also to itself. Voting is a method of reaching decisions in which divergence of interest is openly recognized, and in which the multiplicities of divergence are forced into the Procrustean categories of Yes and No. Significantly, voting is rare in simple societies and in small groups in modern society. Membership in a collectivity implies accepting a share in the collective responsibility for the group's actions as well as a share in the decision to act in a certain way. The local associations in Bremnes were in the main face-to-face groups operating in a conservative environment. Even the producers' cooperatives, which had been responsible for introducing technological changes, were made up of men who had been neighbours for many years, who were related by kinship and marriage, and who were not trying to alter the existing pattern of social relations on the island, even though they might be trying to alter the position of the islanders as against the rest of the region. In these conditions voting was an inappropriate procedure. Furthermore, in voting, the worth of one man relative to another is fixed, and in most voting systems all men have equal votes. When decisions are reached unanimously after discussion, each man gives his own weight to the views of his fellows. In Bremnes individuals presented their views as though they had first been stated by someone else; they spoke tentatively and cautiously; they tried to win the support of divergent colleagues by saying that they agreed with them all. The complex process by which a final decision was reached without the cleavages in the group becoming irreconcilable is one that I am not competent to analyse fully. The process was in part a corollary of the emphasis on equality that was noticeable in Norwegian culture. What is significant for our purposes is that it is a recognizable process which went on in some social contexts and not in others. There was one context in which voting by secret ballot was almost invariably used. When new committee members and officers had to be elected, pieces of paper were handed round, everyone wrote down his choice and folded over his paper before handing it in. Thus the only topic that never came up for open discussion was the relative worth of members of the community. I think there are two reasons for this. The election of committee members was the one occasion at which an immediate decision
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
8l
was absolutely necessary to prevent the structure of the association collapsing. Secondly it is difficult, though not impossible, to discuss the merits and demerits of one's friends in their presence without committing oneself so much that the appearance of general amity is threatened. As it was, most elections to local committees in Norway consisted in re-voting into office the outgoing members, and sometimes a special subcommittee was appointed to draw up in private a list of nominations, so that voting became a formality. In this way the rivalries that threatened the unity of the community were hindered from coming into the open. In Bremnes there was, however, often free discussion about who should serve on those committees involving more hard work than honour. The parish council differed from the other associations on the island in that it was required to act and could not be merely a mutual admiration society. It was the local government in a society that was changing, even though it was changing fairly slowly. The leaky church roof had to be repaired, and as the population increased more classrooms had to be built. The council was under constant pressure from the provincial administration to collect taxes and to spend the money collected. Unlike the missionary societies and chess club and women's institutes, the parish council obtained a large part of the revenue for its projects from the state, and higher authorities audited its accounts, approved its budget and bombarded it with correspondence. It might try to move slowly, but it was continually forced to come to a decision one way or the other on issues about which the community had not yet made up its mind, that is to say, about which there were differences of opinion that had not yet been resolved. Hence from time to time a vote had to be taken. The usual techniques, or as some would say, tricks, were used by the mayor to gain unanimous approval, such as, for example, making the majority record their votes by remaining seated and the minority by standing up; all those who were in two minds about the issue probably failed to spring to their feet. In the same way, members of council tried to avoid having to vote on matters in which they had divided loyalties, claiming that because of ties of kinship and affinity they would be likely to be biased and therefore could not discuss a particular matter fairly. On one occasion I observed, when an unusually controversial matter was up for discussion and three members had, one after the other, spoken briefly to say that they were related to the parties in the case and therefore could not take part in the discussion, the mayor intervened to point out that council members were probably all related in one way or another to the parties concerned, but that nevertheless they must come to a decision. Where associations other than the parish council were concerned, such embarrassing situations could usually be avoided. There was no state administrative machinery behind them to keep them going at all costs, and if serious
82
MODELS OF THE REAL SOCIAL WORLD
latent differences were allowed to become apparent, the organization might split. Formal associations in Bremnes, despite their hierarchical organization, were not authoritarian. The existing structure of social relations in a conservative environment was maintained by seeking for apparent agreement for all decisions. With the parish council, speed of decision was more important and voting was more often employed. In industrial enterprises, both in the marine-engine factory and the fishing industry, where the environment was not conservative and quick decisions were needed, there was a hierarchical structure and this was effective and not merely formal. 5 The form of parish society Although there were many leaders of part of the parish, each of whom operated in certain restricted contexts, there was no overall leadership of the parish valid in a wide range of contexts, such as we are familiar with in the primitive world. We might perhaps call the pattern of public life in Bremnes 'government by committee'. In formal terms there were no long chains of command on the island. Instead there was a host of small organized groups with overlapping membership, and the whole population was enmeshed in a close web of kinship and friendship which linked together all the people on the island, but which also tied them to kinsmen and affines scattered through western Norway, and indeed throughout the whole world. In this system the people formally in positions of leadership were the elected chairmen of the various associations. They held office for a fixed term but were very often re-elected unless they decided to resign. There were perhaps fifty voluntary associations of one kind or another, as well as about forty standing committees whose members were appointed or recommended for nomination by the parish council. All these men occupied positions of some public responsibility. Slightly more in the public eye were the mayor, the rector and his curate and the sheriff. None of these men could be said to represent the parish in its totality to the outside world, and all of them were involved with factional interests within the parish. The sheriff and the rector were perhaps most removed from internal rivalries, but even they took part in politics, although they were civil servants and directly responsible to higher authorities outside the parish. The holders of both these offices had stood as parliamentary candidates in a recent election. They were both elected parish councillors and members of various parish council committees in their own right as well as being exofficio members of other committees. Even more involved in local politics were the mayor and the chairmen of standing committees. In fact there were no living symbols of parish unity, or of the relation of the parish to other social groups, in any but a restricted sense. There was
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
83
no one person in a key position who articulated the parish with a wider social system as happens in many simple societies. The mayor came nearest to this, for he represented the parish on the provincial council, and was sometimes invited to serve as a director on the boards of public utility companies serving the parish. Yet even so he was not the representative of the parish in ecclesiastical or judicial matters or in the affairs of the missionary societies. This lack of a single leader or symbolic head is perhaps due to the fact that the parish was not a corporate group in the same way that, for instance, a minor lineage is a corporate group among the Tallensi (Fortes 1945: 99). Bremnes was a parish, a unit of civil and ecclesiastical administration and part of the kingdom of Norway. Yet even in civil and ecclesiastical affairs the parish looked outward in different directions. Ecclesiastically Bremnes was part of Finnas parish union, which was part of Ytre Sunnhordland archdeaconry, which was part of Bjorgvin diocese; in civil matters Bremnes was an immediate subdivision of Hordaland province; while in judicial affairs it formed part of Finnas sheriff's area, which was part of Sunnhordland magistracy. Two centuries ago Norway was still virtually a Danish colony governed by what we would now call a system of direct bureaucratic rule. The various sections of the bureaucracy were largely separate and the local areas into which the country was divided for different purposes then coincided even less than they did in the 1950s. During the nineteenth century, as the local population gained a greater share in public affairs, changes were made to bring the ecclesiastical, administrative, fiscal and judicial divisions into alignment, but the coincidence was still not complete. Indeed the trend was beginning to be in the opposite direction, as new systems of organization cut across existing alignments, as for example the Home Guard and the electricity supply grid. The parish was a unit in some of these different organizations, but it was not an exogamous or endogamous unit; it was not an economic unit, and from most points of view it had no culture of its own. Its nearest approach to a social centre was the parish church, and it was here that the largest crowds gathered, that common beliefs were affirmed, and changes in social life received public recognition. Yet the church was not as widely supported an institution as it once had been and its sphere of influence had considerably diminished. Although the process of social specialization had not gone as far in Bremnes as it has, say, in a London suburb, Bremnes was definitely not a simple society. The systems of organization within which the people carried on their activities were not congruent with one another. Neighbours, kinsfolk, workmates, fellow members of associations were all becoming different. In general, the mesh of the social network was growing larger. Nevertheless, the organization of Bremnes society was still largely an arrangement of cross-cutting ties and groupings in which not only friends
84
MODELS OF THE REAL SOCIAL WORLD
and enemies, but also leaders and followers were inextricably mixed. No one line of cleavage ever became dominant. The territorial system endured and the industrial system executed commands; but in this society the relations that were valued most highly were still to be found in the shifting middle ground of social intercourse between approximate equals. 6 Post-colonial societies So far we have looked at Bremnes society as an isolated object of study. In conclusion I want very briefly to consider Bremnes in comparison with other similar societies. In reality Bremnes was not an isolated society, and there is the large descriptive and analytical problem of understanding the relation between Bremnes and neighbouring parishes, and between it and the Norwegian state. These problems I shall not deal with here. The problem I want to glance at is a morphological one, of seeing Bremnes as an example of a particular type of society. The centuries of development, and of stagnation, that lie behind the Bremnes society described in this paper can be summarized as follows. From A.D. 600 until 1300, during the Viking Age and later, the Norwegian state expanded to include all Norway, Iceland and part of Sweden, with colonies further afield in Scotland, Ireland, the Isle of Man and Greenland. From 1300 onwards the state declined under pressure from Denmark, Sweden and the Hanseatic League, and suffered badly in the Black Death. In 1380 Norway was joined to Denmark by a dynastic marriage and gradually sank to the effective status of a Danish colony, with a peasantry living largely in a subsistence economy. The administration was carried on mainly by Danes; the Danish language was used exclusively for writing, and commercial and industrial development lagged behind that of Denmark. During the eighteenth century the bureaucracy in Norway began to concern itself with the development of Norwegian, as distinct from Danish, culture and began to agitate for a Norwegian university. There was a considerable divergence of economic interest between Denmark and Norway, which was increased by the vicissitudes of the Napoleonic Wars: communications with Denmark were cut by the British blockade and a Norwegian government was formed. In 1814 a constitution was drawn up, heavily influenced by the ideas of the American Constitution and the French Revolution. Instead of gaining independence the country was, however, joined in a dual monarchy with Sweden. After nearly a hundred years of political struggle this union was dissolved, without fighting, in 1905. In Viking times all free men attended local assemblies with judicial and legislative authority and later there were higher assemblies made up of representatives of the various localities. In the twelfth century the assemb-
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
85
lies began to lose their power, and control of local affairs passed to nominees of the king. Under Danish rule government became more bureaucratic, and the peasants, who provided the bulk of the army, had only the viceroy to protect them against the demands of the land-owning gentry and the Danish-speaking bureaucrats. In the eighteenth century much land passed back into peasant hands, but the common people remained unenfranchised. The 1814 constitution provided for a wide franchise and some wealthier peasants were elected to Parliament. It was, however, not until 1837 that a system of elective local government was introduced. Many of thefirstrural mayors were bureaucrats, such as priests, sheriffs and judges, but gradually more farmers and peasants were elected. The bureaucracy, which even in Danish times had theoretically been open to all, now began to have members of peasant origin. At the same time the status of the bureaucrats as leaders of local society declined and new positions of eminence appeared, the school teachers being perhaps the first group of mainly peasant origin to acquire semi-professional status. In Danish times trade was carried on in the countryside by town burghers, but by the end of the nineteenth century there had been considerable intermarriage between the peasantry and the children of burghers stationed out in the country, and some peasants had started small trading posts on their own initiative. The comparative paucity of capital for investment in the towns meant that small-scale rural enterprises were often started by wealthier peasants rather than by townsfolk. At the same time the economy of the coastal region in the west, the part of Norway to which Bremnes belongs, underwent a change. Since time immemorial there had been fishing for the home market and for export, and as communications by sea and land improved this trade now expanded. Despite considerable set-backs the size of the catch increased over the years as more capital was invested in fishing vessels and nets. Down to the end of the nineteenth century most householders on this coast were both peasant and fisherman, but with the development of commercial fishing a division began to emerge between those who were mainly fishermen and those who were mainly peasants. A fewfishsalteries owned by town merchants and a few small boat-building yards were established rurally in the nineteenth century, and barrel-making flourished as a cottage industry, yet it was only in the 1930s that large-scale rural industrial enterprises began to become important in the economy of the fishing districts. By the 1950s there were several canning factories scattered along the coast and, as electricity became more readily available, further industrial expansion took place. Bremnes, with its marine-engine factory, was more industrialized and had been industrialized for longer than many nearby parishes. Thus a couple of centuries ago there were in each rural parish one or
86
MODELS OF THE REAL SOCIAL WORLD
two bureaucrats living at a much higher standard than the rest of the population, speaking a different language, and moving from post to post without developing marked local affiliation. Below them in status were a few traders, usually burghers of a town; they had more local ties and were not so mobile as the bureaucrats. The rest of the population were peasants, with the freeholders occupying the highest stratum and accounting for about half of the peasantry. Next came the leaseholders, usually cultivating land owned by Danish-speaking gentry or by rich burghers. Under them came the cottars, or labour tenants, who worked so many days a year for the freeholders in return for the use of small holdings. Finally at the bottom of society were the landless labourers, the indigent and vagrants. In this system the peasantry, although forming the great bulk of the rural population, had little say in public affairs, were undifferentiated in terms of occupation and culture, but were divided into ranks based on inherited rights in land. This system ended, and Norway became a quasiindependent State with elective local government largely as a result of international politics and social movements among the bureaucrats and in the towns, that is, as a result of social forces external to rural society. Nevertheless, after a generation or so, peasant leadership began to emerge, the old ranking system among peasants broke down; the status of the rural bureaucrats declined, and the rural economy became more diversified. This sequence of events was in no sense the inevitable consequence of the ending of colonial rule, but it is clear that the break with Denmark supplied the initial impetus that started this train of events in Norway and even in Bremnes. It seems not improbable that similar events may have taken place in other former colonies. One of the major social movements of the twentieth century has been the partial breaking-up of those worldwide empires established by the powers of Western Europe during the nineteenth century and earlier, and there are many older empires now long since fallen into decay. Yet the problem of the rural effect of political independence does not always receive the attention it merits. When a country achieves independence, interest is at first naturally focused on transformations taking place at the centre and in the towns. When the colonial power is driven out by armed insurrection or as the result of a long political battle, the struggle for liberation is at the same time the process by which a new elite is formed to take over from the old colonial governing class. This is presumably what happened in South America, and perhaps in Indonesia. In a large country the formation of a new elite has greater social consequences for the towns and centres of government than it has for the countryside. In general colonial elites tend to concentrate in towns and military camps and are thinly spread throughout the rural areas. Hence the removal of the colonial power may not at first have much effect on rural life. Liberation brings fresh faces in board-rooms and government
CLASS AND COMMITTEES IN A NORWEGIAN ISLAND PARISH
87
offices, but the same people continue toiling in the fields. The ending of colonial government must of necessity bring about changes at the centre of the political structure of a new nation, but the effect rurally may be largely the replacement of, say, a White District Officer by a Black one. From the point of view of rural society the change to national political independence is then an event external to the system. At other times the withdrawal of the colonial power has been an event external to the social system of the colonial territory as a whole. An example of a withdrawal of this latter kind was the ending of Roman rule in Britain, when the critical conditions causing the withdrawal were to be found outside Britain. Yet, however remote from the rural areas may be the causes of colonial decline, in general the rural system will itself tend to change sooner or later. Occasionally rural change may precede political independence, as for instance in Israel, where for once we are fortunate in having more sociological information about rural conditions in transition than we have about changes at the centre. However, I think that we may say that usually the countryside lags behind the towns and the central institutions of a developing ex-colony, and changes in the pattern of rural life and family conditions come later and more slowly, if they come at all. Thus, when a country is for one reason or another left to govern itself, the effect on the rural areas may be slight, delayed or entirely lacking. A full discussion of why Bremnes society developed in the way it did cannot be attempted here. I would merely stress that the achievement of national independence was one factor of significance in that development and that the opportunities we have of studying the consequences of similar events in other rural areas should not be overlooked. The vacuum caused by the withdrawal of a governing elite may initiate one social process, as we have seen in Bremnes with the gradual emergence of part-time peasants in key positions in the structure of government and organized social life. Industrialization is a quite distinct process which has occurred and is still proceeding in countries with widely varying forms of political institutions, some of them colonies, some former colonies, and others that have not experienced colonial status for many centuries. The evidence suggests that in all of these, whatever social system they have had in the past, some form of class society develops as industrialization proceeds; or, as I would rather say, that the emergent societies can at least be described partly in the imprecise vocabulary of social class. The process of industrialization had already begun in Bremnes, and was perhaps largely responsible for such gropings towards a class system as we have noted. Comparative evidence from other societies at an early stage of industrialization, and from those in which new elites are being formed, may throw more light on the ways in which these two processes reinforce or neutralize one another.
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD: A DILEMMA OF PIETIST POLITICS
This chapter was written as a contribution to a Festschrift for one of my teachers, Evans-Pritchard. He was the son of an Anglican priest and a convert to Roman Catholicism, as well as being the author of one of the most impressive works of social anthropology, Nuer religion (1956). The community I studied in Norway was deeply committed to Lutheran Pietism; a contribution dealing with some aspect of this faith seemed to be an appropriate choice for the Festschrift. Evans-Pritchard had stressed the close connexion between the social and economic life of the Nuer and their cosmological beliefs. I tried to do the same for the Norwegians of Bremnes, but with the causal arrow pointed firmly in the direction from religious beliefs to political activity. The somewhat unexpected selection of the leader of the only explicitly religious political party as Norwegian prime minister in a coalition government in 1965 conveniently brought into the limelight some of the complexities and contradictions inherent in a politico-religious stance that had served well for an embattled minority but which now had to be adapted for use in a position of power in the national arena. Evans-Pritchard, perhaps because of his unorthodox, or at least unconventional, personal association with both Catholicism and Islam (cf. Lienhardt 1974: 303), left unresolved where, if anywhere, the boundary might be between theology and the sociology of religion (Barnes 1988b). In my paper I tried to incorporate relevant theological issues into what I hoped would be a fully sociological model of the nexus between religious belief and social action. The paper was written in about 1968 and I have left
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
89
unchanged sentences in which I used the present tense; I think that in the present context, unlike that of Chapters 5, 6 and 7, there is no danger that this usage will be mistaken for the atemporal 'ethnographic present' tense. Nevertheless the reader should be aware that there is no longer a prime minister in Norway who belongs to the Christian People's Party. Many people would say that the basic difference between Christianity and most tribal religions is that Christianity is an ethical and universal religion based on broad moral principles binding on all men at all times and places. By contrast a tribal religion is typically particularist, calling for specific performances by certain persons in delimited contexts; it does not specify how strangers should behave nor what they should believe, nor does it indicate how its followers should act outside the strictly religious situation. But if this is the basic difference, then expressed in these terms we have only a caricature of Christianity and of tribal religions as revealed by modern research. For just as in most tribal religions there are some actions that invite religious censure or activate mystical retribution, so in Christendom the form of the good life is shaped not only by ethical statements like the Ten Commandments and the Sermon on the Mount but also by cosmological beliefs about heaven and hell, the Trinity and the Atonement, as well as beliefs about postulated historical events such as the Incarnation and the Resurrection. The same is true of other so-called ethical religions. Buddhism provides a theory of reincarnation and of escape from reincarnation as well as 227 rules of conduct, and Islam has a hierarchy of angels as well as its duties of the faithful. Nevertheless many would argue, particularly in the present anti-theological and pro-ecumenical climate, that cosmological or metaphysical beliefs influence the behaviour of a believer primarily in specifically religious or liturgical contexts, in prayer and ceremonial, and do not directly determine how he behaves as a citizen in the market-place or the factory or in parliament. In these areas of activity the righteous citizen is guided by ethical principles rather than by his beliefs about the structure of heaven. Yet even in this modified form the disjunction between religious behaviour validated by metaphysics and secular behaviour determined by ethics has only limited taxonomic value. For within all the great world religions we see the historically persistent contrast between faith and works as roads to salvation. Max Weber has more than any other writer examined the sociological implications of the diverse forms taken by this contrast in the various religions, and in the light of his studies we should be alert to the probability that beliefs not only about the fate of the believer after death but about other aspects of the cosmological order as well may
9O
MODELS OF THE REAL SOCIAL WORLD
have a direct influence on behaviour that is seemingly entirely mundane and secular. In societies without writing, the past is encapsulated in the present in many w a y s - i n legends, traditions, genealogies, the pattern of enduring social relations and the continual re-enactment of ostensibly unchanging rites. Continuity in the social life of literate peoples is maintained partly by the same devices, but in addition there are written records which endure unchanged even while they are forgotten, waiting to be rediscovered and put to new uses after centuries of oblivion. Although everywhere our perception of the past is continually being revised in the light of our changing present interests, in a literate society the past still remains accessible in the contemporary records that have been preserved unchanged through time. It is significant that in his phantasy of the new barbarism, 1984, Orwell stresses both the perpetual revision of the history of the very recent past and the destruction of contemporary records; in 1984, as in a non-literate society, the past is hidden and only the latest versions of the myth are in free circulation. These considerations are particularly relevant to the study of social life in western Norway, where a long tradition of literacy, a well-developed system of public records, and a keen sense of historical continuity and development combine to make the recollected and documented past an essential element in the affairs of the present. In this context a fieldworker who necessarily spends only a limited period of time on his inquiries is at a permanent disadvantage. Under tribal conditions the ethnographer who sees the results of his investigations in comparative perspective and who can utilize evidence from archaeology, linguistics and other disciplines may well become better informed about the history and development of the people he studies than are the people themselves. In a literate and diversified society actively concerned with its own past and able to utilize the results of scientific inquiry the position is reversed. The ethnographer can never hope to read all the books all his informants have read and he cannot expect to absorb more than a fraction of the complex and diverse cultural background, partly myth and partly fact, that they have taken a lifetime to acquire. As Evans-Pritchard (1962: 64) puts it, 'Then history cannot be ignored'. The ethnographer has not only to observe and listen to his informants; he also has to use his abilities as a sleuth to pick his way through the jungle of archives and libraries to arrive at precisely those facts which impinge critically on the contemporary scene he is trying to understand. Yet this necessity is also temptation. It is an attractive and peaceful jungle with many byways and it is easy to forget just what one has come to seek. Some ethnographers may never be seen again. What follows is a description of one instance in which cosmological beliefs influence political behaviour. I shall discuss a contemporary situa-
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
91
tion in Bremnes, a district in western Norway where I have been making sporadic inquiries for a decade or so (Chapter 4; Barnes 195 7b, 1960k). By political behaviour I refer here to the way in which people vote in Parliament and on local councils and the support they give to national political parties. The cosmological belief is derived from the writings of Martin Luther, who taught that there are two kingdoms or regimes or regiments ruled over by God, a lefthand kingdom and a righthand kingdom. I make no assumption about the primacy of this or any other religious belief but merely assume that whatever material, mental or genetic causes may promote or sustain any religious belief, the belief soon acquires an autonomy of its own. In 1965, after twenty years of almost unbroken rule, the Labour government in Norway was defeated in a general election and was replaced by a governing coalition of four right and centre parties, one of them being the Christian People's Party. This is the strongest political party in Bremnes. Throughout Norway the party is mainly supported by Christians who follow that version of Lutheranism introduced into Norway at the very end of the eighteenth century by the Pietist evangelist Hans Nielsen Hauge. Pietism has been dominant in Bremnes for the last seventy or eighty years and enjoys even more local support than does the Christian People's Party. The post-war Labour government clashed with the Pietist movement soon after it came to power over the control of a teachers' training college, but in its later years one of the strongest conflicts concerned subsidies for schools for adolescent boys and girls run by Pietist organizations. The Labour government refused to provide funds for the schools. One of the first acts of the new incoming government was to provide this financial aid. During the Labour period, the Pietists argued that education, particularly the moral instruction and character-training provided for adolescents in these schools, was a function of the Christian community and not a matter that could be left to the state, which at that time they saw as being controlled by a government covertly if not openly anti-Christian in its sympathies. But it was only right, they claimed, that the state should encourage this good work for the community, work which was otherwise almost entirely dependent on the financial sacrifices made by individual Christians. Now that the Christian People's Party is part of the new coalition government, with the Minister for Church and Education a member of this party, many Pietists realize that the same arguments may be put forward by Catholics, Seventh-day Adventists, Mormons, Jehovah's Witnesses and other smaller groups, each group demanding subsidies for its own schools. Pietists in Bremnes and elsewhere tend to regard with suspicion all non-Lutheran religious teaching and in particular recall Luther's identification of the Pope as an anti-Christ. Hence opinion in the party is divided. Some members argue that the secular state
92
MODELS OF THE REAL SOCIAL WORLD
has no authority to discriminate in religious matters and that, to be fair, if it subsidizes the schools of one religious persuasion it must subsidize all; others argue that the state should aid only those citizens who preach the word of God and not those who preach false doctrines. Both opinions have their advocates in Bremnes but the majority of party members hold that accepting grants from the state for Pietist schools does not entail agreeing to financial grants to the schools of other denominations. These are the facts I discuss. In this paper I am not concerned with the ethical problems they pose and my interest is limited to the evidence they provide about how the people of Bremnes participate in the national political system. To analyse the facts in terms of this system I have to range far afield over the historical development of church and state in Norway. In the context of this short paper I must necessarily treat this development very superficially, but my aim is to provide merely a brief analysis of contemporary events rather than an adequate historical and theological account of their antecedents. To understand how this division of opinion within the party has come about, and why Bremnes party members incline to what we may call the particularist rather than the universalist alternative, we must look both at the local scene in Bremnes, and at some of the distinctive features of Pietism and Lutheranism in Norway. These relevant features are the relations of the Norwegian state church to the secular government, to the Lutheran tradition, and to the Pietist movement. I mention how the Pietist movement, which is anti-clerical in sentiment, has become responsible for training priests, and why it runs schools of its own. In Bremnes, I sketch the local pattern of industrialization and the varying amount of local support for Pietism. The first question to ask is, why does the Christian People's Party enjoy solid support in Bremnes? The party was founded in western Norway in 1933 and arose out of a feeling that the existing parties had all become inadequately Christian. Bremnes is a district which, within this region, has a reputation for being markedly Christian and it is not surprising that the new party attracted strong local support. However, this salient interest in religion is a phenomenon that has changed with the passage of time. It seems probable that while Bremnes now stands out among the neighbouring districts of western Norway for the support given to prayer houses, missions and other voluntary church activities, this prominence is due more to a recent decline in these activities in other districts rather than to an increase in Bremnes. The trend in Bremnes is in the same direction as elsewhere, but it is less pronounced. It seems thatfiftyyears ago there was a greater degree of consensus throughout the region than there is now about the authority of the text of the Bible and the importance of regular attendance at religious meetings, saying grace before and after meals, and
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
93
abstinence from alcohol. These traits constitute some of the distinguishing external characteristics of the present dominant culture in Bremnes but are not universally followed even there. There are many who go to church and the prayer house only infrequently, who go hunting on Sundays, and do not say grace. Apart from these people who accept the moral superiority of the dominant culture without meeting all its behavioural demands, there is now a substantial minority which actively rejects the dominant culture. In this local context support for the Christian People's Party in Bremnes is an expression of support for the dominant culture in the face of increasing local attack and erosion. On the national level, however, support for the party cannot mean, as it still does in Bremnes, opposition to cinemas, theatres, dancing, sex education in schools, and the building of the secular youth clubs. On the national scene these battles have already been lost and a minority party, even as part of a ruling coalition, cannot hope to reopen the attack. The defenders of the old values can envisage Bremnes only as a remnant of Christian living in an un-Christian nation. If the coalition government is a success, and if the party increases its support, there may perhaps be talk of a national Christian counter-offensive; but this is merely a possibility for the future. The programme of the Christian People's Party includes, as the second of its eight basic points, the improvement of primary schooling and the provision of state subsidies for independent schools for adolescents. The first of its points includes the preservation of Christian instruction as the core of primary school education and the denial of official support to all anti-Christian institutions (Bondevik 1965: 69-70). Other points deal with Christian social justice, temperance, support for agriculture and fisheries, disarmament, equitable taxation and prudent state spending, and impartial Christian democratic courts. Thus the party is distinguished from other coalition parties, as well as from the Labour opposition, in its advocacy of the application of Christian principles to the business of government. This is no new development in Norwegian Pietism, for Hauge, and Gisle Johnson after him, both stressed the importance of Christian witness in the secular world and were opposed to any withdrawal from it. The party gives a general and unspecific endorsement to economic free enterprise, coupled with a determination to improve the structure of the welfare state set up by the former Labour government; the main plank of the party is the protection of Christian values in private life. The public business of government and the private life of the individual meet most forcefully in the national schools. In Norway the great majority of primary and secondary school children, and in a country district like Bremnes all these children, attend schools that form part of the loose national system. The local district council is responsible for building schools, hiring and paying
94
MODELS OF THE REAL SOCIAL WORLD
teachers, and choosing between alternative syllabuses offered by the central government. Education accounts for over half of the expenditure of the local district council, and parents take an active interest in what is taught to their children in the local schools. Furthermore the national school system, as we may call it, has always had a close link with the state church. In the mid-nineteenth century children were taught to read, and adults who were too old to go to school taught themselves to read, so that they could study the Bible. Apart from the three Rs, the only other subject of importance taught in school was Christian knowledge. The district parish rector was, and still is, ex officio a member of the local school board, and instruction in school was linked with the instruction given by the rector to confirmation candidates. For many years compulsory schooling was limited to seven years, from age seven to age fourteen, and for many boys and girls confirmation during the last year of school marked the transition from childhood to the world of earning a living. The transition was often from the shelter of a Christian home and Christian school to the rough and tumble of life in the fishing fleet or in domestic service in the towns. During the twentieth century the school curriculum began to take in many new subjects. Even in Bremnes some children continued their schooling beyond age fourteen, sometimes going on to secondary school after a year with thefishingfleetto save a little money. A retired school teacher summed up the changes in the period 1900 to 1950 by saying: 'Christianity used to be the major subject and now it has become just a minor subject'. Thus the concern of the Christian People's Party with education is easy to understand, particularly at a time when post-Sputnik pressures, in Norway as elsewhere, have led to enlarged curricula, longer periods of schooling, overcrowded classrooms, and an acute shortage of school teachers. This tradition of education based on the Bible goes back to the beginnings of Lutheranism, with the work of Francke at Halle University in Germany. In nineteenth-century Norway, at least in the countryside, literacy and piety went hand in hand. Many of those who were interested in religion became teachers, and the teachers who succeeded professionally were often lay preachers. As the syllabus came to include more subjects, the Christian component in teacher selection and training became less significant. In the last few years, the shortage of trained teachers has led to the employment in primary, and even in some secondary, schools of so-called 'students', young men and women who have recently passed the matriculation examination but who have not yet gone to teachers' training college or university. Thus the school can no longer be seen as primarily concerned with the transformation of Christian infants into Christian adults; it has become a school of skills rather than a school of values. In these circumstances we can understand why the
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
95
Pietist movement seeksfinancialsupport for its own schools to supplement those provided by the state. To understand the attitude of the party towards state aid we must also look at the relation between church and state. Here we are not specifically concerned with Bremnes, for the broad outline of the relation between local church and local organs of central government within the Bremnes arena is common to all districts in Norway, though there are some special features in Bremnes we shall look at presently. The Lutheran church in Norway is a state church in that there is a government department responsible for church affairs (it is significant that this is the Department of Church and Education); its clergy are civil servants and are paid from official funds; all citizens are members of the church unless they have formally registered themselves as non-members. These conditions have persisted more or less unchanged from the Reformation through the centuries of Danish and Swedish rule into the present era of independent nationhood. Not only is there a state church but all citizens, except registered non-members, are required under the constitution to bring up their children in the Lutheran faith. Hence atfirstsight it is surprising that the argument, familiar enough in other contexts, that the state should not concern itself with religion has any relevance for this particular state and this religion. Indeed, a simple identity of the interests of the Lutheran evangelical Christian church and the Norwegian state was implied in the article of the constitution, now amended, which used to read 'Jews and Jesuits must not be tolerated'. But in fact we are here dealing with a Lutheran church which has become in certain parts of the country strongly Pietist in sentiment, and not, say, with an Eastern Orthodox church constitutionally able to live in century-long harmony with a Byzantine Christian monarchy. Nor has there ever been in Norway a Western mediaeval type of tension between church and state. The Reformation, and the new relationship between an Evangelical or Lutheran church and a Protestant prince, were hammered out in Germany and the results were imported ready-made via Denmark to Norway in 1537. During the three hundred years following the Reformation, the church in Norway became an Erastian state church, and was in fact the most pervasive instrumentality of the central government, with its clergy stationed farther out into the countryside than any other branch of the bureaucracy. In the eyes of country folk it became strongly identified with the Danish colonial establishment and urban elite and, eventually, with the dangerous doctrines of the Enlightenment and theological rationalism. It was with this church already in existence that the Pietist movement brought to the Norwegian countryside at the beginning of the nineteenth century not only its own version of the Lutheran tradition but also an interest in the basic tenets of Lutheranism which in Norway seems not to
96
MODELS OF THE REAL SOCIAL WORLD
have accompanied the Reformation itself. Hence it comes about that during the last hundred and fifty years Pietist arguments based on a Lutheran interpretation of the Scriptures have been advanced in radical criticism of the status and teaching of the manifestly Lutheran church and of the ostensibly Lutheran state. This fact, together with Hauge's forthright disapproval of sectarianism, helps to explain why, to such a substantial extent, Pietist dissent has been contained within the institutional fold of the state church. The position is well stated in the proud slogan of the principal organ of the Pietist movement, the Inner Mission: 'In, but not under, the state church'. How is this relation between church and state derived from Luther's teachings? The Dano-Norwegian institutional arrangement of a national church with total membership and assimilated to the state public service follows one of the several solutions outlined by Luther for settling the problems of church and state. It was the solution followed throughout Protestant Germany. Nevertheless Luther's views on this matter shifted substantially in the course of his lifetime. It is difficult for the outsider to summarize his intellectual position, for some commentators argue that his teaching must be understood in the light of his conviction that the end of the world was imminent (e.g. Forell 1954: 15), whereas others argue that he looked back to primitive Christianity rather than forward to the millennium (e.g. Wolin 1956: 34-5). In these circumstances it is apparent even to the non-theological anthropologist that Luther did not develop a clear theory of worldly society. His voluminous writings provide authoritative ammunition for those advocating various other conflicting solutions for the worldly status of the church, particularly when the state can no longer be conceptualized as the prince who happens to be a Christian. Here I am concerned only with the interpretation of his teaching accepted by Pietists in Bremnes. Luther discusses at length the duties of the Christian, as a 'world-person' rather than as a Christian, to obey the civil authorities, but he also criticizes these same authorities bluntly. He writes: The offices of princes and officials are divine and right, but those who are in them and use them are usually of the devil. And if a prince is a rare dish in heaven, this is even more true of the officials and court personnel. (LW 13: 212; cf. WA LI, 254, 10-13; Mueller 1954: 55)
This attitude of critical obedience was found in western Norway during the nineteenth century, where it was extended to the church and translated into action by Pietists who would gather at the church an hour before the service was due to begin, to pray for the rector who was to preach and whose theology they mistrusted. The policies of the Christian People's Party are directly influenced by
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
97
Luther's doctrine of God's two governments or regiments (cf. Thomson i960). I speak with hesitation on matters outside my competence, but it seems, according to Cranz (1959: 159-73), that in his 'Commentary on the Sermon on the Mount' (WA XXXII, 299-544; LW 21: 1-294), Luther argues that every Christian exists simultaneously in two realms or governments, a spiritual realm where he is subject to God's spiritual government and a worldly realm where he is under God's worldly government. The offices, or in modern terminology statuses, of prince, judge, lord, servant, wife, child, all belong solely to the worldly realm, even though the individuals who fill these offices may be Christians. In the worldly realm God rules through reason, whereas in the spiritual realm God rules alone. The temporal organization of the church on earth, with its offices of bishop and priest, is part of the worldly realm and while it, along with the family and the state, is ordained by God as part of the worldly realm, the church does not occupy any special position within that realm. The relative status of the two realms emerges clearly in a passage from a sermon preached by Luther on 15 December 1532. Here he says Likewise also secular government may be called God's kingdom. For he wills that it continue and that we be obedient to it. But it is only the kingdom of the left hand. His rightful kingdom where he himself rules and where he appoints neither father nor mother, emperor nor king, henchman nor policeman but where He is himself the Lord is this: where the Gospel is being preached to the poor. (WA XXXCI, 385, 6-11 andLII, 26, 21-7, cited in Mueller 1954: 43; cf. Cranz 1959: 172, f. n. 195) The ambiguity in the word 'rightful' is present in Luther's original German text. Hertz (i960) seems to have overlooked this nice example of the pre-eminence of the right hand, in the field of religion itself. It is important to note that Luther does not equate God's left hand with damnation, as in Matthew 25, v. 31-46, but with secular life here on earth. With this relationship between church and state in mind, we can look again at the distinctive features of the Norwegian social scene, remembering that these Lutheran doctrines form part of the contemporary thinking of Pietists in Bremnes and other parts of Norway. Most of the books and periodicals on the shelves in Bremnes homes are religious works, and these usually include expositions of Lutheran and Pietist doctrine. At the beginning of the nineteenth century the Pietist movement was introduced into Norway by Hans Nielsen Hauge, and gained much support in the west of the country, including Bremnes (Hamre 1964; Molland 1951; Straume i960). The movement stressed the value of personal experience of Christian salvation and personal morality and minimized the distinction between clergy and laity. It became a movement of believers, the leaven within the larger body of members of the state church. It was opposed to the liberal interpretation of Christian doctrine then espoused by many
98
MODELS OF THE REAL SOCIAL WORLD
clergy. In many parts of the country it was associated with the nationalist movement that followed the adoption of the 1814 constitution and the transfer of Norway from Danish rule to Swedish tutelage following the Treaty of Kiel. It was also associated with the movement to revive the indigenous dialects of the countryside in a composite country language in opposition to the Danish speech associated with former colonial rule. The history of Pietism and these related movements in Bremnes and elsewhere does not concern us here, except to note that by the beginning of this century this laymen's movement had gained sufficient support in the nation to challenge the liberal clergy on their own ground by setting up a theological college of its own in Oslo. During the twentieth century the supporters of liberal views have largely withdrawn from the church altogether, and at the present time the great majority of priests entering the Norwegian church receive their training in this college; yet Pietism remains essentially a laymen's movement and is referred to in these terms. It remains committed to the view that service as a priest is no higher a vocation than any other acceptable to God. The main body of the Pietist movement is no longer actively anti-clerical or against the state church, and a great deal of effort is made by influential Pietists to ensure that if possible priests appointed to churches where Pietists are numerous are themselves of Pietist persuasion or sympathy. Pietists often have an ambivalent attitude towards the church and its officials. This was well expressed in a speech given at a farewell feast in honour of the Bremnesborn man, a Pietist, who had become a priest and who, after completing a period of duty as curate in Bremnes, had been promoted to a rectorship in another parish. The chairman of the prayer house, where the feast was held, expressed the feelings of the neighbourhood by saying that all were very sorry that he was leaving, though they had to be pleased that he had secured promotion. The curate was admonished that however far his talents might take him in the ecclesiastical hierarchy, he must always remember that he was really a layman in priest's clothing. This remark was not only the highest tribute the chairman could pay, for the believing laity are the salt of the earth, but also a warning that the outward signs of ecclesiastical rank belong only to God's worldly regiment and not to his spiritual kingdom. As well as a theological college, the Pietist movement founded many other enterprises, such as missionary societies and temperance hotels and daily newspapers, which do not come into our story. Two kinds of activity do, however, bear directly on the plight of the Christian People's Party that we are trying to analyse. Throughout the Norwegian countryside followers of the Pietist movement set up prayer houses where meetings for prayer and Christian witness were held, organized by the local group of believers and addressed by their own members or by itinerant preachers
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
99
sent out by regional Pietist organizations. Meetings were arranged so as not to clash with the official services held in the church, and the local clergy, if sufficiently Pietist in their views, were often invited to speak in the prayer houses of their parish. They then spoke as personal Christians, speaking to their equals in the sight of God, and not as civil servants appointed by the king to preach God's word. The prayer houses came to be the local headquarters of a territorially organized national movement, the institutional aspect of Pietism as it were, which paralleled the territorial organization of the state church. The church was official, included everyone, provided facilities for baptism, confirmation and marriage, and was authoritarian in as much as clergy were appointed ultimately by the king on behalf of the central government. The prayer house was voluntary, was supported by those who had been born again in Christ, and was egalitarian, at least between men; women were inferior to men in both systems. This combination of organizational strength and lack of commitment to the ideological institutions provided by the state led naturally to the decision to establish voluntary Christian schools for adolescents. Just as the prayer house supplements but does not replace the church, so the voluntary schools supplement but do not replace the schools provided by the state. Examples provided in Germany and Denmark, together with various practical considerations, dictate the choice of boarding-schools for adolescents, attended for six-month courses at the age of sixteen to eighteen or so. Whereas confirmation at the age of fourteen is a ceremony of the state church, embracing all children as a matter of course, the voluntary schools provide an opportunity for those beginning their lives as autonomous adults to demonstrate that they have decided, or are deciding, to be children of God rather than children of the world. In a sense it is the pervasiveness of Christian instruction and orientation in the state schools that necessitates the creation of voluntary Pietist Christian schools. These general conditions have manifested themselves in specific form in Bremnes and have affected local attitudes to the recent dilemma. We have mentioned that until recently a high degree of conformity to Pietist principles was achieved in Bremnes, possibly to a greater extent than in neighbouring districts. In part this was probably because until the beginning of the century Bremnes remained an undifferentiated fishing and small-farming district, exporting to the towns of Norway and to the United States of America those who were ambitious or who would not conform. Casual labourers came to the district in connexion with fishing and mining and many of these were scarcely children of God, but no permanent settlements of migrants were established. After the turn of the century industrial enterprises began to be established by local entrepreneurs whose managerial drive was matched by their steadfast faith in the
IOO
MODELS OF THE REAL SOCIAL WORLD
tenets of Pietism. Indeed, the teachings of Hauge on the importance of commercial and industrial initiative in the Pietist way of life provided an ethical model for industrialization in Bremnes. The outward forms of Pietism were enforced with sanctions just as its inward manifestations were encouraged through prayer meetings and religious concerts. For example, under Norwegian law hard liquor can be bought only at the retail branches of the official Wine Monopoly, the nearest branch to Bremnes being sixty or seventy kilometres away in Bergen. This arrangement has been in force since the 1920s. Liquor ordered by workers in the largest factory in Bremnes was sent by coastal steamer from Bergen to its port of call nearest Bremnes and was there trans-shipped for delivery to the hamlet where the factory is located. The founder of this factory was also skipper of the boat used for trans-shipping goods. Stories are told of how, when he discovered a case of liquor addressed to one of his employees, he sank it in deep water, refunded the cost to the employee, and dismissed him forthwith. More generally, the high degree of local control in financial matters, including the granting of subsidies to voluntary bodies, enabled the Pietist majority, through the local district council, to protect Bremnes from the secularizing tendencies of the second quarter of the twentieth century. In other parts of Norway, and even in another part of the ecclesiastical parish that includes Bremnes, there have been various breakaway movements from Pietism that have led to the establishment of independent Lutheran churches or to sects that are not even Lutheran. Until about five years ago, none of these had gained a foothold in Bremnes. The only public buildings in the district where meetings could be held were the church, the prayer houses and the schools, and a single lodge belonging to a Christian temperance organization. Since assemblies out of doors are usually not feasible in the Bremnes climate, it was almost impossible for organizations openly hostile to the dominant culture to hold public meetings. Unapproved activities such as dancing by adolescents took place only clandestinely in boat houses or at the chilly cross-roads. It was therefore easy for the dominant group to be intolerant in matters of faith and morals. In the last five years or so, the local battle has entered a new phase. The introduction of television has brought a keener recognition of the possibilities of other styles of life. Better communications have brought the cinemas and dance halls on neighbouring islands within easy reach of young people, who now drive cars from ferry to ferry whereas their parents went all the way by row-boat. More significantly, the economic prosperity that Norway has enjoyed during the last ten years and the provision of abundant electric power to Bremnes from the mainland have led to a shortage of workers in the expanding industrial enterprises in the district. Factories and workshops can no longer be run patriarchically and the
THE RIGHTHAND AND LEFTHAND KINGDOMS OF GOD
IOI
proportion of the population attending the prayer houses seems to be declining. There is, however, still no youth club and no public dancing, and Pentecostalists are not allowed to use school buildings for their meetings. On the local front the supporters of the dominant Pietist culture are hard-pressed but have not yielded ground. Diminution of local support has, as we have seen, been accompanied, somewhat unexpectedly, by national success. Bremnes supporters of the Christian People's Party, after long years in opposition, find themselves forced to discuss national political issues in a context that is unfamiliar and radically different from the one they face at home. There are some party members who do not accept that new tactics are necessary and who hold that the sole function of their representatives should be to preach the word of God in parliament to their colleagues in other parties. But the very existence of a Christian political party is based on the recognition that Luther's notion of secular government by a prince who happens to be a Christian no longer applies to Norway. The party must do more than preach; it exists to restore by legislation the distinction between God's spiritual government and his worldly government. In particular it endeavours to ensure that the worldly government does not intrude into what properly belongs to God's spiritual government. Its policy is that Christians shall render unto Caesar only those worldly external things that are his and shall render unto God those internal things that are his alone, particularly the minds of schoolchildren. The policy and tactics of the party while in opposition, resisting the efforts of an anti-Pietist government to secularize the school system further, were clear enough. Now that the party is in power, the correct course of action is not so easily determined. In opposition the party could hold uncompromisingly to its principles but in power it is faced with the common political necessity of compromising if it is to translate any of its principles into practice. The difficulties are particularly great for the supporters of the party in rural areas like Bremnes, who are faced with a sudden change of scale in the relationships they have to consider, and with an unfamiliar political arena where they cannot estimate in advance the likely profit and loss from compromise. One solution to these difficulties is to appeal to the second clause of the Norwegian constitution, which establishes the evangelical Lutheran religion as the official religion of the state, and to argue that so long as this clause stands, Lutheran organizations, including those that are Pietist, have a special claim to state support. Yet to link the Pietist movement with the constitution in this way would seem to lead it into the trap that caught the church when it became an organ of the state. It would convert the laymen's movement into part of God's worldly government, the lefthand kingdom of God, and remove it from the righthand kingdom and the direct experience of salvation. But it is too much to expect Bremnes party
IO2
MODELS OF THE REAL SOCIAL WORLD
members readily to abandon a claim for special treatment for Pietist organizations and to advocate state support for all schools, including those run by Catholics. This is a step that even the secular Labour government did not take. The dilemma remains. The complex task of reconciling ideology and organization is met with in many contexts. Many situations can be found that parallel the one I have described. The similarities and differences between the Bremnes dilemma and those faced in one-party states based on anti-institutional ideologies are particularly striking. These comparisons lie beyond the scope of this paper. All I hope to have demonstrated is that in order to understand political action in a literate society, with its past encapsulated not in the malleable material of genealogies and myths familiar to us from tribal societies but in the more refractory stuff of books and printed articles of belief, it may be necessary to go back even to the text of a sermon preached four hundred and thirty-five years ago. More important, in large-scale societies, political decisions have to be referred to the various arenas in which they are arrived at and to which they apply, and these do not always coincide. The criteria used to decide between alternative actions may be broad ethical principles or may be derived from specific cosmological beliefs. The goals or pay-off may be here on earth or in the realm of the spirit or, as in the present case, both at the same time.
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION WITH SPECIAL REFERENCE TO AUSTRALIA
So far, we have been concerned in these chapters with social institutions existing within a comparatively homogeneous society. We have progressed from the face-toface relations of kinship, through the somewhat largerscale relations with lineages and between residents in a rural local area, to those found between national political actors. Despite the variations in scale, all these systems of relations have involved aggregations of people who, despite differences in allegiance, have nevertheless shared a broadly specified common culture. In the next two chapters we move on to consider models spanning more than a single society and characterized by cultural heterogeniety and inequalities in power and resources. Both the chapters deal with aspects of colonialism. My first experience of colonialism was in central Africa, in what were then the British protectorates of Northern Rhodesia and Nyasaland, now Zambia and Malawi. My Oxford doctoral dissertation was a study in the conquest of an African tribe and its incorporation into a colonial regime (Barnes 1967J). My understanding of colonialism had been deepened through contact in London with Keith Hancock, and I was impressed with his treatment of colonial frontiers in his Survey of British Commonwealth affairs (1937-42). At a conference on race relations in 1954 I presented a model of colonial expansion and consolidation based on evidence from throughout southern Africa (Barnes 1955c). After visiting Papua New Guinea, then under Australian rule, I tried to generalize this model to take account of the differences between the indigenous societies that had been conquered in the two areas and of the differences in the power available to the 103
IO4
MODELS OF THE REAL SOCIAL WORLD
two sets of colonizers. In addition, bearing in mind the Dutch 'police action' in Indonesia in the late 1940s, as well as numerous violent independence movements elsewhere in the world, I introduced the notion of a one-sided war. If both sides in a conflict perceive themselves as fighting a war, each grants to its enemy a formal equality of status which legitimates its opposition. In general, colonial powers under attack have strenuously denied this legitimacy to their subject populations. Although there are now comparatively few recognized old-style colonies left, one-sided wars continue to be fought all over the world.
1 Loss of indigenous sovereignty In an age of crumbling empires, when most former colonial territories have become politically independent, it may still be instructive to study the processes by which empires and nations were built up. What political forms were involved and how were they modified and developed? The new nations of Africa and Asia that emerged after 1945 have political structures that bear a generic resemblance to the established democracies and dictatorships of Europe. These forms of organization seem to be a necessary requirement for membership of the United Nations, and are widely held to be essential if there is to be a flexible industrial and commercial economy, centralized administration and widespread literacy. None of the United Nations enjoys an entirely subsistence economy or relies only on communication by word of mouth. They all have courts of law, standing armies and at least the beginnings of a bureaucracy. In particular, all the United Nations are nations, or states; there are no stateless societies among them. Thus they are drawn from only one broad category of known and possible political structures, a category that has come to dominate the world scene. Many other forms of political life are known from the past and have been described for the more remote parts of the world in more recent times. Those people who followed, or still follow, these other forms of social organization we call by a variety of terms, of which 'tribal' is perhaps the least presumptuous. Tribal political structures declined both as the tide of empire rose in the nineteenth century and, for different reasons, as these same empires withered away in the twentieth century. Erstwhile tribal peoples either were absorbed into existing economic and political systems of Western type or sooner or later copied them for their own use. Much has been written about the adoption of Western forms of organization by nations newly independent. Here I shall consider what happened to tribal political life under colonial impact at a stage when
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
105
independent nationhood was either a very distant vision or not an objective at all. In particular I shall compare conditions in various other parts of the world with those found in Aboriginal Australia and Papua New Guinea in the late 1950s. By that date most tribal people in Africa, America and Oceania already lived in some form of association with Europeans or with others who had adopted at least the technology if not the social institutions of Europe. With few exceptions Europeans and their tribal neighbours formed plural societies in which the two main segments of the population differed in ways of life, values and customs. In general political power was divided unequally; the Europeans were conquerors and the tribes were conquered. There were few tribes who could still make war on their own account, or who were interested in making war only against other tribes. In the political life of modern states the ability to make war is a good test of sovereignty; the test works fairly well for the tribal world if we extend the notion of warfare to include feuding and organized revenge. Almost everywhere except perhaps in parts of New Guinea and South America legitimate war-making had become the prerogative of those few political entities which recognized one another as sovereign states. There were somewhat more than one hundred of these in the whole world. Warlike acts by other groups or by individuals were categorized by these sovereign states not as war but as revolt, rebellion, insurrection, sedition, terrorism, banditry, brigandage, mutiny, piracy, faction fighting or murder. In the struggle for political independence these were, and still are, terms of denigration, not euphemisms; terrorists and bandits are not accorded the privileges of prisoners-of-war. Two or three hundred years ago, and further back in time, there were many more war-making bodies than there are now. Warfare was an activity much less co-ordinated than it later became and the right to make war was more widely recognized. In those days the attacks by the nations of the West on non-Western peoples who had not yet been brought under Western control were perceived as wars, not as police actions. Where warfare was avoided, treaties were made between high contracting parties who in formal terms were assumed to be equally free and sovereign, even if in fact one party was a European leviathan and the other a non-European midget (cf. Barnes 1967J: 68-72). The trend was for tribal peoples to lose their sovereignty to the West. In many other aspects of social life, tribes adapted themselves to conquest by a bewildering and fascinating variety of responses. Yet in political affairs a uniform trend is quite clear; desite small and short-lived exceptions, tribal political systems were eroded. In very many instances it was still possible in the 1940s and 1950s to study as a going concern the contemporary economic system of a tribal or erstwhile tribal group, or its kinship structure or its religious beliefs. Only in a limited sense could one study its
IO6
MODELS OF THE REAL SOCIAL WORLD
political system. There was almost always some pax (britannica, australiana, or the like) which had to be assumed and whose existence affected profoundly the doings of the people being studied. Eyewitness accounts of primitive warfare belong in the main to the nineteenth, not to the twentieth, century (cf. Turney-High 1949). Even when warfare did still take place, it was usually what we might term 'one-sided' war, in which, although one side saw itself at war with an external enemy, the other regarded the disturbance as merely domestic trouble requiring police action. 2 Analytical framework The widespread erosion of tribal political systems provides our data. Although the general trend is manifest, the process of erosion was not uniform. We therefore ask if there are any systematic variations in the process as between one tribal people and another, one historical epoch and another, one mode of conquest and another. For a preliminary analysis, this complex question can be simplified by considering only two main variables. Indigenous political systems can, without too much injustice, be divided among three categories; the diverse processes of White conquest can, with perhaps rather more injustice to the facts, be arranged into four phases. Fortes and Evans-Pritchard's Introduction' to African political systems (1940a: 6-7) provides a convenient classification of tribal systems generally. First there were those groups of food gatherers who lived in small, politically autonomous units. I shall mention the Bushmen of southern Africa, the Australian Aborigines and the Siriono of Brazil. Secondly there was the category of primitive states, typified by the Zulu, Ngoni, Ashanti and Nupe, in which there was centralized authority and other attributes of statehood. In many instances the size and precise location of a primitive state seems not to have been determined by its environment, so that the unit wasfissile,with portions frequently breaking off to join neighbouring states or to found new independent states. There were other primitive states where, although control of the unit was disputed and its boundaries fluctuated, the unit itself remained relatively intact. An example was the Barotse empire in what is now Zambia. Thirdly there were those stateless agriculturalists who had no centralized authority. Examples were the Nuer, Tallensi, the Plateau Tonga and the peoples of the Papua New Guinea highlands. The other variable, phases of conquest, provides four categories (Barnes 1955e: 169-70). We distinguish the presence or absence of a colonial administration and of settlement involving the alienation of substantial portions of land. If both an administration and settlers were absent, any
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
IO7
dealings that the tribal society might have had with a Western country were regarded by both parties as external affairs. This phase prevailed in the early days of White settlement at the Cape of Good Hope, in West Africa until the late nineteenth century, in the Pacific during the voyages of exploration, and in North America before treaties began to be made. Treaty-making usually marks the end, or one possible end, of this phase. A treaty between a conquering Western power and a tribal society usually differed in two respects from the treaties generally made between two Western nations. Almost all treaties with tribal peoples were agreements that the Western power should take the tribal group under its protection. In the heyday of imperial expansion in Africa after the Berlin West African Conference of 1885, British agents were provided with a supply of printed treaty forms which could be completed with the names of the chiefs and tribes they met with on their travels (cf. Hanna 1956: frontispiece; Perham 1956: 671). Because of its content, each tribe usually made only one treaty; there was never opportunity or occasion for another. This generalization is not true of North America, where the indigenous tribes were regarded, at least in the eyes of constitutional lawyers, as foreign nations with whom treaties could be made even after they had been in close contact with White settlers for many decades. Treaty-making was usually followed by the arrival of one or more representatives of the conquering or colonizing power, from the solitary Adviser or Resident, as in an Indian princely state, to the complete apparatus of Agricultural Departments, Magistrates, Labour Officers and the like in many an African colony. Colonizers did not always make treaties before they arrived. Sometimes agreements of a less formal kind were made, as with the Ngoni in what is now northern Malawi. Sometimes the colonial power extended its influence gradually and more or less peacefully, as in the highlands of New Guinea or in eastern Samoa; or it followed on the heels of military conquest, as with the Fort Jameson Ngoni, the Maori, the Ashanti and the Kaffirs of eastern Cape Province. Whether the colonial administration established itself peacefully or by force, it usually introduced a regime of what we may term administrative paternalism. Administrations were not always in the van; sometimes the settlers were ahead. On the northern frontiers of the Cape settlement during the eighteenth century, and in most of Australia most of the time from 1788 until, say, 1900, European settlers on the moving frontier were in contact with the indigenous inhabitants and of necessity had to take the law into their hands even as they took the land with their own hands. Colonial administrations operating without the complication of settlers were generally paternalistic, and often benevolently paternalistic, towards the tribal people they conquered; settlers were on the whole forced to take a much shorter-term view of their own interests. They had to concentrate on
IO8
MODELS OF THE REAL SOCIAL WORLD
securing control of the land they wished to exploit, and on defending their property and stock against the attacks of the indigenous inhabitants. In varying measure, depending on the particular type of enterprise on which they were engaged, they tried to secure their labour supply, either by establishing friendly relations with the local people, or by kidnapping, capturing, recruiting or enslaving enough labourers to suit their needs. These actions sometimes led to conflict between the frontier settlers and the civil administration, as in British North America after 1754 (Collier 1947: 197). Administrative paternalism might well be combined with White rural settlement. The colonial administration was then concerned not only to maintain law and order, but also to regulate conditions of employment for the tribes and to protect them against excessive exploitation. However, the real change in the regime did not usually come about until the indigenous inhabitants of the country began to earn money in ways other than as unskilled plantation or farm labourers, and began to spend their money on things other than the limited range of Western consumer products available in the plantation store. A considerable degree of political and social autonomy was possible under a paternalistic regime; the tribal population could be left to govern itself provided the peace was not disturbed. With the growth on the one hand of industrial production with its demand for a stable urban indigenous proletariat, and for a wide market for its consumer products and, on the other, of native entrepreneurs and petty industrialists, this measure of autonomy became impossible. The indigenous peoples began to find that, in political affairs at least, they had to struggle on a national scale. In Australia, the numerical preponderance of White over Aboriginal, and the ineluctable frontier quality of most of the interior of the continent, obscured the emergence of this last phase of White conquest, which we may term industrial integration (cf. Barnes 1968a). In Papua New Guinea this stage had not been reached when political independence was granted. But it can be seen clearly enough in the condition of the Maori in New Zealand, the Bantu in South Africa and even the Hawaiians in Hawaii. This analytical scheme overlooks two factors which locally have sometimes been of great importance. In the first place, missionaries often acted independently and far in advance of administration and settlers. After the mission had gained converts, a quasi-colonial administration arose, relying mainly but by no means exclusively on supernatural sanctions but less inclined to leave well alone (cf. Gann 1958: ch. 2). Secondly, the scheme recognizes only those settlers who sought large tracts of land. In some parts of the world the invaders were mainly interested in gaining mineral wealth and, while demanding a secure labour supply, left the local population in undisturbed occupation of most of their land. But large-scale
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
109
mining usually took the administration along with it, as for example in the New Guinea highlands. Where traders, as distinct from miners and settlers, went ahead of the colonial administration, as notably in West Africa, the process of conquest usually remained in the first phase, of external relations, although in the early days, where the indigenous inhabitants were weak, some traders did settle down to establish small empires on their own account. 3 Food gatherers Let us now turn to the simplest type of indigenous political organization, as found among food-gatherers. In most parts of the world, peoples who relied on this simple technology were subject to pressure from pastoral or agricultural neighbours long before any Western administration or Western settlers appeared on the scene. They had already been forced into the less attractive portions of the environment, into the jungles or deserts or arctic wastes. These marginal areas were seldom of immediate interest to the Western invaders, and the small numbers of hunting and collecting peoples were left to live their own lives with little direct interference from outside. Thus in territories where head-taxes were levied on the agricultural population, we often find that the swamp, desert and jungle dwellers were exempted. It was usually more difficult to introduce modern facilities, such as schools, roads and dispensaries, in such areas than in closely settled regions, and because these facilities could serve only small communities, they were relatively more expensive. The environment did not lend itself to cash cropping, and the low population did not constitute a significant reservoir of labour for Western enterprises. Hence interaction with the invaders took place only on the periphery, if at all. Unfortunately small bands of nomads living at a low level of technological development and existing often on the edge of starvation were not only hard to administer but also hard to study; little detailed information is available about the way in which the Western assumption of sovereignty over the country inhabited by people of this kind affected them. Bushmen in southern Africa were, compared to their agricultural and pastoral Bantu neighbours, left very much alone, despite the seventy years or more that the various territories were under Western sovereignty. Pygmies in the Congo basin and the related peoples of the Great Rift Valley seem also to have been able to govern their own affairs while under colonial rule, unless they chose to abandon their seminomadic life style and to settle down under the wing of a mission or government post. The Siriono of Brazil studied by Holmberg (1950) were in a similar condition. The response of such peoples to Western intrusion, and indeed to approaches from other non-Western peoples with a higher
IIO
MODELS OF THE REAL SOCIAL WORLD
technology than their own, was to flee. So long as the deserts were large enough and contained some wild life and water, or the jungles deep enough, the old way of life could continue. Under these conditions, it would be surprising if any great changes in political organization took place. The density of population and the maximum size of the local food-gathering group were controlled by the environment, and were likely to decrease rather than increase under the new regime. There was no improvement in communications. Hence we would scarcely expect that centralized authority would develop, or that the size of the political unit would extend beyond the small band. So long as the people remained subsistence hunters, no large-scale organization was likely to emerge. If the administration appointed its own nominee as headman, it had to find some way of ensuring that his followers did not melt away when he endeavoured to enforce his authority over them, for there were few sanctions at his disposal other than the force that he might be able to call upon from the administration. It was difficult to send punitive expeditions into the jungle. Just as the food-gatherers did not adapt to the invasion, so the invaders did not adapt to the presence of the primitive food gatherers. The hunters merely fled and the invaders merely rejected or ignored the hunters. Thus in Taiwan the Chinese did not attempt to administer the food-gathering peoples of the interior, but were content to confine them in a demarcated area (Okada 1955: 382). In South Africa the Bushmen, who were hunters, became the hunted. Similar procedures were adopted in Australia. Elkin (1951: 165) comments ... a semi-nomadic, food-gathering, and therefore scattered people, with neither settled villages, anchored gardens, nor centralized organization, has no ... points of resistance to the newcomer and his ways, nor means of recovery. Moreover, the obvious absence of these features gives the invader (settler, administrative officer, or missionary) the impression that the natives are almost cultureless and that whatever he does can interfere but little with them. Therefore, he is very unlikely to respect native ways, customs, beliefs and values, or to adjust to these his method of economic, administrative, or spiritual invasion. He sees no objective symbols of their existence.
4 Aboriginal Australia and the Transvaal When the colonization of Australia began in 1788 conditions were unusual in that peoples living by hunting and collecting were spread over the whole of a continent. There were no indigenous pastoralists or agriculturalists who might have gradually pressed the Aborigines back into the less hospitable and fertile parts of the region. Furthermore the culture of the Aborigines differed from that of many other food-gatherers.
INDIGENOUS POLITICS A N D COLONIAL A D M I N I S T R A T I O N
III
People who live in intimate dependence on the natural environment build up an elaborate body of knowledge and lore about flora, fauna, localities and natural phenomena. The Australian Aborigines developed an unusually strong attachment to specified sites, and the policy of flight in the face of invasion was not so readily open to them. On the other hand, each Aboriginal community was linked with many distant localities by a systematized elaboration of interpersonal ties. The White invaders gradually pushed back the frontiers of settlement in much the same way as, for example, the Bantu had pushed back the Bushmen and other pygmy and pygmoid groups in Africa a few centuries earlier; but in the face of this movement Australian Aborigines did not retreat into the central deserts in the way in which, for example, the Siriono retreated deeper and deeper into the jungle in the face of advancing Brazilian settlement (Holmberg 1950: 62-3). The European invaders of Australia were not peasants and did not settle the country closely. Furthermore they were not subsistence cultivators or subsistence pastoralists, but were producing for a market and were able to employ labourers. In most parts of Australia the settlers were well in advance of the administration. The stage was thus set for the partial elimination of the Aboriginal population, and its partial conversion to a peon class attached to White pastoral properties. The same process seems to have taken place in the Transvaal and in parts of Natal at about the same time. In southern Africa the Bantu who came to live permanently on European farms were traditionally cattlekeepers and maize growers, not hunters, and hence perhaps fitted more efficiently into the sheep and cereal economy of their White overlords. They did not have the same tradition and recurrent need to go off on long journeys that characterized Aborigines in Australia. Politically the Bantu of the Transvaal had a tradition of centralized authority infissilestates. But when the Boer and British settlers invaded the high veld and the uplands of Natal, the indigenous population consisted for the most part of splinter groups of refugees who had been harried by the empire-building activities of the Zulu, Swazi and Southern Sotho. The White farmers offered protection of a kind to people accustomed to authority at a time when their own leadership was proving inadequate. In both countries the administration eventually caught up and even overtook the settlers on the frontier. The administration asserted its monopoly to dispense justice and laid down conditions of employment. In neither country did anything substantial remain of indigenous political organization, whether traditional or modern, despite the persistence of major cultural distinctions. In both countries the native population became dependent on the White economy principally as labourers, in Australia as casual labourers or skilled cattle hands, in South Africa
112
MODELS OF THE REAL SOCIAL WORLD
as farm labourers or mine workers. Sovereignty passed to the White government. 5 Primitive states When indigenous people lived in a well-ordered state with powerful centralized authority, it was economical for the conqueror to govern through the native ruler. It is a mistake to imagine that this technique was created by Lugard in the 1900s, even if he invented the term Indirect Rule. The technique goes back to the Romans, if not earlier. The principles of the technique are clearly expressed in a memorandum written by Captain Thurston, a British colonial administrator in Uganda, in 1897. He says It has always been the practice of England to govern her distant dominions, as apart from her Colonies, whenever feasible by the system of Protectorates; by which system their administrators are placed under a native Prince who governs by the advice of a native Protector. The advantages are obvious; for the people through force of habit, love for the person, or the prestige of his office, naturally submit to the orders of their Prince. The Prince himself through the instinct of selfpreservation if through nothing else, usually willingly obeys the orders of his protector, and those orders are further disguised under the name of advice, and are conveyed in such manner as to as little as possible destroy his prestige or wound his susceptibilities. By this means pressure when it is necessary is brought to bear on one person only, the Prince, and not on the whole population. Even when the Prince withholds his ready cooperation from the protector, the cases of Egypt, of Zanzibar and of Uganda tend to prove that the system can still be employed with a full measure of success, (cited in Low 1973: 14-15)
Most of the studies by social anthropologists of the political effects of conquest relate to primitive states. In describing decentralized societies, whether stateless societies or the political structures of food-gatherers, it is easier to ignore the results of conquest. Furthermore almost all these studies have been concerned with the same phase of the process of conquest, that of paternalism (Barnes 1967J; Busia 1951; Fallers 1956; Gluckman 1940, 1951; Hunter 1936; Kuper 1947; Mair 1934; Mitchell 1956; Schapera 1956). We can state briefly the main features reported in these studies. It is only with primitive states that a clearly demarcated phase of peaceful external relations occurred, forming a prelude to the introduction of administrative paternalism. There had to be an effective indigenous ruler with whom the Western power could make a treaty of nonaggression, protection or trade. Conquest was often achieved when the invaders made an alliance with a legitimate but not dominant faction in the indigenous state; with Western help the faction then achieved dominance. An indigenous ruler might be successful in utilizing Western
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
113
assistance to conquer his neighbours. Resistance to invading pressure often promoted increased concentration of power. Once incorporated into the colonial administrative framework, the indigenous political structure, or rather that portion of it utilized by the conquerors, tended to become ossified. Territorial groups were utilized and non-localized social groups ignored. Political units that formerly varied continually in size and status were frozen or underwent discontinuous changes. New officials, particularly clerks, rose to power. Chiefs who formerly were highly competitive specialist politicians relying on popular support found themselves appointed to relatively secure and non-competitive administrative positions; they depended on the support of the invaders, and tended to lose touch with their people. Whereas in the past the chief legitimately favoured his kinsfolk, in the new dispensation to favour a relative was nepotism and therefore wrong. Where cash cropping developed, chiefs and other aristocrats might be able to utilize their traditional prerogatives to become much richer than their subjects; where labour migration was the main source of money, chiefs who had to stay at home might find themselves poorer than returned migrants. Dissatisfied subjects could not flee to another chiefdom as they might have done in the past, and it became necessary to set up some kind of council or assembly where popular dissatisfaction could find an outlet. Where some traditional assembly was utilized, membership was usually more sharply defined and more restricted than in the past. The new social order lacked the sanctions of a traditional religion, and Christian beliefs were rarely presented in such a way as to support the position of local rulers, new or traditional. The doctrine of the divine right of kings was not for export. 6 Stateless societies The characteristics of primitive states under a regime of colonial paternalism are not our main topic. I am more concerned with food-gathering peoples and with stateless societies. However, in several parts of the world colonial administrations assumed that all native peoples under their charge were organized in states, and sometimes in states of a specified kind. In any colonial territory administered as a unit it was convenient to have a uniform system of native administration that could be applied by field officers wherever they were stationed and which could easily be adapted to territory-wide action. An indigenous model, if it appealed to the administration, might be used even in areas where the local indigenous political system was of quite another kind. Thus in Uganda, where the British were impressed with the efficiency of the Ganda state system, Ganda models were used for the administration of other parts of the Protectorate. In some of these areas, the former political system had indeed been similar to that of
114
MODELS OF THE REAL SOCIAL WORLD
the Ganda, but in other parts there had been stateless societies of various kinds. To make the conversion more thorough, the British appointed Ganda notables as paramount chiefs of these non-Ganda areas. Another stateless society that was treated as a state was the Plateau Tonga, an egalitarian matrilineal stateless people of Northern Rhodesia (now Zambia) (Colson 1948, 1962). Their territory was bisected by a railway line built a few years after they had been brought under British control at the beginning of the twentieth century. Land on either side of the railway line was alienated to White settlers, who began to raise cattle, and to grow maize and tobacco, employing labourers from the Tonga and more distant tribes. In the early stages of contact the British ignored the Tonga, for they were considered to be tributaries of the Barotse empire, with which the British negotiated a treaty which, inter alia, transferred rights in Tonga country to the British in return for protection and payments to the Lozi, the Barotse rulers. Later, Tonga villages, which had been fairly fluid residential units, were made the basis of administration. Villages were grouped into small districts and a leader appointed to each district. Colson considers that the district leader was probably the indigenous ritual leader. In 1918 chiefs were appointed over larger areas. Some of the men selected as chiefs were ritual leaders or their descendants, others were rainmakers or leaders of rain shrine cults. A genuine effort seems to have been made by the administration to appoint those who might have a traditional claim to chiefly office, for it seems not to have realized that indigenously there were no chiefs. Those appointed to office were succeeded by their heirs, for the administration assumed that the position had been, or ought to be, hereditary. In this way small groups of matrilineal kin obtained a monopoly of office which had power over the whole chieftaincy. The government did not intervene to alter the rules of succession it had itself introduced and chiefly matrilineages began to emerge in what, fifty years earlier, had been an egalitarian society. One chief was given higher status than the others. The people paid no particular respect to the chiefs, whom they called government chiefs. One reason for acquiescence in this state of affairs was that the proximity of the railway line made it possible for industrious and ambitious individuals to become relatively prosperous. Hogbin has shown how difficult it could be for an administrator to realize that one of his native subordinates was exceeding his powers, and how hard it was to take effective action even when the administrator did realize what was happening. Indeed, in a stateless society, it was almost impossible for a man appointed as headman or chief to carry out his duties efficiently in the eyes of the administration without at the same time exceeding his powers. He had the force of the administration behind him, but he had to assert his authority over his people and demonstrate to them
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
115
that he must be obeyed; they did not automatically do what he said. His own definition of his role was unlikely to coincide exactly with the definition intended by the administration, and there was no traditional model for him to follow. Hence he was likely to do either too little, and to be acceptable to the people but inefficient in the eyes of the administration, or too much and become a tyrant. Since his status was unprecedented, there were usually no traditional mechanisms for limiting his power. Since he was a nominee of the administration, he was bound to get its support, and often there was no effective way in which the common people could bypass the local autocrat and complain directly to the administrator. Those appointed to recognized posts might have some traditional claim to office, even if it was to an office lacking secular authority, as with the rain shrine priests who became chiefs among the Plateau Tonga. Sometimes they lacked any traditional legitimacy at all. Hogbin has reported how in Malaita, where traditionally the leaders of district groups were men who owed their position not to descent but to their success in organizing feasts, the administration appointed headmen who seem not to have been traditional leaders but were men who had some experience of the constabulary run by the administration. Maekali, one of Hogbin's main informants, had served in the Malaita constabulary for twelve years before being appointed headman of a territorial division. In 1940 the administration introduced a scheme of Native Councils and Native Courts. The councillors selected were in many cases the traditional district group leaders but Maekali remained as president of the Native Court and acted as chairman of the Council (Hogbin 1939: 143, 1944: 262).
Whatever the traditional form of political organization, the local representative of the colonial administration had to employ some members of the indigenous population full time in order to carry on the day-to-day business of administration. They might be employed as policemen, porters, clerks or even as administrative officers. No system of administration seems to have been able to operate without this buffer of local employees. It would have been impossibly expensive to operate with only expatriate labour, whether or not it was thought to be in the best interests of policy to do so. The group of indigenous public servants, if they may be so called, tended to identify themselves with their employers and to adopt in many respects the values and attitudes of the administration. They might be able to exercise fairly complete control over the activities of their superiors, particularly if the expatriate administrator could not speak the local vernacular (cf. Mitchell 1949: 153-4). The indigenous public servants were often able to exercise considerable power even when they were dealing with traditional primitive states. With stateless societies, there were usually fewer checks against the abuse of authority imposed from
Il6
MODELS OF THE REAL SOCIAL WORLD
above, and their positions were therefore likely to be even stronger. These same considerations applied to other individuals who gained the support of the indigenous members of the public service. 7 Australian Papua New Guinea In terms of this analysis, the chief characteristics of Papua New Guinea while under Australian colonial rule were that the indigenous political systems were very largely stateless, and that with some exceptions the administration was on the frontier, not the settlers or the missions. As in Aboriginal Australia the traditional systems did not provide hereditary princes and rulers who might have become the channels for implementing the policies of the administration. The amount of force that could be deployed was much greater than in the conquest of Africa in the nineteenth century, where the colonizers were often operating months away from their bases, threatened by disease, with very vulnerable communications and difficulties of supply that kept their forces small. Hence in Africa it was necessary for survival and success to find local allies. These conditions did not apply so forcefully in Papua New Guinea, despite the very rugged terrain and the insularity and hostility of the inhabitants. If they had, the conquest of Papua New Guinea would have been even more difficult than it was, for stateless peoples make poor allies. Even stateless peoples had leaders; but their authority was usually transient, and might be restricted to one domain of social life. When an hereditary chief became the nominee of the administration, it sometimes became an embarrassment to have to accept his son as the successor to an administratively recognized position. Traditional leadership, particularly if it was largely ceremonial, might be too determinate for the purpose of the administration. On the other hand, with an erstwhile stateless society, the problem facing the administration might be how to give permanence to a leader who might have achieved his outstanding position in a traditionally acceptable manner but who was likely to be challenged in a few years' time. If, as in much of Papua New Guinea, a leader owed his popular support to his success in giving feasts or in arranging ceremonial exchanges, how could he continue to enjoy public confidence when he switched his attention to such activities as seeing that pigs were not allowed to wander at will, or that paths were kept clear, or that ill people were taken to the dispensary for treatment? If there were two conflicting systems of evaluation of his performance, to what extent could he continue to enjoy simultaneously the admiration of the administration and the people? In particular, in those societies where prowess in warfare was one of the most important criteria for positions of leadership, how could the
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
117
former war leader become the upholder of the pax australiana and eschew indigenous warfare? Unfortunately we do not know the answers to these questions, and indeed probably will never know. The questions are directed at a model of latter-day colonialism which, if African experience is any guide, would have to have run for a couple of generations before possibly reaching a relatively steady state. But, as so often happens, a model that seems to tally well with the facts at one point of time, and which is ready to run indefinitely, is soon made irrelevant by new and usually unforeseen developments and has to be replaced by a new model. In Papua New Guinea, long before anything like an equilibrated colonialism had been reached, political independence was achieved, more because of external factors than because of internal pressures. With political independence the colonial model ceased to be of any help in understanding the political process in Papua New Guinea, and a new model, incorporating many more variables than we have mentioned here, was called for. 8 Industrial integration In a fully integrated society, the traditional forms of indigenous political organization cease to be relevant, though they may well survive in transformed symbolic guise. Groups of people can continue to manage their own affairs only so long as their affairs are mainly their own. Once people begin to participate effectively in the economic and social life of a wider society, they cannot be kept out of its political life. Even if the indigenous population remains a sharply demarcated unit in the social structure, as for example with the Black people of South Africa, political activity by indigenous men and women forms part of the political life of the nation. If power is concentrated at the centre, political activity, whether legal or illegal, parliamentary or extra-parliamentary, must be aimed, either directly or at several removes, at achieving control at the centre. Whether or not there is continuity in leadership from the old to the new regime is largely a matter of chance. It has been argued that stateless peoples can adapt to major political changes of this kind more easily than those who have belonged to primitive states; new leaders, trade union presidents, bishops, priests and congress chairmen can emerge without exciting opposition from the older hierarchy of paramount chiefs, clan leaders and the like. Yet aristocrats of the old regime do sometimes achieve positions of leadership in the new order, as for example the Hawaiian prince who was one of the first delegates from the Territory of Hawaii to the United States House of Representatives; he went as a Republican. When incorporation takes place without assimilation, as for example in South Africa, or with the Jews in the ghettos of mediaeval Europe, some
Il8
MODELS OF THE REAL SOCIAL WORLD
limited autonomy is possible and the cultural traditions of the enclave are one factor in determining the form that the internal organization takes. But even in this limited sense, indigenous political institutions depend on recognition by the wider society. Furthermore, in external affairs, the group and its leaders have to contend with the larger political system. 9 Politics and administration In this paper I have tried to outline some of the principal features of the process of conquest as affecting indigenous political systems of different types. On the way I have tried to fit into the scheme of analysis Aboriginal Australia and colonial Papua New Guinea. The examples I have used have been selected casually, but I hope they may suggest parallels and divergencies elsewhere. One important distinction to make, in discussing forms of political structure, is that between administration and politics. M. G. Smith makes the distinction between what he regards as two aspects of the wider category of government. He defines action as political when it seeks to influence the decision of policy. Policy decisions define a programme of action, implicitly or otherwise. The execution and organization of this programme is an administrative process. (Smith 1956:48) Administrative action, according to Smith's usage, consists In the authorized processes of organization and management of the affairs of a given unit'. Whereas politics is a struggle for power, administrative action is 'defined by authority, and is inherently hierarchic' (Smith 1956: 49). In a modern state we make a sharp distinction between professional politicians and administrators and we do not expect, and do not find, the same qualities in the two categories. In the same way we make a distinction between political activity and administrative activity. Under almost any scheme of colonial administration, the recognized chiefs, headmen and councillors were intended to be mainly, if not wholly, administrators, like their superiors in the hierarchy. In the traditional society prior to contact, positions of leadership called for the exercise of political rather than administrative skill. In those well-developed primitive states with a distinctive administrative and judicial machinery, as for instance in the Nupe and other West African kingdoms, the junior positions were mainly administrative; but even here the chiefs acted, and were expected to act, as political leaders. Official reports contain many instances where a chief or leader showed that, however courageous a warrior and adroit a politician he may have been, he made an ineffective subordinate administrator. But the difficulty of adapting a political elite to
INDIGENOUS POLITICS AND COLONIAL ADMINISTRATION
119
administrative tasks is not fundamentally a personal one. The difficulty lies rather in the method of recruitment. When leaders were selected because they were diligent in acquiring esoteric knowledge, as in Aboriginal Australia, or because of their outstanding managerial skills, as in the highlands of Papua New Guinea, it is not surprising that few of them were efficient or cooperative administrators. Furthermore, loss of sovereignty did not remove the need for political leadership. In part the invaders supplied, or endeavoured to supply, all the leadership that was thought to be necessary. They debated with one another about how 'their' indigenous people should be protected, exploited or developed. The task of implementing the policies decided upon was left to the administrative hierarchy, from the highest colonial public servant down to the lowest village headman. In a fully integrated society, this division of function might have worked, but it could scarcely be expected to provide effective leadership when the mass of the people did not participate, even as spectators, in the policy-making process. Hence we had on the one hand a perennial crop of chiefs, headmen and others who became inefficient administrators because they endeavoured to play politics; and on the other the development of native councils and similar institutions of local government, in which the indigenous people could argue with one another about alternative courses of action or inaction. The effectiveness of these new institutions depended in part on the extent to which they were truly political institutions, concerned with the distribution of power, and not mere debating societies. A native council that always acquiesced in the policies of the colonial administration could not be an effective political instrument. One of the main practical problems facing any plural society composed of conquerors and conquered was to adjust to the unequal distribution of power and sovereignty. It had to ensure that despite differences of culture that divided its people the struggle for power could be carried on in political institutions that were effective and which enjoyed the support of all sections of the community. The penalty for failure was more 'one-sided' wars.
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY: A STUDY OF AN NGONI GROUP IN NORTHERN RHODESIA
This chapter, like the previous one, deals with a social field marked by heterogeneity, by the imposition of colonial rule over indigenous peoples, but the focus of attention shifts from political control to perceptions of the past and the implications of the introduction of writing. While writing the paper in 1950 I began to think about our concept of history, so beginning a line of discussion carried further some twenty years later in my paper on Levi-Strauss (Chapter 9). The paper is based on observations made in thefieldin central Africa between 1946 and 1949. Writing so soon after I had left thefieldI described what I had seen in the present tense; it would have seemed excessively pedantic to have written about customary practice of such recent date in the past tense. Now, nearly forty years later, the present tense has become inappropriate in the light of the major changes that have occurred during the intervening period. I have therefore changed the tenses as necessary, partly to emphasize that the situation I describe no longer holds, and partly to avoid any suggestion that I am writing in what anthropologists call 'the ethnographic present' tense, a literary device intended to detach tribal societies from their historical context so that they could be studied as museum specimens, uncontaminated by Western influences. This is precisely what I do not want to do. In the paper I refer to the people among whom I worked as the Fort Jameson Ngoni. The township of Fort Jameson is now known as Chipata but I have left the name unchanged to emphasize that my account relates to the colonial period and not to the present independent I2O
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY
121
regime. An interesting essay throwing much light on how in the 1980s Ngoni men and women of Chipata district perceive their past has recently been published by Lukhero (1985).
In non-literate societies there is, strictly speaking, no history; for there are no documents, and history, in the sense of historiography, is essentially the study of contemporary documents. There are only legends by which, to a greater or lesser degree, present conditions are related to or explained by alleged former conditions (cf. Evans-Pritchard 1950: 121). In fully literate societies, with a well-documented past, legends take different forms and it is then often difficult to distinguish legend from history. In modern societies many socially significant legends relate to the immediate past rather than to antiquity. It is of interest, therefore, to examine what kinds of legends are current in societies shifting from non-literacy to literacy and to what extent they have a history. In many Western cultures the earliest surviving documents are chronicles or codes of law. At an early stage writing was employed to record facts of wide interest rather than things of significance to only a few. When Africans in colonial societiesfirstbecame literate they employed their skill in writing letters to one another. Only later did individuals begin to write essays, histories and novels and so develop a vernacular literature. They learned to write in an environment already containing a great quantity of vernacular printed matter issuing from European-controlled sources. The spatial separation of kinsfolk brought about by labour migration made letter-writing worth while and the ephemeral nature of these pencilwritten documents was no disadvantage. It required greater resources than most individuals and Native Authorities could muster to produce a family or tribal chronicle in permanent form. In the 1940s the Fort Jameson Ngoni were at this stage of transition. Many men and some women could read, if not write, and we could no longer call the tribe non-literate. Yet a society of this kind behaves in many ways as if it were non-literate. In particular, information about the past is transmitted orally and laws and customs are not committed to writing. In this paper I discuss some of the ways in which this information was manipulated. I distinguish between tales of fighting, related to the period 1821-1898, which we may regard as tribal legends of origin (Barnes 1967J); beliefs about customary behaviour relating to the same period; private legends which served to explain the form of relations between present-day groups; and lastly history proper, based on documents describing contemporary events. Early travellers all speak of the various Ngoni groups as fierce warrior tribes interested only in destruction and plunder. With the spread of European control Ngoni raids ceased. In villages inhabited mainly by
122
MODELS OF THE REAL SOCIAL WORLD
women and children, with scraggy animals and eroded gardens, the modern traveller who looked for signs of a military tradition would see only khaki greatcoats and battle-dress. Among the Fort Jameson group, if not elsewhere, the great cowhide shields and the regiments which carried them disappeared at the beginning of the century. Yet the battles of the past were still remembered and were called to mind in songs sung at drunken parties, in tales told to children and, importunately, to sympathetic strangers. Many of these songs were in the old Ngoni language already largely forgotten, and singers could not give precise meaning to the songs they sung. They were relics of the past learnt parrot-fashion and, in the context in which they were sung, they did not require any specific reference. In our own culture we do not need to know who were the April rainers before we can sing 'Green grow the rushes, 0'. Prose tales, however, were in ordinary speech and described military exploits of people known to the speaker, or preparations for raiding ordered by the Paramount Chief and carried out by his lieutenants. The most frequently recounted battle was the fight between Ngoni and British troops in 1898. Many men and women were still alive who remembered the fighting or even took part in it. The cataclysmic changes that followed the Ngoni defeat and the unusual weapons used combined to make the battle memorable. It was the only admitted defeat, for all other tales told only of victory. The Ngoni defeated the Bemba, Wisa, Cewa, Kunda and so on. We know from contemporary travellers' accounts that though on the whole the Ngoni were successful against their neighbours, their run of successes was not unbroken. There is evidence that on occasion even the Cewa raided the Ngoni and took captives. We may regard the form taken by these legends as due in part to the process found in most cultures by which victories are remembered and defeats forgotten. But the distortion introduced was not only the elimination of defeat. In general, the Administration and the Missions encouraged cooperation between tribes and treated all tribes as equal. They emphasized the values of peace rather than of war, and condemned pagan rites. In the light of this teaching legends were modified to conform to modern values. The barbarity of the raids was minimized and the good qualities of the old military discipline were extolled. Sometimes men said that the old armies never fought but only danced, or that people from other tribes were not captured but all joined the Ngoni of their own free will (as in fact some did). It was said that there was never any ancestor worship; the names of the Paramount Chief's ancestors were mentioned in prayers merely 'as King George is mentioned now'. This process by which the Ngoni conquerors of yesterday were made to look like the European conquerors of today was no new phenomenon. The Ngoni identified themselves with Europeans rather than with their matrilineal neighbours even before the European conquest.
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY
123
Wiese (1900: 200-1), himself the main agent of their defeat, wrote 'They are very proud, looking down on all the other tribes with contempt, and think they show great courtesy whenever they assure the European traveller that the Whites may be related to them'. Another traveller, arriving from the south in 1896, related how he ran into an armed patrol on the outskirts of Ngoni country. His carriers ran away but he called out to the patrol that he was a friend of the Paramount Chief, speaking in Zulu. He then heard them saying 'He is an Ngoni and a white man'. Later, when he had satisfactorily answered questions on Zululand geography, the Paramount Chief said of him 'He is a chief... he is an Angoni, one of us' (Durban man 1929: 19, 25). My own informants told me that the old Ngoni language was closer to English than to Nyanja. This reinterpretation of the past in the light of the present did not produce a consistent picture of a proto-European society. The inconsistencies of the present remained and were reflected in the ambivalent attitude adopted towards the past. Although in one context people affirmed that everything was well ordered and peaceful in the past, in other contexts they dwelt on the unlimited power of the Paramount Chief and the fierceness of his warriors. An aristocratic man once said to me at a beer party, If this was the old days I would kill all these Cewa drinking here'. Yet the general impression people usually tried to convey was of a disciplined, proud and successful people. This picture was accepted not only by Ngoni themselves but also by the European neighbours of the Ngoni. For them it was one feature of the good old days when Africans still smiled and were nature's gentlemen. Their picture of the Ngoni past contained fewer concessions to contemporary White values, for the very distinction between Whites now and Blacks then increased the attractiveness of the picture. Ancestor worship and polygyny were not unwelcome in a Black Arcadia. In the 1940s Ngoni compared themselves with Europeans and Indians living nearby and envied their apparent power and wealth. They sought to rationalize their own relative failure in terms of their past. It was in this connexion that the defeat of 1898 was important. Although they claimed to have conquered all their African neighbours, Ngoni did not claim success over the British. Instead, defeat at the hands of the British was cited in explanation of every present-day failure and of every departure from traditional practice. Ngoni said, If we had not been defeated we would not have lost our land and we would not now go hungry'; If we had not been defeated the tribes would not have been mixed up and we could follow our customs properly' and so on. In the past, captives taken by the Ngoni were rapidly indoctrinated into Ngoni ways and themselves became proud Ngoni; Ngoni complained that, although conquered by Europeans, they had never become Europeans.
124
MODELS OF THE REAL SOCIAL WORLD
The general legends of the past from which these notions were derived were common knowledge. In the main they were passed on orally and were maintained by the teaching of tribal history in schools, the appeals to old military traditions made by recruiters for the Northern Rhodesia Regiment, the conversations of European planters comparing one tribe with another as labourers, the reports of administrators and the writings of ethonographers. The link between the latter-day Ngoni and their illustrious forebears was also symbolized in the office of the Paramount Chief. He bore the name of his warrior great-grandfather, and the formal aspects of his relation to his people were explained by reference to his ancestors' achievements. He could not attend funerals or take part in some of his own marriage ceremonies because 'he had already conquered all others'. He could not be expected to walk through the bush with the District Officer because 'he had already conquered the whole world'. He basked in his great-grandfather's invincibility, for in this context the defeat of 1898 was forgotten. The relations between the Paramount Chief and his eleven minor chiefs were determined partly by the distribution of power among different branches of the royal lineage before the European conquest, and the division of the country into these minor chiefdoms was a visible reminder of that distribution. Associated with these tales of military exploits and royal power were myths about former customs. Ngoni said that in the old days they used to hunt, and marry, and build villages, in a distinctive way and that now they did so differently. Frequently when I asked what would happen in suchand-such a hypothetical situation I would be told what my informant thought would have happened before 1898 rather than what he had seen happen recently. Thus if I asked what payments did Ngoni make in marriage I was usually told that there were two payments, 'snuff-box' and bride wealth, consisting of one or more beasts. Further inquiry revealed that these payments were nowadays made only infrequently and that the period my informant had in mind was before 1898. More detailed investigation suggested that in many pre-1898 marriages these two payments were not made and that when bridewealth was paid it did not always consist of beasts (Barnes 195 if). Ngoni did not deny these facts but they did deny their relevance. To them the distinctively Ngoni way of marrying was with these two payments, and it was this process alone that merited the vernacular adjective cingoni. They regretted that people no longer married in this way and blamed the Europeans for it; but this form of marriage remained part of the distinctive cultural heritage of the tribe. As we might expect, notions of this kind did not provide sanctioned standards of conduct nor ideal patterns towards which Ngoni might strive. They might be used to explain or justify a state of affairs, as we have seen,
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY
125
but they did not directly determine action. An individual who departed from traditional custom, in these terms, was not necessarily censured. Indeed, men who were knowledgeable about the glorious past did not follow traditional ways any more than did less well-informed people. According to tradition the Ngoni were a virilocal people, yet if a man went to live with his wife's relatives his own kinsfolk could not force him to return by appealing to this virilocal tradition. Traditionally a man should avoid his mother-in-law. Yet I once saw an aristocratic man, a great authority on traditional history, on leave from work sitting on a verandah chatting with his mother-in-law. I said to him, 'I thought that a man should avoid his mother-in-law'. He laughed and replied, 'Yes, so he should. This behaviour is disgusting, and as soon as I come to live here permanently I shall see that customs are properly followed'. In certain situations the Fort Jameson Ngoni appeared as a tribe in contrast to other tribes. Tribes were administrative units, each with its own chiefs, its own area and sometimes its own District Officer. Representatives from each tribe in the region sat together to form the Native Urban Court and the Provincial Advisory Council. The Ngoni representatives on these bodies tended to present distinctively Ngoni views to contrast with the views of their colleagues. Thus there might be a discussion on wills. There would then be an Ngoni view, a Cewa view, and so on. The Ngoni representative would repeat the old story about bridewealth, virilocal residence, the rights of the eldest son, and these would be recorded as being the Ngoni custom in the matter. Even when discussing an apparently modern issue such as wills, the tendency was to go back to the mythical past. It was in terms of this past, in reality as well as in myth, that Ngoni were different from their neighbours. In the present, as contrasted with Europeans, Ngoni and Cewa were much the same. Ngoni paid bridewealth in the same way as the English eat roast beef. It was a national characteristic, but one located in the minds of men rather than observable in their actions or their aspirations. Instead of the scarifications and totemic observances that distinguished their neighbours, Ngoni expressed their tribal identity by having distinctive customs which, conveniently, need not be followed. In the Native Urban Court a similar process appeared to operate, but here, unlike at the Advisory Council, there was a greater need to come to grips with real situations that required action, and tribal distinctiveness was tempered with practical reality. In the purely Ngoni Native Courts, however, appeals were made to traditional practice in a different way. The courts were required by the Administration to follow tribal custom and people knew that the courts were bound to do so. In fact the courts had continually to deal with new situations and to make decisions that were unprecedented. This was done under the guise of drawing attention to
126
MODELS OF THE REAL SOCIAL WORLD
some good Ngoni custom that had been neglected. For example, a man came to court saying that he was always quarrelling with his wife and that he wished to divorce her. The bench granted the divorce and awarded the woman 30s damages. The litigant protested. The junior member of the bench, a man aged about twenty-five years, said 'Don't you know, it has always been the custom in this court to award 30s damages against men who divorce their wives?' Yet this was a comparatively recent practice and the litigant's protest seemed to me to be quite justified. The young man had been on the bench for only about eighteen months. Even without Administrative stimulus, deliberate acts of legislation were not unknown among the Ngoni, but these required considerable discussion and probably a tribal meeting. It was easier, particularly for a court member as junior as the man in question, to appeal to the unwritten corpus of tribal custom when introducing a new rule. An analogous process in our own society is perhaps concealed under the fiction that judges do not make law. Ngoni did not quote specific precedents in court and in this undocumented environment new decisions, if they were not challenged, soon became part of what had always been the custom since time immemorial. We may then summarize by saying that their common store of notions about the past provided Ngoni with a means of distinguishing themselves from similar groups at a time when the cultural distinctions between these groups were in fact becoming less and less. In the courts, appeals to the value of traditional custom were made to gain acceptance for new rulings, while in ordinary life breaches of custom were widespread and were not censured. When we consider legends associated with particular groups within the tribe the picture is rather different. Separate groups in the population remembered, and sometimes acted in the light of, particular incidents in the past which linked them to other similar groups. Thus village A might be linked with village B because of some incident in the past when the erstwhile leader of A captured the leader of B, or because at one time the leaders of A and B were brothers. These incidents, and the present-day inter-group relations with which they were associated, were not known universally but only by the groups concerned and by those near them. Yet in many ways these incidents were of greater practical significance than the more widely known tribal history. The relations to which they gave rise, or rather such of these that had persisted to the present time, were remembered at funerals, at the formal exchanges of beer, and in the settlement of quarrels. There were at first glance no sanctions to maintain relations of this kind, and indeed the very existence of two villages A and B might be due to a quarrel followed by a split. But although, for example, when an old woman died in A there was no way of directly forcing a messenger to carry the news to B, there was in fact a good deal of social
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY
127
pressure on both groups to keep alive the connexion between them. This arose from the general responsibility placed on everyone to preserve the peace, that is, to preserve the existing social order intact. If there were quarrels in a village the headman was blamed by the chief for not controlling his people well. All disputes were best settled out of court, even if they might have to be referred to court later on. Therefore a quarrel in village A brought to light all the alliances and cleavages which linked A with other villages and differentiated it from them. The headman of A would appeal to B for support, invoking the historical incident linking them, and would hope to settle affairs of his village with the authority of B's approval behind him. For if the headman of B was a potential ally he was also a potential rival, and the dissident section of A might invoke its link with B and transfer its allegiance to him. Thus although the details of the relation between A and B, and the historical incident connected with it, might not be widely known there was considerable external pressure on them to maintain their relation. Connexions of this kind, mainly between villages, but also between chiefdoms, played a large part in contemporary Ngoni life. Men walked twenty miles to a funeral and women carried pots of beer all day in response to ties of this kind. Thus we see, on the one hand, that tribal legends were widely known, related to pre-conquest times, and did not directly influence conduct; and, on the other, that legends associated with inter-group relations were known only by the groups concerned, might relate to either before or after 1898, and did directly influence action. Why was there this difference? We may regard the former legends as of significance mainly in external relations and the latter in internal. Within the tribe, and more particularly within the minor chiefdoms, inter-group relations were still largely determined by Ngoni themselves, however much they might have altered in form and content since 1898 and however much Ngoni had become part of a wider social system. Movement between villages was only slightly controlled by the Administration and in quarrelling, in marriage and adultery, beer drinking and hoeing, Ngoni men and women chose their own allies and enemies. Incidents in reality gave rise to inter-group relations and were subsequently invoked to maintain them. History was preserved not in books but in the names of groups and in the remembered connexions between them, and as these connexions gradually changed so local legendary history was gradually revised. After 1898 in external relations the Ngoni never acted as a tribe except within the framework of European-controlled institutions. No myth of origin was needed to explain the fact that Ngoni and Cewa representatives sat together on the Provincial Council save that the Administration had put them there. The world of external relations was one of cultural heterogeneity and comparatively adequate documentation. In this field
128
MODELS OF THE REAL SOCIAL WORLD
legends had only a limited social function. They differentiated tribe X from tribe Y but did not relate them to each other. Lastly we must consider Ngoni history proper. The many alleged accounts of Ngoni history published in Europe and South Africa seemed to have had little influence on the Fort Jameson Ngoni. Several histories had appeared in the vernacular, and tribal history had become a subject taught in the local primary schools. All these vernacular accounts were based more on the recollections of old Ngoni men and women than on contemporary documents, of which there were very few. We must therefore regard them as written legends rather than as proper histories. Differences between one version and another sometimes reflected the different political affiliations of those providing the information. Read (1936: 467) and Winterbottom (1950: 17) point out that informants from northern Nyasaland (now Malawi) asserted that their Paramount Chief was the heir of Zwangendaba, the leader of the Ngoni on their trek from southern Africa, while those from Fort Jameson said that their Chief was the heir. The period from 1821, the start of the trek, to 1898 was treated in detail in the vernacular histories and the remaining fifty years dismissed in a few lines. It seems likely therefore that in time the tribal legend will be ossified, or at least that there will be one version widely known. The vernacular texts were, however, written in the school language, Nyanja, and not in Nsenga, the language of ordinary village speech, and were learnt largely parrot-fashion, sometimes even with unintelligible words included. It is uncertain how much of what was learnt under these conditions would be carried over into life outside the classroom. We might well hope that with the increasing documentation of modern life the period of legend would end and Ngoni history proper would begin. Unfortunately this is not the case. It is fairly easy to find out from contemporary documents what went on in tribal affairs from 1890 to 1898, but with the imposition of British rule a curtain descended. Official reports grew shorter and shorter with the years, for there was less and less to record. Towards the end of the British South Africa Company's rule the published information on Ngoni life was little more than the annual statement, covering the whole of Northern Rhodesia, that 'the conduct of the natives has been satisfactory' (cf. Barnes 1967J: 116). From the beginning of Colonial Office rule in 1924 reports lengthened but these dealt with the history not of the Ngoni tribe but of the Eastern Province and of Northern Rhodesia. These larger regions became the significant units of government, not individual tribes and chiefdoms. The Fort Jameson Ngoni became more and more closely integrated with the wider society and less and less had any history of their own. When the Ngoni were making history they lacked the means to record it; when the means became available they had no history to record.
THE PERCEPTION OF HISTORY IN A PLURAL SOCIETY
129
Thus we see that in this transitional society internal and external relations, as well as the level of technical skill, influenced both the way in which people related themselves to their past and also the timing and content of their written history.
Part II
MODELS OF SOCIAL SCIENCE
8
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
The chapters in the first part of this book are focused on various features of the world - relations of kinship, lineage, social class, political party, and of conqueror and conquered - all of which are to be encountered mainly away from libraries and university campuses. In the second part we turn to examine an activity that typically is pursued on the campus, the study of social science. Relations of kinship and the like are of course to be found on university campuses, and likewise the university does not have a monopoly of the study or practice of social science. Nevertheless, there is a valid contrast between the largely non-intellectual content of the relations which are dealt with by the models discussed in the first part and the mainly intellectual activities considered in the chapters of the second part. A common theme runs through the second part, the contrast between what I call the natural science paradigm (cf. Barnes 1980a: 25-35), a n idealized model of the procedures of inquiry and analysis typically followed in the natural sciences, and the modes of inquiry and analysis appropriate to the social sciences. This contrast has a long history, notably including the Geisteswissenschaft controversy during the latter half of the nineteenth century, a history that I largely ignore. The essays presented here were all written in response to problems I encountered in my own work, problems that arose because of my wish to apply, as far as possible, the methods of the natural sciences to the study of social institutions and relations. I was trained initially as a mathematician and later found myself working in a wartime radar research establishment. I remember being 133
134
MODELS OF SOCIAL SCIENCE quizzed about what I was going to do when I was demobilized. I replied that I would try to apply to the study of social phenomena the same principles of investigation as I had used on material phenomena during the war. In a sense the following chapters document my progress (or if that word is too value-laden, my progression) from that naive statement towards a position that takes adequate account of the inherent differences between the two branches of science. The first of these chapters begins with some general comments on these differences and continues with a discussion of the significance of time in the study of social life, a topic that I return to in Chapter 9. The paper was first given at the annual meeting of the Sociological Association of Australia and New Zealand in 1967.
I thought that it might be of interest on this occasion to comment on a broad topic that I have been thinking about for many years: what meaning can we give to the claim that sociology is a social science? More generally, what does the word 'science' mean when we modify it with the adjective 'social'? I have touched on this question elsewhere (Introduction; Barnes 1958a) but here I shall tackle it from a fresh direction. I shall deal only with a limited aspect of the question, an aspect that I think is of general interest and should be of general concern. I shall look at two concepts, feedback and real time, which are usually associated with the natural sciences, and inquire how they are relevant when we endeavour to work as social scientists. The social sciences take the name of science, and thus indicate their aspirations towards scientific status; this entitles us to judge them by the criteria of science. The failure of the social sciences to achieve these aspirations is shown clearly enough in the fact that, in ordinary speech, when we say that a person is a scientist we imply inevitably that he or she is a natural scientist and not a social scientist. There is no balanced division between two kinds of scientist, natural and unnatural. Instead we have a division into the sheep and the goats, the proper scientists and the social scientists. In this sense, the social sciences are not science. By the social sciences we usually mean economics, psychology, anthropology, sociology, political science, demography, criminology, human geography and a few other disciplines; the boundaries of the category are vague and no one bothers to fight about them. These disciplines call themselves social sciences, and not social arts, or social humanities. This is not merely because work in the social sciences is often inartistic and sometimes inhumane, for throughout the social sciences there is a basic orientation towards something thought to be characteristic of science. Even though some social scientists tend to view with alarm the latest
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
L35
scientific and mathematical innovations in their subjects, it is the relationship of the social sciences to the other sciences rather than to the arts and humanities that raises difficult and important problems. The name science suggests an ordered body of concepts, propositions, scales of measurement and, in some cases, procedures for practical application. On the whole, these features are not prominent in the social sciences. They are certainly to be found in economics, and to some extent in psychology and linguistics, but are present only in fragmentary and disjointed form in political science and human geography, in anthropology and sociology. Here we find an embarrassingly large array of analytical concepts; typically, anyone writing an article feels free to introduce his own new scheme of analysis or to select from the large number of established schemes. As we would expect, these concepts do not fall into any single comprehensive and generally accepted conceptual system. Everyone practises his own combination of analytical eclecticism and inventiveness, and such consensus as there is springs from charismatic teachers who impress their analytical schemes on their pupils by force of personality rather than by the tested evidence of their scientific investigations. Hence we have a professional arena where there are a few foci of concentration, distinctive 'schools' of sociology, each with its own specific interests, vocabulary and revelation of the truth. In between the foci are hundreds of unattached investigators taking concepts and techniques from here and there, inquiring into whatever topic comes their way and reporting on it in whatever terms seem plausible. From this field of activity emerge innumerable propositions and hypotheses, but these do not hang together to form an ordered structure of knowledge about the social world. Many of the propositions put forward are inherently incapable of falsification in that there is no possible evidence that, if discovered, would refute them. There are scales of measurement, but these apply to only a small fraction of the characteristics discussed. There are many recipes for practical application, but these spring as much from native common sense and unanalysed insight as from scientific theory. I do not need to say more on this tale of woe. A quarter of an hour spent trying to collate the aims, concepts and arguments of, say, three different elementary textbooks of sociology and social anthropology, preferably one American, one French and one from Oxford, will demonstrate what I mean. Nor is it necessary for me to spend much time on the other side of the contrast: the aseptic logical imperturbability of mathematics; the four basic forces that in physics link the elementary particles with the stars and galaxies; the molecular transformations that in biology link the inheritance of blood groups with pains in the back. We know that quasars are hard to comprehend, that eminent statisticians argue about the evidence for smoking as a cause for lung cancer, and that professionally designed bridges and cooling towers
I36
MODELS OF SOCIAL SCIENCE
sometimes fall down. But these puzzles, arguments and failures only stimulate growth in the ordered body of scientific knowledge and do not destroy the validity of natural science as such. I think the contrast is clear enough, and I want merely to register my preliminary point, that sociology is not at present very scientific. Some, but not all, of my colleagues recognize this state of affairs, and of these, some, but not all, deplore it. They advocate a remedy which calls for closer copying of the techniques of the natural sciences, a more rigorous methodology, an approach to the search for knowledge closer to the laboratory than to the study and the studio. My argument here is that in some important respects, though certainly not in all, sociology should spurn the model of the natural sciences, and should face up scientifically to problems which hitherto have largely been enountered in the arts. This opinion derives from a consideration of the effects in social inquiry of the phenomenon of feedback and of the significance of real time. This may seem a strange thing to say, for we tend to think of feedback as a process found in control mechanisms and similar machines and as having nothing much in common with arts like history and literature. Nevertheless, it is a concept that has been used in an imprecise sense in discussions of social and other systems from the point of view of cybernetics (von Foerster 1952), though its utility as a precise term seems to be low (Ashby 1956: 53, 8 0 - 1 ; cf. Barth 1966). I shall use the word only in an imprecise sense. Likewise we hear of real time chiefly in connexion with computers. It appears as the one manifestation of time that cannot be speeded up by the electronic sophistication of the computer. When working in real time, 'the computer interacts with the external environment and carries out appropriate action as the situation develops' (Greenberger 1966: 196). We can feed into a computer statistics relating to, say, births and deaths that have occurred during several centuries and obtain an analysis within a few seconds. But however powerful a computer we use, we cannot feed it with information about events that have not yet happened. The computer must be as patient as mankind, and wait until future becomes present and past. This may seem far removed from the work of many sociologists, for we do not all use computers, or even slide rules, and do not all assume that the society we study has the characteristics of a machine. But the notion of real time applies to any analytical contraption with an input and an output; computers and social scientists are merely two examples of this general class. I think that the connexion becomes clear when we consider the difference between the phenomena investigated by the two kinds of science. In the natural sciences, feedback occurs between one part and another of the system being studied, whether it be a device for orienting a satellite towards a star or a mechanism for keeping a whale at a
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
137
constant temperature. No feedback occurs between the phenomena being studied, the satellite or the whale, and the investigator who studies them, the astronomer or the zoologist. As the water in the tank grows colder, the blood courses more vigorously through the veins of the whale, not the veins of the zoologist who is on dry land. In the social sciences, however, the position of the investigator is quite different. In many instances he finds himself part of the social field that he is trying to understand. By his actions he disturbs or modifies the field in various ways, some expected and others unexpected, some convenient and others most inconvenient, as we shall see. Obviously there is some analogy here with the Heisenberg principle of uncertainty in quantum physics, but the analogy is not very close, for in social science the investigator does not merely interfere with the field but is himself constrained by it in a way that has no parallel in physics, as I understand it. In parenthesis, I suggest that had so exact a relation as Heisenberg's, involving a natural constant and the product of two measures of precision (Ikenberry 1962: 25-6, 74-5), been discovered in the social sciences, we would have hailed it as a principle not of uncertainty but of certainty (Bohm 195 7: 85n). Rigorously defined equations and even inequalities are still very rare birds in our fields. To resume: in the arts, unlike the natural sciences, we are inextricably concerned with the impact on the observer, the reader, the hearer, or the viewer, of the art object, whether this be a painting, a novel, a poem, a play or a symphony. For example, greatness in a work of art is sometimes defined as the quality of producing irreversible changes in those who encounter the work. 'Life has never been the same for me since I saw King Lear', we may say as evidence for the play's greatness. Some forms of art need no audience, and the Venus de Milo may look as beautiful as ever when the Louvre is closed and empty. There is no feedback, for though we are affected by the sight of the statue, it is not altered when we look at it. On the other hand a superb performance of Swan Lake requires a responsive audience as well as outstanding dancers. Feedback is here not only significant for evaluation but necessary for execution. It is often the same in the social sciences. The interaction between the young social scientist, encumbered with tape recorder and camera, and his object of study, say a gang of Surfies discussing their week-end strategy, lacks, alas, the ecstatic quality that belongs to Swan Lake; also, with luck, it lacks the irreversible quality we associate, say, with first hearing Beethoven's Ninth Symphony. Nevertheless, in the field situation there is involvement, response, constraint, two-way interaction; and these processes have not to be merely felt and expressed intuitively and artistically, but in social science also have to be domesticated, as it were, measured and analysed objectively so that if necessary they can be fed into a computer.
I38
MODELS OF SOCIAL SCIENCE
Recognition of the importance of feedback has come only gradually and unevenly to the social sciences. Studies carried out among distant illiterate tribesmen or semi-literate slum dwellers in the metropolis did sometimes influence or disturb the lives of the people studied; more often the early pioneers of our discipline changed their view and understanding of society as a consequence of the studies they carried out. But these were independent effects which did not cumulatively interact with each other. In particular, publication of results was an event significant to the authors of books and papers, not to the peoples described in these works. It is true that even now some studies can be carried out under quasi-laboratory conditions, with informants who are scarcely aware that they have become the subject of scientific scrutiny, or with investigators who are isolated from their objects of study by the passage of time. But even in the most superficial inquiries, it is increasingly the case both that informants behave under scrutiny in a way significantly different from their behaviour under ostensibly unscrutinized conditions, and that knowledge of the results of any one inquiry influences the way in which people respond to the next. I am not simply referring to the dislike people may have to reading seemingly neutral facts about themselves. We know, for instance, that after the publication of the results of the i 9 6 0 census in the United States, demands for recounts were made by some chambers of commerce in cities whose population had declined (Becker 1964: 274). Nor can I here discuss the way in which the interests of the observer determine to a large extent the phenomena he selects as significant for study, a topic which, for example, has recently been excellently discussed by Srinivas (1966: ch. 5; cf. Beteille 1965: 9-12) for the case of an Indian sociologist studying Indian society. This is an instance of feedback but of a different class from those I shall confine myself to here. What I have in mind are those situations where the investigation itself, or the publication of the results, leads to changes in the behaviour of the group studied, so that either it becomes impossible to continue with the investigation, or to carry out any subsequent investigations on the group, or the subsequent investigations are necessarily different from those that have gone before, since the group concerned has changed its characteristics. Both the speed with which fresh and disturbing information about itself flows back into the social system, and the specificity of the influences that are generated by social inquiries, are phenomena that vary greatly. At the one extreme, for example, we have the quick and specific influence exerted by the publication, before an election, of the results of public opinion polls about how people plan to vote; everyone seems agreed that knowledge of poll ratings influences how people actually vote, though just what happens is still obscure. A second and unspecific example is the knowledge, ambient in most literate democracies at least, that there are investigators, mainly women, who go
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I39
around speaking to strangers and asking them seemingly arbitrary questions. Whether or not individuals are prepared to accept this as legitimate, at least it is recognized as a special kind of activity, and answers given to these questions are likely to bear an increasingly complex relationship to answers that might be given to similar questions if these arose in the course of normal everyday intercourse. A third example, specific and slow, can be seen when an institution meets to consider a publication describing its activities as they were observed several years previously; the institution may deliberate how to put its house in order, or how to prevent its secrets being further disclosed. There are many other examples that could be cited, but these should be sufficient to indicate what I have in mind. The phenomenon has emerged most clearly in connection with community studies, in particular with the study carried out in 'Springdale' in upper New York State (Vidich and Bensman 1964) and in a more dangerous context in the United States Camelot project in Chile (Walsh 1965: Horowitz 1965, 1967; Bernard et al. 1965-6; Beals et al. 1967; American Anthropological Association 1967; Deitchman 1976) but it is present, in less dramatic form, in almost any study we now undertake. This raises certain difficulties. The investigator is no longer a man from Mars, looking scientifically down the microscope. To the people being studied, the investigator falls into some pre-existing category, either a man or a woman, either something like an administrator who asks useless questions, or like a busybody from the Town Hall, or perhaps something like a newspaper reporter, or even another of those college boys who come every summer with their notebooks. Because of this, the investigator finds that certain kinds of behaviour are expected and even required of him. If he asks only for the sort of information readily given to visitors he may have no difficulty but he will learn little of value. Once he begins to probe into the secrets of the family or the village or the sports club, he is likely to meet resistance. He may even be told that 'These are not the sort of questions they asked last year'. In part the investigator finds a role assigned to him specifically as an investigator, and in part he finds himself bound by the fact that the people he is studying perceive him as, say, male, unmarried, impecunious and probably unorthodox. Hence he is expected to behave as other impecunious unmarried males and his moral and political opinions are suspect. In addition he must ask only certain types of questions. If he is to do good work, he has to extricate himself gradually and non-traumatically from these bonds, so that he can become free to associate with, talk to and observe, old married men as well as young unmarried ones, and women as well as men; he has to gain the confidence of his informants so that despite his appalling ignorance of local customs and etiquette they are prepared to treat him seriously and to discuss with him the politics and
I4O
MODELS OF SOCIAL SCIENCE
intrigues that make up so much of social life (cf. Berreman 1962). Probably there will always be some limitations he cannot overcome; for instance it is generally easier for a woman observer to study local politics if this is a largely masculine pastime than it is for a male observer to study exclusively feminine domestic activities. A team of assorted investigators can achieve more flexibility, but at the cost of additional complications. In part, the success that the investigator has in gaining the confidence of the people he is studying will depend on how he handles the problem of feedback. In any community information and knowledge are valuables that people husband carefully, and no one tells his rivals and enemies what his next moves are going to be and when they will take place. Yet the investigator is also seeking local knowledge. If he is told facts in confidence and does not pass them on, then the balance of local politics is unchanged, even though there may be a change in the relationship between the investigator and his informant. If the investigator can get his subjects to tell him their secrets, either because he has wealth, or can provide services, or is just a nice chap, then he is helped in his scientific task. But if he filters any of the private knowledge that he has acquired back into the community, his field of study is contaminated in two ways. Suppose for example that he is told by C that A is not really the father of B. If the investigator passes this information on to others, then more people than before come to know this fact, and their actions and attitudes towards A and B may alter. Secondly, C's position is changed by the investigator's action. He has been made to share his secret with others and may find it difficult to go on behaving as if he did not know that there was any doubt about B's paternity. Worse still, he may become generally known as the scallywag who let the cat out of the bag, or the skeleton out of the cupboard, and who told the investigator something that others had been careful to hide from him. So the investigator would seem to be well advised to keep his own counsel, to tell no one about A and B, and to repay C in kind, with information about the private life of, say, Christopher Columbus or some other suitably remote figure. But unfortunately the play is not so simple. C may have told the hapless investigator about A and B in the hope or expectation that the investigator would pass on the information to some outside person to whom C has no direct access but who is on the point of rewarding A in some way dependent on there being no breath of scandal in his family; or it may be that C expects this piece of information to lessen the growing friendship between the investigator and A, which threatens C's relationship with the investigator; or perhaps C hopes that the investigator will pass on the information within the community so as to further some elaborate ploy that C has in progress. If the investigator has been in close contact with the community for a long time then he may be well enough informed to be
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I4I
aware of all these possibilities and interpretations, and to choose between them. But if he has newly arrived, knows little about the history of local intrigues, is worried that he is not getting on with his task fast enough, and is acutely lonely, then he may jump at C's offer of friendship and will record eagerly all C has to tell him. Only later does he learn that C is an inveterate scandalmonger who says the first thing that comes into his head. But the problem does not end when the sands of time run out, the grant expires, and the investigator has to pack up his notebooks and return to the ivory tower. Even if there is no direct feedback from the research seminar or the doctoral thesis, or even from the scholarly lecture to an audience of professionals, the moment anything appears in print the process of interaction between investigator and the people he investigates begins again. Some years ago a colleague of mine published a book about social life in a village somewhere in Britain, and disguised the location of the village by using vague maps and by referring to the village under a pseudonym. The day after publication, a television crew moved into the village and began filming a programme on the villagers who were so scientifically described and disguised in the book. In this case no great harm was done, and I think it is still safe for my colleague to return to his old haunts, but this is not always the case. In the United States, certain tribes in the south-west whose religious ceremonies had been described in print have responded by refusing ever again to allow a stranger to study their customs. In the recent case in New York State mentioned earlier, the inhabitants of a small town, whose political life had been described in a paperback using false names and disguised locations, reacted by lampooning the investigation team in a Fourth of July parade, and by expressing their opposition to any further inquiries. It might be argued that these reactions after publication do not matter, for is not the scientific task already completed? Does it matter what happens afterwards? This would be to misunderstand the situation completely. In the first place the scientific task is never completed; the same team may wish to return to the same community to follow through longterm processes and changes, or others may wish to go there to check the findings of the first investigators. Secondly, if one community is known to have found exposure to scrutiny by social scientists unpleasant, other communities may come to hear of this and refuse their cooperation when their turn comes. Thirdly, the investigators during the course of their work must inevitably become beholden to the people they study for an immense amount of help and cooperation which they can never repay. It is dishonourable then to ride roughshod over the feelings of the community by exposing its secrets to public gaze. The real problem is to determine why it is that people are often
142
MODELS OF SOCIAL SCIENCE
embarrassed to find themselves described in print. It may be that scandals are brought out into the open, or that good deeds done by stealth are publicized, or that facts that everyone knows but no one admits are revealed and have to be faced openly. Whatever the reasons, the feelings are quite widespread. We all like to read about the way other people run their lives, but we do not relish the detailed analysis of our own lives becoming public property. This is a problem that social anthropologists, for instance, have had to face ever since they began to look at the total setting in which even illiterate tribesmen live their lives. I remember naively remarking many years ago to a British District Officer in central Africa that I was studying him too, and not just the Africans for whom he was administratively responsible. I had to learn quickly to express these awkward facts more discreetly. In some situations, it has been possible for investigators to go through the text of their reports with members of the community being studied, so that the final text has had the agreement of the people whose actions it describes. This has been done notably with the Glacier Metal Works project in London and in general has the advantage that significant new data are likely to come to light in the process. But there are often practical and theoretical difficulties against this procedure. It is a very lengthy process, so that publication may be delayed for a year or so; it entails continued contact between the investigator and the community, which may be impracticable if they are situated thousands of miles apart; the community may be of a kind such that no one can take responsibility for anyone else, and may be such that if A agrees, then B is bound to disagree. More importantly, it is a characteristic of virtually all social life that it is based on an imperfect flow of information, and on the evasion of unpleasant facts wherever possible. That is why the publication of a report on a community that states all the unpleasant facts in black and white is so traumatic an event in the life of that community. If the community can be brought to the point when it will agree to face up to the facts about itself, than it is in a real sense no longer the same community. In some therapeutic investigations this may be the desired end, but most inquiries in social science are not aimed therapeutically, and are planned to leave the community undisturbed in its unregenerate condition. Referring again to my own specialism, there is an interesting indication in the monographs written by anthropologists of the shift from studying tribal peoples as scientific objects, with no possibility of feedback, to studying human beings like ourselves, highly capable of talking back at us. In the old days, the misdemeanours of Chief X, the scandals in clan Y, and the religious secrets of village Z were all published in full, with no attempt at disguise. The only things that were not published in plain English were those sexual practices that were thought to be offensive to the innocent metropolitan reader; these were included in appendices written in Latin.
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I43
Nowadays the facts of sex are given for all to read, but the identities of X, Y and Z are concealed. The erstwhile innocence of the metropolitan reader's daughter has been replaced by the chief's knowledge of the law of libel and his ability to pay for lawyers to fight his case (cf. Barnes 1967b: 195). Even if the particular topic investigated is one that is unlikely to put the investigator in danger of the law of libel, or of the Official Secrets Act, publication still remains a matter of importance to the community under investigation. For example, suppose that a social scientist Alpha steers clear of such topics as the structure of political life, or the administration of justice, or the prevalence of bribery, any one of which might lead him into difficulties over publication. Suppose instead that Alpha decides to play safe, and to confine his investigations to, say, folk tales or the prestige ranking of occupations. These can, in fact, be sometimes politically dangerous topics, but suppose that in this instance they have no political implications. He collects his data, annotates and analyses them, and publishes his findings in full in some obscure learned periodical. Alpha's findings are declared to be important by the experts. Can they accept his data? Suppose that his learned colleague Beta wishes to repeat Alpha's investigation. Can he do it? In the first place, some years have elapsed by now since Alpha did his work, and some of the greybeards who provided Alpha with his information have died, so that Beta cannot question them. Others are still alive, but their memories have been dimmed. Since Alpha did his work, TV, the radio, the cinema and missionaries have reached the community, and the original folk tales have become mixed with tales about David Crockett, Moses and the man from UNCLE. Local views on the ranking of occupations may have remained stable but the introduction of new occupations, the mechanization of some industrial processes, and the varying successes and failures of prominent local enterprises may all make it difficult to compare directly data collected a few years apart even on so neutral a topic as this. Even if a benevolent autocracy has kept at bay all these insidious external influences, the inquiries made by Alpha, the value he placed on these old tales that most people already thought unimportant, and the bounty he offered to those with long and interesting tales to tell, may have combined to stimulate a local revival in the telling and elaboration of folk tales, so that they begin to be told under conditions and in contexts different from those of Alpha's day. Even if no one in the community manages to obtain a copy of Alpha's learned dissertation, it is clear that the situation that his colleague Beta arrives to investigate is significantly different from that investigated by Alpha. In fact, Alpha may have stimulated people to tell him folk tales by promising that they would see their stories in print, and probably several copies of his obscurely published monograph are circulating in the community by the time Beta arrives. Those who disliked Alpha are likely to tell Beta that Alpha has got
144
MODELS OF SOCIAL SCIENCE
the facts wrong, while those who accept Alpha's version will turn to it for authoritative rulings if Beta questions their testimony. These are some of the reasons why replication is so hard a condition to attain in the field of social science. In my own work in Africa I remember how I tried at first to collect accounts of tribal history as a check on the various published accounts that had already appeared from the group I was studying. I soon found that if in doubt my informants turned for guidance to copies of the same published works I had read, or to vernacular school textbooks based on them. I found I was investigating not the reliability of earlier investigators but the thoroughness with which my informants had absorbed what had been published (Chapter 7; Barnes 1967J: 4). It may well be argued that there are many kinds of social inquiry to which these considerations do not apply. For instance, to go back to an earlier example, if a sample of informants is asked to rank various occupations in terms of prestige, I think it is unlikely that, say, the prestige attached to the occupation of casual labourer will be further lowered by the publication of the fact that 97 per cent of informants rate this occupation lowest in the set. As yet, prestige in this domain seems to be determined by factors other than the results of sociometric tests, and the results of these tests are usually presented in a form unlikely to have any public impact. But we have only to move a little way, to inquiries about social class for instance, to find that what previous investigations have revealed, or are thought to have revealed, influences how present informants think about their social world. Even if in some sense social class is an illusion, it is certainly a social as well as a sociological illusion, and hence ceases to be an illusion at all. Some glossy magazines appear to be dedicated to making the illusion real. Seeley, in an analysis of what I would call feedback in his Crestwood Heights study in Canada, writes: We are not far from the point where what the social scientist said yesterday is part of today's mores ... Who can think, talk, or write today without explicit or implicit use of terms such as mores, folkways, social structure, culture, role, self-conception and social status? . . . How many college students have used Kinsey to justify what they otherwise might not have done? (Seeley 1964: 159) In general a significant part of the social system, at least in Western democracies, is social science, which is thus embedded in the social environment that it seeks to analyse and which modifies this environment by studying it. Even in less sociologically sophisticated social systems, the activities of social investigators become part of the system they study, though the notion of social science itself may not. In either case, the sociologists have to work with a much greater self-awareness than is called for in the natural sciences.
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I45
There are ways in which this limitation can be partly overcome, but the problem remains serious. One reason for this lies in another basic point of difference between the social and natural sciences. In our inquiries, real time is often a significant dimension of the phenomena we seek to understand. Suppose for simplicity that, in our example, Alpha does not affect in any relevant way by his presence the flow of social action that he observes. Then what he observes is essentially unique; there is no guarantee that he or anyone else will ever see the same kind of things again. In the natural sciences this is seldom the case. In physics, an experiment performed on a Monday is likely to give the same results as a similar experiment carried out on Tuesday; the day of the week is not a significant parameter. In geophysics a thousand ages are but a moment gone and much of what we measure today we can hope to measure a hundred years hence and get the same answer. But in social life real time is usually one of the variables in our equation. We study the social life of people who are creatures of history, and we have to take them as we find them. We say that people grow old and die, and others take their place, but this happens only in an analytic and approximate sense, for the place the new generation takes has often shifted slightly. The king dies, and he leaves the office and position of kingship greater or lesser than he found it by the actions he has taken or failed to take during his reign. We cannot experiment, or at least we can experiment only in a very limited sense; for the most part we can only observe. We cannot even repeat many of our observations, for when we return to the scene a second time, we are no longer looking at the same scene. Furthermore, we ourselves have changed, for the investigator grows older at the same rate as his informants. The light-hearted innocence that characterized him on his first field trip has been replaced by the informed wisdom of middle age; or at least this is what his informants expect him to have. Consequently he has to ask his questions in a different form and he gets different answers. He is assumed to know already much of the culture he studies and the fresh information passed to him is of quite a different quality from that he received before. Likewise he finds that he is expected to interact with the same individuals as before who are now much older. He cannot so easily make contact with fresh individuals in the same social positions as before, for he is himself older than he was and, more importantly, is known to be older by his informants. Again, we have to note that these considerations apply most forcefully where a single fieldworker is engaged on an intensive study of some delimited social field over an extended period of time. If a team of quasianonymous students administer some unseen investigator's questionnaire to an ostensibly virgin random sample of respondents drawn from a large universe, then many of the difficulties I have mentioned do not apply. This
I46
MODELS OF SOCIAL SCIENCE
method of inquiry brings us much valuable information and we would be greatly handicapped if we could not use it. Nevertheless it is a method that reveals problems rather than solves them. Once we leave the easily surveyed prairies of sociography and turn to the thickets and forests of sociology proper where the going is tougher, once we seek to understand social processes and mechanisms, we have to shorten our focus and thus begin to run into these difficulties. Intensive investigation entails some amount of feedback between observer and actors. Social processes are not instantaneous, and the periodic time of many processes is significantly long compared to the lifetime of the observer. Many social processes take a generation to complete, and it is seldom that a single observer can follow through a complete sequence from start to finish. We have to improvise to build diachronic models from synchronic data. Let me take an example from my work in Norway. I am interested in seeing how leadership in farm families is transferred from father to son. I have been able to collect some limited public information from official records over the last hundred years about that moment in the developmental cycle of the family at which legal title to the farm is transferred from one generation to the next. But this legal conveyance occurs on only one day in the course of a social process extending over twenty years or more. The growing son leaves school and starts work on the farm as his father's unskilled and almost unpaid assistant. He goes away for some months to an agricultural college, returns full of new ideas and works part-time on the patrimonial farm, taking better-paid temporary jobs as well to acquire enough money to contemplate marriage. Often the father sells the farm to the son when the latter marries, but continues to work actively for many years. It is he who now becomes his son's almost unpaid assistant, not unskilled but obsoletely skilled. This is a composite picture of what happens. In the limited period of fieldwork that I have carried out in Norway I have been able to observe a variety of families, each at a different phase in this protracted process, and thus have been able to build up a synthetic generalized picture of what presumably happens to some extent in each family (Barnes 1957b). This analytical procedure is a poor but necessary substitute for observing the full sequence of phases in one or more families, for the families available for study are haphazardly placed at various points in the cycle, and differ in many other respects from one another. The transition from synchronic evidence to diachronic interpretation has many analogies in the natural sciences, as for instance in the Hertzsprung diagram, whereby data on the luminosity and surface temperature of stars are made to yield information about their age and evolution (Sturve 1950: 31-46). In social inquiries, the method is often used to elucidate cyclical processes, where analysis is complicated by the fact that non-repetitive and cumulative
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I47
changes are occurring in the society studied alongside these cyclical processes. Adequate written contemporary records can enormously extend the range of available information, but adequacy has to be judged relative to the problem in hand. We are indeed fortunate if some earlier writer happens to have written down just the very facts we are now interested in. For instance, we can compare the Anzac Day ceremony we observe with published accounts of earlier ceremonies, but we have to make do with whatever happens to have been recorded about those earlier events. I am sure that it is possible tofindout what music was played at the coronation of Queen Victoria, but it is unlikely that we shall ever know much about the distribution by sex and age of the crowd that cheered her on her way to the Abbey. The great advantage of the fruit fly to the geneticist is that its generations are so much shorter than his own; he can hope to study flies from a very large number of generations during his own working lifetime. The geophysicist studying the evolution of the earth has only one earth from which to gather his information, and most of the processes he is interested in take place so slowly that his information has to be treated as referring to a single instant of time. Like the social scientist and the astronomer, he then has to devise ways of relating synchronic information to diachronic hypotheses. Nearer to ourselves, the forester may have to study the growth and maturity of trees planted by his professional fathers and grandfathers. But when we study the life of man in society, the observer and the observed mature and grow old at exactly the same rate. Hence we always have to strike an uneasy balance. We can either look only at those aspects of social life that change very slowly, if at all, like, for example, fundamental social principles, so-called, such as the principle of equity, or justice, or primogeniture, or representative status, or structural phenomena like social stratification, suicide rates, or assimilation; or we can look only at the individual instance and treat it as unique, whether it be the French Revolution or the career of Sir Robert Menzies or a day in the life of a middle-class family in Mosman. The third choice is to steer a middle course, to admit that events never exactly repeat themselves, but to seek to identify what occurs regularly and what is unique to the occasion. While I regard this as the most fruitful choice, it is also the hardest to follow. It calls for great skill in observing and recording the ephemera of social life, the conversations and the gossip as well as the unchanging proverbs and dogmas. In the language of the theory of games, we have then to study not only the rules of the game but also the choices actually made in play. Since the play never comes to an end, and the rules are gradually revised, and we have the opportunity to make observations only for a limited period, there is always a great deal of uncertainty about our analysis. Some recurrent social events have a periodic time of less than a
I48
MODELS OF SOCIAL SCIENCE
generation and any one observer may hope to be able to study two or three or half a dozen sequences in his own lifetime. Provided that his framework of analysis does not change too much as he advances in wisdom, he may be able to isolate the invariate from the variable features. General elections are an obvious example. There is something unique about every general election, and even more so about every by-election. Nevertheless, the growing number of studies of elections demonstrate some probabilities in election behaviour, and have destroyed some beliefs about elections that were previously held. Here I interject a plea for greater recognition of the value of making the same kinds of inquiries in different places and at different times in almost routine fashion. We all like to be professional innovators and are apt to disparage work by our colleagues that we regard as merely repetitive and confirmatory of what we already know and expect. Yet more replication of empirical inquiries would give a firmer base for our interpretations and it is unfortunate that replication is frequently impossible due to changes in social conditions. It is doubly unfortunate that we are so prone to tinker with our analytical procedures (Bonjean et al. 1965). Analytical innovation often rules out the possibility of comparability with earlier inquiries while providing little or no increased understanding of our own data. One of our greatest needs, though certainly not our only need, is a growing corpus of data collected within the same analytical and methodological framework. Only then can we test effectively the adequacy of any scheme of analysis, and confidently discard it if it proves wanting. We certainly need in sociology all the new ideas we can get, but new modes of inquiry should be planned so as to derive the maximum amount of information from studies that have been conducted earlier. A final point. As we try to determine what features of social life are significant for inquiry and analysis, we have to look at the significance of these same features in the lives of the people experiencing them. The experience of the observer is always matched by the experience of the actors he observes. A man votes in many elections during his lifetime, and he does many other things of significance many times. He goes to several weddings and to many funerals, he acquires expectations about what will happen, and his behaviour and that of his fellows may remain more constant and predictable than it might otherwise be because of the practice or reinforcement the established pattern of behaviour continually receives. But any man attends his own father's funeral only once. The occasion is unique not only for any observer who happens to be studying him, but also for the actor himself. It is reasonable to expect here a greater range of variation in what actually happens than in processes in which, as it were, the actors are rehearsed more frequently. I think that I have said more than enough to indicate the limitations of
FEEDBACK AND REAL TIME IN SOCIAL INQUIRY
I49
applying to social inquiry the model of the natural scientist working in a laboratory, studying objects that cannot talk back at him. There are many other ways in which the fact that we are ourselves human beings embedded in society influences how we go about studying other human beings equally embedded in a social matrix. The closer we come to studying our own social life, the shorter the sequence of feedback becomes and its effect becomes more significant. Thus for instance when an institution like a factory or a university decides to investigate its own operations, using some special team of inquiry, changes are likely to occur even before the team gets to work, in the hope that the results of the inquiry will reflect more favourably on the various parts of the institution concerned. These changes are often aimed to produce maximum effect in minimum time and do not necessarily coincide with the changes designed to produce longer-term results that might otherwise be recommended following the inquiry. I certainly do not suggest that for this reason we should restrict ourselves to the study of remote and powerless groups of people, or to good safe topics that can be safely reported on in academic obscurity without a flicker of response from the people whose behaviour we describe. Nor should we accept as final and unalterable those situations where an inquiry into an institution has caused other institutions of the same kind to resist proposals for further similar inquiries. On the contrary I hold that the justification for social inquiry lies in the opportunity and ability it gives us to understand better how we ourselves behave in society. It is because our ultimate professional goal includes understanding ourselves that we have to give attention to the phenomena I have tried to outline.
TIME FLIES LIKE AN ARROW
Like the papers reproduced as Chapters i and 5, this essay was written as a contribution to a Festschrift, but for one that never came to fruition. I had published a study of one aspect of the work of Claude Levi-Strauss in which I had tried to subject his writings on kinship to close textual scrutiny (Barnes 1971m: 101-75). Other commentators had, so it seemed to me, been greatly inspired by his work but had selected from it valuable and stimulating ideas rather than analysed the corpus as a whole. He had become a cult figure, a trend I could understand but which I deplored. I thought that it would be salutary to contrast his work with that of two other writers on kinship, Murdock and Fortes, whose frameworks of analysis were very different from his. Almost all aficionados of Levi-Strauss found my study insufferable; one reviewer asked plaintively why the study of kinship should be so dull {TLS 19 71). An invitation to contribute to a Festschrift for LeviStrauss gave me the opportunity to present a paper in which I was free to choose whatever ideas I liked from his writings, without bothering too much about how they fitted, or failed to fit, with other parts of his work. As I said elsewhere, whatever else they are, contributions to festschriften should be festive (Barnes 1974a: 448). Given Levi-Strauss' repeated use of the contrast between structure and history, I chose his treatment of the notion of time as my starting-point, paying particular attention to his use of time in models of different types. This enabled me to clarify my thinking on the discipline of history, a topic I had begun to grapple with many years earlier when writing Chapter 7. I5O
TIME FLIES LIKE AN ARROW
1 Introduction It is scarcely a historical accident that two anthropologists who have made major but quite separate contributions to our understanding of human thinking have also made significant analytical use of the concept of time, have written at length about the relation between anthropology and history, and have sought to delineate the position of anthropological study itself in the intellectual and political history of our own civilization. Yet though they grapple with the same topics, they remain far apart. Witchcraft, oracles and magic among the Azande (1937) differs from La pensee sauvage (1962a) in scope, method and conclusion, while Anthropology and history (1961) diverges from the Lecon inaugurale (1966c) in concepts, objectives and intellectual affiliation. 'Nuer time-reckoning' (1939) perhaps lies closer to Levi-Strauss' work (cf. 1963b: 289), but the gap is still large. What brings Evans-Pritchard and Levi-Strauss together is an interest in determining a set of contrasts and similarities dividing and uniting the two disciplines of history and anthropology; they are interested, as it were, in ensuring that the search for an efficient division of labour between the disciplines does not degenerate into a demarcation dispute (cf. Lefort 1952: 99-104). Many of their professional colleagues have ignored this issue. There are certainly several contingent or historical reasons why these two writers in particular have concerned themselves with this interdisciplinary relation, but part of the explanation lies, I think, in their common search for order or pattern in data that are, at first glance, characteristically disorderly, random and chaotic. Human thought, so we might imagine, is free of the constraints that systematize human action and, short of 1984, we are all free to think as we please, however coerced by culture and society we may be in our acts. And yet, so it seems, if we look carefully enough we can discover a pattern or structure in our thoughts that contrasts sharply with the apparently arbitrary and unpredictable sequence of historical events. Evans-Pritchard and Levi-Strauss respond to this paradox in complex but different ways, and the views of neither can be stated adequately in a few sentences. Evans-Pritchard, like many of his British colleagues, writes in the shadow of that after-dinner remark by a legal historian made nearly seventy years ago: ' . . . by and by anthropology will have the choice between being history and being nothing' (Maitland 1911: 295). Evans-Pritchard (1961: 20-1), roughly speaking, chooses the first alternative, whereas for Levi-Strauss (1962b: 45; 1963b: 24; 1964b: 546) history and ethnology are a 'true two-faced Janus' and have a 'mutually complementary nature', so that historians should confine themselves to the study of process and anthropologists to the study of structure. Evans-Pritchard agrees that anthropology and history are inseparable but rejects Levi-Strauss' division of labour, while
152
MODELS OF SOCIAL SCIENCE
Levi-Strauss designates Evans-Pritchard as 'one of the champions of purely historical ethnology' (1966b: 161 f.n.). It is impossible in this article to tackle both Evans-Pritchard and LeviStrauss; either alone is a more than adequate challenge to the would-be commentator. Here I shall simply attempt a few remarks prompted by LeviStrauss' handling of the notion of history and of related concepts in which the idea of time enters in some form or other. If my remarks help to clarify the reader's appreciation of Evans-Pritchard's position, so much the better, but that is not my main objective. In his more recent writings, Levi-Strauss' formulation of some of his arguments has been tempered by his debate with Sartre and his reaction to 'a tendency which appears to me very prevalent in contemporary French philosophy, of considering that historical knowledge is knowledge of a type superior to others' (Levi-Strauss et al. 1963: 648). However tactically important this reaction to Sartre's Critique de la raison dialectique (i960) may be, it seems to me a separate issue. I am not competent to discuss it and shall try to steer clear of it. 2 Invention or discovery? Poetry or science? One of the activities of present-day scholars in a wide variety of disciplines is model-building, whereby they attempt to construct some kind of logical picture isomorphic with a portion of the world of reality or imagination. Levi-Strauss is certainly one of the most notable of model architects. The profusion of available models has made us all aware that the design of a model depends greatly on the skill and perception, and also on the whim, of the builder as well as on the characteristics of the phenomenon it must mirror, but we are less in agreement about the origin of the bricks used in models. Are these also custom-built, or are they merely drawn from some pre-existing standardized pool, a kind of Platonic G.I. store? For example, most people would regard 'pattern variables' as one of Parson's inventions rather than one of his discoveries. On the other hand, in Fortes' view, kinship is an 'irreducible principle' which, it seems, exists, and always has existed, independently of the Tallensi and the Ashanti and of anything that Fortes may have written. It is characteristic of Levi-Strauss that, more than his colleagues, he has been explicitly concerned with the epistemological status of his analytical tools. This is well demonstrated in his treatment of those tools he uses to discuss time and the absence of time. Equally characteristically, it is often hard to reconcile his general pronouncements about methods with instances of these methods in use. For although synchrony and diachrony, the contrasted pair he uses most frequently in connexion with time, are usually treated as analytical categories, conceptual tools for use by an inquirer as he sees fit, Levi-
TIME FLIES LIKE AN ARROW
I53
Strauss also writes of these and other categories as if they had an autonomy of their own. I think that this is the aspect of his work that, more than all others, puzzles the reader schooled in the empiricist tradition and habituated to a mundane style of exposition. It is important not to dismiss this characteristic as a Gallic idiosyncrasy. Two separate issues are involved, both crucial to understanding what LeviStrauss has to say. First, these analytical categories are presented by LeviStrauss not as 'all my own invention' but as categories that are already there in the human mind waiting to be discovered. The methodological apparatus he uses is not contrived or fabricated but is imposed on him by the data; his task is a 'search for mental constraints' (1966c: 54), and for the Immanent necessity' which presides over the genesis of the acts of the human spirit (1967b: 124). Hence, in a sense, the categories already have autonomy, for they exist in their own right before we are fortunate enough to discover them. Therefore we should not be too surprised that although Levi-Strauss (1966c: 55; Levi-Strauss et al. 1963: 631, 633) accepts the description of his intellectual position as 'Kantianism without a transcendental subject' (Ricoeur 1963: 600, 618; cf. Lefort 1951: 1409-10, f.n. 1; Levi-Strauss I966d: 294), his categories nevertheless often behave as if they had private transcendental lives of their own. Thus he speaks of diachrony as the victor in its (his, her?) battle with synchrony, of the collaboration between diachrony and synchrony, of societies that take the side of 'history' against 'system', of other societies that deny history, and of myths that, in a certain way, 'think themselves out among themselves' (1966b: 154, 231, 233, 234, 236; 1966c: 56). The reader who likes to keep his feet on the ground may prefer to treat these expressions as mere figures of speech, embroidered perhaps by translation, but they are so persistent in Levi-Strauss' work that he should realize that he must come to grips with them if he is to appreciate more than a part of what is being said. For my second point is that we misunderstand Levi-Strauss profoundly if we treat these flourishes as merely stylistic exuberances. Many readers, it seems, hold that much of the difficulty in Levi-Strauss' writing, as well as much of its attraction, derives from his literary style rather than from the complexity of the ideas he is trying to convey. If only he would write more prosaically he would be less entertaining but more intelligible, testable and scientifically useful. I think this is a mistaken view. It is easy for the English-speaking reader to be impressed with the differences between the modes of presentation of Levi-Strauss and his Anglo-Saxon colleagues, but more impressive are the similarities of style, vocabulary and form of argument manifest in the published exchanges between Levi-Strauss and his French critics and commentators. Every writer, whether he be poet, humanist or scientist, writes to communicate to an audience, and in this
154
MODELS OF SOCIAL SCIENCE
world of messages diverse audiences understand diverse codes. The Parisian code differs significantly from those in use in Britain and America. In a rough way we can arrange styles of writing along a continuum, with at one end poetry, appropriate to a dialogue between human beings sharing a common culture, polysemic, ambiguous rather than redundant, and, as Levi-Strauss (1963b: 210) says, essentially untranslatable. At the other end of the continuum is the computer programme, designed for a dialogue between machines, austerely monosemic, never ambiguous, with a controlled amount of redundancy, supremely translatable and, of course, absolutely unreadable. Most writing in the social sciences in English falls firmly into the middle range of this spectrum, too specialized for Les temps modernes, too imprecise to come anywhere near a computer, and yet striving towards precision and the elimination of ambiguity. Poetry uses the language of everyday life, while the computer is forced to use a special vocabulary and grammar of its own; social science fights uneasily to validate its claim to be science and yet not lose its intelligibility to the educated layman. In a different arena and in a different tradition, LeviStrauss faces in a different direction. With him, as Sontag (1966: 70) puts it, 'there is more awareness of the adventure, the risk involved in intelligence'. Leach (1965a: 36) comments on Levi-Strauss' readiness 'to make sweeping generalizations about human nature'. Indeed we may regard La pensee sauvage as well as Le cru et le cuit (1966c: 48, 56) as a modern myth in the minor tradition of Engel's Origin of the family and Freud's Totem and taboo. But the real contrast with Anglo-Saxon social science arises because the style and vocabulary of structural analysis are taken from the major European humanist tradition of Plato, Kant, Hegel and Marx, with all its imprecisions, imagery, connotations, layers of meaning and glories, tempered only by the scientific austerity of structural linguistics. In this sense it is the most stupendous and audacious example of bricolage we have seen for a long time. Levi-Strauss has chosen to remain within the European literary and philosophical tradition. His success in doing so may explain in part his reluctance to develop a meta-language for structural analysis. He rejects not only meta-processes but also any higher order of concepts scientifically insulated from the conscious and unconscious manifestations of the thought of cultural man. Maybe in his view such concepts are unthinkable by definition. Instead, his tools of trade are taken from this tradition, so that the operations of Vesprit humain are analysed in terms of a subset of its products. As anthropologist Levi-Strauss sees himself as astronomer among social scientists. Though his telescope certainly exploits optical relations that are not invented but immanent and waiting only to be discovered (1966c: 56), the telescope itself is manufactured from local European material. Hence his astronomical reports are evocative as well as
TIME FLIES LIKE AN ARROW
I55
informative to a European reader. Leach (1967b: xvii) writes approvingly that 'Levi-Strauss often manages to give me ideas even when I don't really know what he is saying'. Yet he can hardly attribute these ideas to LeviStrauss. Other people scan the pages of the Bible or Quotations from Chairman Mao Tse-Tung at random in search of inspiration; structural analysis, if it is to be science, as Levi-Strauss (1964b: 548; 1966c: 118) insists, rather than poetry or preaching, should give us more specific directives. Leach gives only half the story when he says that 'Some passages of Levi-Strauss when translated into English seem almost meaningless. Yet to believe that the only things worth saying are those which can be writ plain in the English language seems to me a very arrogant assumption'. For this is precisely the arrogance of science, that seeks to dissect, analyse and measure the world around us and to make verifiable statements capable of being writ plain in any language, including the language of mathematics. These are not the only statements worth making, but we lose both experience and understanding unless we know what kind of statement we are listening to. In the language of the communication engineer, the 'harmonic ambiguities' that Leach (1965b: 26-7) detects in Levi-Strauss' transmissions cannot be dismissed as mere random noise; they form part of the message. These translated 'almost meaningless' statements are intrinsically meaningless only when read as science; as poetry they are meaningless only because they have been translated. To my mind, Levi-Strauss cannot both run with the poets and hunt with the computers. 3 Dichotomy and contrast Typically, Levi-Strauss advances his arguments with the help of a series of dichotomous contrasts, and we can well begin by looking at some of these. We have already mentioned process versus structure and history versus ethnology, as well as synchrony versus diachrony. On a couple of occasions (1963b: 286; 1964b: 542) he combines two dichotomies, empirical observation versus model-building, and mechanical versus statistical models, to obtain the equivalent of a two-by-two matrix, with the four cells occupied by ethnography, ethnology (social anthropology), history and sociology. He frequently contrasts synchrony with diachrony, the axis of simultaneity with the axis of successiveness, and discontinuity with continuity. In his work on kinship and elsewhere, he contrasts reversible with irreversible time, and there are also straight, circular, progressive, empty, non-cumulative, statistical, psychological, visceral and other kinds of time as well as micro-time and macro-time. Strictly speaking, these contrasts might be better described as between different kinds of temporal process rather than between different kinds of time, but
156
MODELS OF SOCIAL SCIENCE
the nomenclature is now too well established to change. This is not the place for a complete catalogue of all Levi-Strauss' contrasts, or even of all his kinds of time, nor am I competent to undertake it. But just as Canberra restaurateurs blithely and unknowingly attempt the impossible task of combining synchronic and diachronic cuisines with their slogan: 'Chinese and Australian meals', so may I attempt the first few steps of the equally impossible task of trying to systematize ideas and concepts that are, in my view, invincibly antisystematic. My first comment is that not all these contrasts are orthogonal to each other. Some terms seem to be nothing more than synonyms, and give colour to the analysis but no additional dimension. Thus for example 'progressive time' and Irreversible time' seem to be the same. Likewise Levi-Strauss refers to 'undulating, cyclical, reversible time' in the terminology applied to members of male Ego's matrilineage among the Hopi (1963b: 301). As far as I can tell, all other instances of 'reversible time' might equally be termed 'undulating' and 'cyclical', so that no narrower discrimination within the category of 'reversible time' is entailed by the two additional adjectives. They are thus analytically redundant. Indeed, 'cyclical time' would be a better label for the category. It is clear, for instance, that the Northern Aranda system of eight sub-sections uses a non-cumulative mechanical model, with no statistical or evolutionary features. Hence the time it uses would seem to be 'reversible' rather than 'irreversible' (cf. Levi-Strauss 1963b: 286). Yet although, as we say, the system repeats itself every four generations, the matrilineal cycles in the system run always in the same direction, in the same oriented temporal sequence of generations, and cannot be reversed. Hence it seems better to describe the system as using a model in cyclical time, with periodicity of four generations. If a model in cyclical time has only two states, as with the Kariera section system, then we may be justified in referring to reversible time, but this is possible only because a two-state model is a limiting case of a multi-state model. In Newtonian mechanics it is easy to construct multistate or infinite-state models in reversible time. Given a Newtonian model of a pendulum swinging to and fro in a vacuum, we can reverse the direction of time without changing the model. I cannot think of a similar model of social events or cultural signs. In most instances where LeviStrauss draws a contrast between reversible and irreversible time, it can be rephrased as a contrast between cyclical and non-cyclical models, all in irreversible time. There are other instances in Levi-Strauss' work where it seems that although the various constrasted pairs are not synonyms of one another, they are nevertheless usually correlated. Thus somewhat lyrically LeviStrauss writes that indigenous social institutions 'are always at a safe distance from the Scylla and Charybdis of diachrony and synchrony, event
TIME FLIES LIKE AN ARROW
157
and structure, the aesthetic and the logical...' (1960b: 73-4). More prosaically, he states that anthropology uses time that is mechanical, reversible and non-cumulative, whereas historical time is statistical, oriented and non-reversible (1963b: 286). Here, the contrasts are between anthropology and history, mechanical and statistical, and reversible and non-reversible. Oriented time does not contrast with non-cumulative time, and can only be a synonym for non-reversible time; the non-cumulative term is thus left uncontrasted. Ignoring this complication, we are left with anthropology and mechanical, reversible time versus history and statistical, irreversible time. In the light of other statements made by Levi-Strauss, we can interpret this to mean, not that statistical models are always in irreversible time, nor that anthropology uses only mechanical models, but rather that most anthropological models are mechanical and use reversible time, whereas most historical summary statements (scarcely models) are statistical and use irreversible time. This exegesis is quite redundant for anyone moderately familiar with Levi-Strauss' writings, but is needed for the reader who accepts his aphorisms at face value. Similarly, Levi-Strauss' reaffirmation that 'the time dimension is of minor significance' in connexion with the difference between history and anthropology (1936b: 285 and f.n. 15; cf. Kroeber 1935: 545-7) has to be understood in the light of his numerous discussions of the precise ways in which the several time dimensions are variously used in the two disciplines. Synchrony and diachrony are terms that appear continually in LeviStrauss' later work. For British readers at least, it is easy to make the mistake of assuming that the contrasted terms mean the same for LeviStrauss as they did for Radcliffe-Brown (1952: 4; cf. Smith 1962: 76-7), whereas in fact the two pairs of usages have little in common. Both writers agree in associating the two terms with simultaneity and successiveness respectively (Levi-Strauss 1966b: 52, 227n). But for Radcliffe-Brown synchronic description is concerned with enduring characteristics, while diachronic description is mainly concerned with systematic change. These associations are lacking for Levi-Strauss, who takes his terms from Saussure (1959: 80-1; cf. Evans-Pritchard 1961: 17) and incorporates into his own work the Saussurian view that while it is legitimate to speak of a synchronic law as a 'principle of regularity', 'diachronic events are always accidental and particular' (Saussure 1959: 93). This is an extreme position not shared by all linguists, and I do not agree with Douglas (1967: 67) that any analysis modelled on linguistics is necessarily synchronic. Indeed Levi-Strauss himself refers approvingly (1963b: 88-9) to Jakobson's 'Principles of historical phonology' (1949; cf. Ricoeur 1963: 599; Merleau-Ponty 1964: 122). Yet I think Gaboriau (1963: 586, 590-1) goes too far when he says that, unlike Durkheim and Saussure, Levi-Strauss (1963a: 68-71; 1966c: 177) asserts, though without demonstration, that
I58
MODELS OF SOCIAL SCIENCE
structural analysis can be applied to the study of process. On the contrary, the Saussurian emphasis on synchronic order leads easily to Levi-Strauss' view that the close association with history required of the anthropologist in his analytical task is primarily to ensure that the raw data can be cleansed of all those accidental and particular features that stand in the way of comprehending the synchronic structure (cf. 1950: xix-xx; cf. Matarasso 1963: 203). The anthropologist's task is 'a question of setting up an inventory of mental enclosures, of reducing apparently arbitrary data to order, of reaching a level where necessity reveals itself as immanent in the illusions of freedom' (1966c: 53). Here Levi-Strauss stands irreconcilably opposed to those of his colleagues who argue for the affinity between anthropology and history, since 'Both deal with events and occurrences, not with man's nature or the timeless necessities' (Arensberg 1957: 98). Accordingly, Levi-Strauss argues that the ethnologist and the historian must stick each to his own last, even if they are permitted sometimes to interchange tools and roles simultaneously. He affirms that we 'cannot at one end and the same time define rigorously a state A and a state B (something that is possible only from outside and in structural terms) and relive empirically the transition from one to the other (which would be the only intelligible manner of understanding it). Even the human sciences have their relations of uncertainty' (1962b: 45; cf. I966d: 304). Maybe; but some of the uncertainty derives from the way the contrast is phrased, as Levi-Strauss' own range of statements suggests. For the viewpoint just cited is matched by others that put the matter differently, as when he says that although at some future date the study of diachronic structures may be possible, for the present it is simply too difficult (1963b: 3, 21; LeviStrauss et al. 1963: 649; cf. Heusch 1958: 806; Gaboriau 1963: 591). At the other extreme he argues (on the same occasion!) that the only possible diachronic or evolutionary models of society would be independent of all environmental and biological factors (Levi-Strauss et al. 1963: 638). This would surely make them models of Vesprit humain pure and unadorned rather than of human society. In his 1953 article on 'Social structure' Levi-Strauss gives tentative approval to the introduction of 'a vectorial factor for the first time . . . on an objective basis into social structure' by Murdock in his demonstration of the irreversibility of direct change from matriliny to patriliny. When this article was republished in revised form in 1958, this passage was deleted, suggesting that Levi-Strauss' views on the possibility of structural models in irreversible time had hardened, despite his earlier exposition of a series of apparently irreversible structural transformations for Australian marriage class systems (1949: 273; 1953: 530; cf. 1963b: 287, 306-7). These apparent contradictions or inconsistencies may be resolved in
TIME FLIES LIKE AN ARROW
I59
many different ways, but two comments may be made confidently. First, the conscious-unconscious contrast is overlooked in some of the statements mentioned. If we stick to the straightforward view that 'History organizes its data in relation to conscious expressions of social life, while anthropology proceeds by examining its unconscious foundations' (1963b: 18; cf. 1950: xxv), and keep in mind the contrast between the unpredictability of what is consciously said and the stability of the unconscious grammatical rules, then it follows that systematic or structural models can be used in the study of unconscious phenomena only, by anthropologists and not by historians. Thus Levi-Strauss accepts Cuisenier's argument that models can be used in economics more than in other social sciences because economics studies processes that are largely immune to conscious and voluntary control (Levi-Strauss et al. 1963: 649-50). But when the conscious-unconscious division between the aims of historians and anthropologists is waived, he then appeals to Heisenberg for an analogy to preserve a diachronic-synchronic division. I find the analogy quite unconvincing and would look rather, though without accepting it either, at Levi-Strauss' assertion (1962b: 45) that processes exist only for those who personally experience them; there are no 'metaprocesses' that will subsume the irreducible experiences of an aristocrat and a sans-culotte during 1789 and provide a processual model of the French Revolution. Thus Levi-Strauss' view of process is diametrically opposed to those of many of his anthropological colleagues and of other social scientists such as, for example, Easton (1966: 143), who writes that his aim is to determine 'the life processes of political systems'. Leighton (1959: 221-4) argues that it is confusing to use the term 'structure' to refer to both 'configurations of apparent motion' and 'configurations of apparent substance' and would reserve 'structure' for the latter. This is a recommendation about nomenclature only. Levi-Strauss goes much further than this. Apart from attaching the term 'structure' to configurations of relations rather than of 'substance', he would seem to deny the possibility that 'configurations of apparent motion', i.e. processes, can possess underlying structural invariants at all. Secondly, these seeming contradictions spring in part from an excess of dichotomy. The simple contrast between synchrony and diachrony fails to make the necessary distinctions between, for example, diachronic models with periodic times of a generation and a millennium, and between noncumulative and cumulative diachronic models. By lumping together, for example, homeostatic and continually expanding models of segmentary lineage systems (cf. Meggitt 1965a; Barnes 1967a), and by putting Kroeber's and Toynbee's models of civilizations and cultures into the same basket as, say, Fortes' (1958a) developmental cycle, Levi-Strauss makes
I6O
MODELS OF SOCIAL SCIENCE
the diachronic task appear to be more difficult than it really is. His own strictures on the analytical limitations of the concept of binary opposition in structural linguistics (1966b: 66; cf. Leach 1964: 1112) might well be redirected. For in his later general pronouncements, which reflect his preoccupation with the study of myth, he loses sight of his earlier generalizations based on studies of kinship. In 1945 he wrote that: Kinship ... exists only in self-perpetuation ... the initial disequilibrium produced in one generation between the group that gives the woman and the group that receives her can be stabilized only by counter-prestations in following generations. Thus, even the most elementary kinship structure exists both synchronically and diachronically (1963b: 47). Hence we have the elaboration of concepts of various kinds of time found in kinship systems. Likewise, in the brief reference to 'subordination structures' in his 1953 paper on 'Social structure' (1963b: 309-11), LeviStrauss envisages the study of 'diachronic changes of the structure', but this is a line of inquiry he has left to others (cf. Heusch 1958: 804-5). hi study of myth there is necessarily a different emphasis, for though as a kind of narrative a myth is usually diachronic, and hence has a beginning and an end, as myth it is Interminable' (1966c: 48). In other words its structure, in Levi-Strauss' terms, is primarily synchronic, even when the structure also has a diachronic dimension, as is made clear in the discussion of the study of Asdiwal (1967a: 17). In the myth, sequences are subordinate to schemata; myth needs time 'only in order to deny its place' (1966c: 61). It is interesting that in his only recent paper on kinship, in contrast to his major studies of myth, Levi-Strauss has returned to the notion of diachronic models of myths, in the distinction he draws between what he calls 'palaeolithic' and 'neolithic' models (1966a: 16). However great may be the difficulties of making satisfactory diachronic models isomorphic with all the complex historic and prehistoric sequences we know of or postulate, if we lower our sights we can find many configurations of facts scattered through time that can be ordered by reference to an appropriate diachronic model. Just as Levi-Strauss appeals to a diachronic physical model of the precession of the equinoxes to illumine the symbolic value of constellations to contemporary South American Indians and classical Athenians (1964a: 242), so for example may a diachronic social model be used to explain, partially and imperfectly no doubt, wars, migrations, soil exhaustion, lineage segmentation, and so on (Barnes 1967J: 29-63). This may be regarded as the study of infrastructures, and therefore history, not part of ethnology proper (1966b: 130-1; Levi-Strauss et al. 1963: 638) or, more charitably, as the study of what is actually happening in Plato's cave, 'the historical and local modalities of the relations between man and the
TIME FLIES LIKE AN ARROW
l6l
world', rather than the study of the shadows on the wall (1966b: 117, 214). But surely structures are fair game for us wherever they are to be found. As scientists we are no longer prisoners and no part of the cave is a sanctuary from scientific inquiry. The analytic closure presaged in LeviStrauss' statement that 'the world of mythology is round' (i966d: 7) can be avoided, as indeed is indicated by that irrepressible curiosity about the real world of people, plants, animals and stars which bursts through even in the most shadowy parts of his work. Likewise, to escape from an amusing but ultimately sterile ballet of symbols in which history and anthropology, synchrony and diachrony, consciousness and unconsciousness, continuity and discontinuity, reversible and irreversible time dance endlessly round each other until the audience decides to go home, we have to break down the dichotomies, establish continua and feed in more facts. 4. Histories and diachronic models Despite his delight in binary contrasts and dialectical somersaults, LeviStrauss has gone some way towards dissecting the different notions we usually lump together under the label of history. But while 'history' is a word of everyday speech that necessarily has many meanings, the label 'diachronic model' still, I hope, belongs to the technical language of science and can be given a precise meaning. Levi-Strauss (1966c: 117) writes that 'certain facts arise from a statistical and irreversible time, others from a mechanical and reversible time'. If we combine this with his association of history with statistical models, noted above, we can easily draw the false inference that statistical models are necessarily diachronic, particularly if the data they utilize are drawn from several points in time. But this is not so. If, for example, we look at the account of Yako marriage given by Forde (1951), characterised by Levi-Strauss (1963b: 284) as supplying both a mechanical and a statistical model, we find that the frequency of various incidents of marriage is given for a succession of quasi-temporal periods, defined by the marriages entered into by members of successive age-sets. Thus we have a set of statistical models (of a very rudimentary kind indeed, nothing more than a statistical summary), each associated with a particular temporal period. Looking at the sequence of models, we can make statements about the average values of certain frequencies during the interval stretching from the date of the earliest model to the date of the latest. We can also describe the trend of frequencies during the same interval. But the state of affairs in any one model tells us nothing about what to expect in any other. We can arrange these simple models in chronological order only by referring to a characteristic that is extraneous to each model, in Forde's example the ordinal number of the age-set associated with it. To explain secular trends we have
162
MODELS OF SOCIAL SCIENCE
to appeal to other extraneous factors (the price of palm oil, opportunities for government employment, and the like) not forming part of the models. Hence the amalgamated model of Yako marriage, covering the total time span for which Forde has information, is diachronic only in a weak or fortuitous sense. It is open to the criticism brought by Saussure (or his editors) and repeated by Ricoeur (1963: 599) and Douglas (1967: 67) that diachrony is subordinate to synchrony, and is understandable only as a comparison of earlier and later synchronic states. The same comment can be made on the studies by Sutter and Tabah (1951) of consanguineous marriage in France, cited with approval by Levi-Strauss. If we turn to the study by Zelditch (1959) of Navaho marriage, also welcomed by Levi-Strauss (i960: 200; 1962b: 42), we find here an explicitly constructed model in which the state of affairs in one generation partly determines what happens in the next succeeding generation. Thus this model is diachronic in a much stronger sense than those used by Forde, and is not subject to the Saussurian criticism. Whereas Forde's models are essentially synchronic, Zelditch uses irreversible time; the state of affairs in one generation does not determine what happens in the preceding generation. The time used is not cumulative, for the effect of the pattern of marriages in any one generation on patterns in succeeding generations is quickly damped down. On the other hand the model does not repeat itself, Aranda-wise, after a fixed number of generations, nor does it eventually settle in some stable equilibrium state. Hence it may be described as a model in irreversible, non-cumulative, non-cyclical time. The same description applies to the model used by Kunstadter et al. (1963), except that the cumulative/non-cumulative contrast may be inapplicable (cf. Levi-Strauss 1966a: 16). In both examples, oriented noncyclical time is an integral dimension of the model. Thus it is clear that the construction and utilization of truly diachronic models has already begun. The two models mentioned here are models that attempt an isomorphism with events, 'what actually happens', rather than with marriage rules. It seems probable to me that diachronic models may be obligatory for the successful study of the structure of events, but only ancillary to the study of the structure of signs (cf. Fortes 1949b: 58-61, 84). But that is another contrast (or jigsaw) that cannot be examined here. I merely comment that if, as Levi-Strauss (1963b: 240) maintains, structural dialectics promote historical determinism 'by giving it a new tool', then diachronic models supply it with a calculus. In the use of diachronic models, structural analysis joins forces with other intellectual traditions within anthropology and sociology to achieve greater precision and explanatory power (cf. Godelier 1966: 863-4; Schapera 1962; Smith 1962: 79-85). On at least three occasions Levi-Strauss (1962b: 45; 1965: 51; 1966b: 238; cf. 1963b: 16-17; 1966c: 56-7) asks if it is possible to write a history
TIME FLIES LIKE AN ARROW
163
of the French Revolution, and each time says that it cannot be done. All the historian can do is to construct a myth, based on his own selection from the facts. 'History is therefore never history, but history-for' (1966b: 257). This contradiction alerts us to the shifts in meaning given to the word 'history'. Gaboriau (1963: 597) draws attention to Levi-Strauss' (1966b: 250-1) complaint about the multiplicity of meanings that Sartre gives to the term 'history', a topic I must eschew in this article; but LeviStrauss himself follows the usage of everyday speech in assigning more than one meaning to histoire. 'History-for' must refer to the results of the historian's labours, and not to 'what actually happens in history', the raw material of the historian. Thus history is myth with dates. But in other places, as for example when he speaks of several types of historical sequences - seasonal cycles, human life cycles, the cycle of exchanges when he says that 'all societies are in history and change', that there is 'a sort of fundamental antipathy between history and systems of classification' (1966b: 232, 234), and that 'cold' societies 'resist desperately any modification of their structure which would permit history to invade their midst' (1966c: 121), Levi-Strauss seems to be referring to 'what actually happens'. More often he uses a label like 'demographic evolution' or the 'contingencies of history' (1966b: 73, 1966a: 16) for the input to the historian-computer. When, with the development of Western civilization in mind, he writes (1956: 154) that cumulative history is produced by coalitions of cultures and not by cultures in isolation, presumably he refers to both input and output. When he speaks of history in contrast to anthropology he refers, as it were, to the programmes used in the computer. Though the input may be continuous, the output is not, and it is history in this latter sense, Levi-Strauss stresses, that is discontinuous. We might also add that history is discontinuous only in a certain sense, for we recall Levi-Strauss' earlier statement that 'Thus, there is, in the history of the human mind, a fundamental opposition between symbolism, with its character of discontinuity, and knowledge, characterized by continuity' (Levi-Strauss 1950: xlvii; Mendelson 1967: 126). Inasmuch as history, as part of knowledge, is more than dates, it takes on the quality of continuity. Different kinds of history, Levi-Strauss argues, may be arranged in a hierarchy, and as we move up the hierarchy we pass from 'history which teaches us more and explains less' to 'history which explains more and teaches less' (1966b: 262). This hierarchy is only sketched in, with biographical and anecdotal history at the bottom and cosmology at the top, but we should be on our guard against assuming that the various kinds of history form a strictly ordered set. Cunnison (1951, 1957) has given an admirable demonstration of both the incompatibility and
164
MODELS OF SOCIAL SCIENCE
irreducibility of various middle-range histories within one society in the Luapula valley. Levi-Strauss' statement that: 'Le propre de la pensee sauvage est d'etre intemporelle' (1962a: 348) may well apply to Australian Aboriginals, with their concept of the Dreaming which 'was, and is, every when', as Stanner (1956: 52) puts it. Yet other non-literate peoples, like those described by Cunnison, appear to delight in a plurality of axes of irreversible time. Levi-Strauss would seem to suggest that the ancestral tales told on the Luapula are historical myths and not true histories as found among literate peoples (Levi-Strauss et al. 1963: 636). But if socalled 'true' histories are really myths with dates, these legendary dateless histories of tribes and clans are better regarded as belonging to a class of myths intermediate between dated histories and interminable myths, distinguished from the latter by the primacy of the diachronic sequence (cf. Levi-Strauss 1966b: 258). Thus we see that dichotomy has, as it were, overplayed its hand. In the structure of myth we discover synchrony and diachrony in elemental contrast, and here Levi-Strauss is an exciting and adventurous guide. But the simple and dramatic contrasts between black and white, left and right, sky and earth, and the like that we usefully employ or recognize in structural models of myths and kinship systems prove to be inadequate tools when we try to discover and describe the structure of historical processes or of interdisciplinary relations. To tackle these tasks we have to distinguish several kinds of history, and several kinds of model, and though our analysis may lose dramatic simplicity, it gains, I think, in adequacy and utility. To make better models of 'what actually happens in history', or even of what people think happened, whether it be crosscousin marriage or lineage segmentation, or eventually the French Revolution, we have to consider periodic times, cumulative and noncumulative effects, and other non-binary notions. The architect of models has to call in the engineer. 5 Cutting time down to size A restatement in less pedestrian language may perhaps help to vivify, though scarcely to clarify, some of the distinctions I have been trying to make. Let us construct a scale and see where along it we can place statements by Levi-Strauss and other writers who have concerned themselves with time. The notion of time may be treated as more, or as less, like other dimensions and as requiring more, or less, special analytical attention; how time is treated determines the point on our scale. At one end of the scale is the simple slogan of Marxist eschatology: 'History is on our side'. This seems never to be Levi-Strauss' view; it is one of the senses in which he is 'not a Marxist in the ordinary sense' (Steiner 1966: 34). Not
TIME FLIES LIKE AN ARROW
165
far from this point of the scale is the position indicated by Auden (1940: 104), when he writes: And the poor in theirfirelesslodgings dropping the sheets Of the evening paper: 'Our day is our loss. 0 show us History the operator, the Organiser. Time the refreshing river.' In this view, the synchronic world of events, recorded in the field notebooks of the journalist-ethnographer, is disorderly. Only by turning to the historical progression of events can we detect the long-term diachronic order. Here we are still a long way from any position adopted by LeviStrauss. We are nearer to Radcliffe-Brown (1952, 203), who recognized social evolution as a reality and thought that it should be studied. But he too mainly sought synchronic order, not in the world of events itself but in a first-level abstraction from it. If for the moment we take a rather myopic interpretation of history, we have at this point on the continuum Fortes' position, as taken in his article on 'Time and social structure', where he contrasts the apparent absence of regularity from the contemporary scene with 'structure' seen as 'an arrangement of parts brought about by the operation, through a period of time, of principles of social organization' (Fortes 1949b: 84). Both British and French structuralists are more concerned with synchronic than with diachronic order, but the two groups aim at different kinds of synchronic models. If Auden sees order in the river of time, RadcliffeBrown does not want to jump in. As Stanner (1967) eloquently puts it: The conviction that the particular events of social life, abstracted to general features, are the facts to which all concepts and theories must be applied rested directly on a Heraclitean view of reality. A point that his critics thus never realized was that his fundamental viewpoint was historical through and through but that, no Cratylus, he merely claimed Hume's 'privilege of the sceptic' towards the possibility of making inductive generalizations about history The Heraclitean logos ... posed an essential problem: what was discoverably coherent in the social, as part of the cosmic, process within which history ran? Radcliffe-Brown found his answer, not in trying to trace the river from source to sea, but in what constrained and shaped it over steady courses to be what it was at places and times. It was not always changing in all respects at once. While Radcliffe-Brown is examining the relation between the banks and the river, Levi-Strauss turns to the water, looking, not at the continuous flow, but at the pattern of bric-a-brac floating on the surface. He writes (1966b: 232): This enables us to understand the appearance of a permanent conflict between the structural nature of the [totemic] classification and the statistical nature of its demographic basis. The classification tends to be dismantled like a palace swept away upon the flood, whose parts, through the effect of currents and stagnant
166
MODELS OF SOCIAL SCIENCE
waters, obstacles and straits, come to be combined in a manner other than that intended by the architect ... the great lesson of totemism is that the form of the structure can sometimes survive when the structure itself succumbs to events. Here, the passage of historical time contributes only disorder: 'There is thus a sort of fundamental antipathy between history and systems of classification'. The ethnographer works carefully with historical information not to establish diachronic process but to uncover the underlying synchronic structure in its various transformations (Levi-Strauss et al. 1963: 638) just as the actor's mind moves from empirical diversity to conceptual simplicity to meaningful synthesis (Levi-Strauss 1966b: 131). Levi-Strauss (1963b: 304) criticizes Radcliffe-Brown for having a 'philosophy of continuity not of discontinuity'. Since Radcliffe-Brown (1957:12) appeals to the views of Heraclitus it is fitting that, according to Lefebvre (1966), we should link Levi-Strauss with Heraclitus' adversaries in the school of Zeno. Is is just another of those historical accidents that one of the most unequivocal instances of a symbol for 'time' in all the corpus of mythology cited by Levi-Strauss (1966b: 143) is Zeno's arrow, as reported by La Flesche (another accident?) (1921: 99; 1925: 207, 234, 364) for the Osage of Missouri? Thus symbolized, time may be domesticated and tamed, can be grasped and broken. We have come near to the far end of our continuum, where time is neither the dominant dimension, nor the dimension to be excised, but rather just one more dimension like any other, perhaps not even a dimension at all. We are not quite at the end point. Much of La pensee sauvage is taken up with the 'constantly repeated battle between synchrony and diachrony' (1966b: 155), and these warriors appear as pairs of orthogonal axes in Le cru et le cuit (1964a: 95, 232). In other words, there is still something special about time. We have to deal with synchrony contrasted with diachrony and cannot simply ignore time, in the way we can often ignore colour or space. Even in Levi-Strauss' taxonomy of systems of kinship terminology, with its profusion of kinds of genealogical time - open, closed, empty, circular, curved, reversible and irreversible - we still do not treat time as analytically homologous with space, colour, volume, weight and the like. The antepenultimate step brings us to music. While we listen to music we become immortal, Levi-Strauss (1966c: 61) argues, whereas, so I suppose, when we attend a performance of Hamlet or look yet again at Charlie Chaplin in The gold rush, we are merely going on holiday, escaping to some exotic stretch of time. But when we escape musically into immortality from the familiar context of the world of events, time travels with us; we do not escape from time. There is still a beginning and an end and these are not interchangeable. The time we dally with is 'irremediably diachronic' and oriented; if we play the tape backwards our heavenly
TIME FLIES LIKE AN ARROW
167
immortality becomes hellish. We have to turn to Levi-Strauss' other comparison and study a musical score, rather than listen to a performance, really to achieve timelessness. We can confront one page with another and experience beginning and end simultaneously. For a musical score is a particular kind of graph, with time as one of its dimensions, as Levi-Strauss (1963b: 212) noted many years ago. In drawing a graph, we treat time like any other variable. This is the last step but one. We have reached the world of the mathematician and his child, the computer. Here, unless we become entangled with 'real' time (Chapter 8), we do not escape from time but rather conquer and domesticate it, so that its graph becomes a butterfly in our collection, or more likely a snake, if we follow Levi-Strauss (1966b: 91) in his (rather than the Murngin) interpretation of the graph of Arnhem Land rainfall. There is still one more step. We have to move on from music and mathematics to where time loses even its dimensional quality and the arrow of time ceases to point at all. Although music and mythology may be 'mechanisms designed to do away with time' (1966c: 61), neither can make a slave of time. Happily, linguistics shows us the way; we can reduce 'time' to a mere word, competing vainly for special treatment in a universe of words. The point is well made in the following verses inspired by a discussion of the structural differences between the two sentences 'Time flies like an arrow' and 'Fruit flies like a banana'. It is significant that this discussion (Oettinger 1966: 166-9), which has the problems of machine translation in mind, forms part of a symposium on computers. The two ends of our stylistic continuum are thus brought into alliance. Now, thin fruit flies like thunderstorms, And thin farm boys like farm girls narrow; And tax firm men like fat tax forms But timeflieslike an arrow! When tax forms tax all firm men's souls, While farm girls slim their boyfriends' flanks; That's when the murd'rous thunder rolls And thins the fruit flies' ranks. Like tossed bananas in the skies, The thin fruit flies like common yarrow; Then's the time to time the time flies Like the timeflieslike an arrow. (Schroeder 1966) There seems to be nothing comparable to this irrelevant treatment of time in any of the myths analysed by Levi-Strauss. There is a Greek myth in which time runs backward (cf. Leach 1961a: 128), but the classical Greeks had already begun that cumulative tradition of speculative thought of which we are the fortunate heirs. Maybe there is a major difference
168
MODELS OF SOCIAL SCIENCE
between the myths of tribal peoples and those of our own day, with H. G. Wells' The time machine (1895) as the watershed. If Levi-Strauss were to extend his inquiries to those neo-myths we paradoxically call science fiction, structural analysis might reveal a domain where the battle between synchrony and diachrony is transformed anew.
IO
KINSHIP STUDIES: SOME IMPRESSIONS OF THE CURRENT STATE OF PLAY
In the late 1960s I began work on a book about the study of kinship but in 1969 my move to Cambridge entailed giving up the work, at least for the time being. The uncompleted text was published in 19 71, and in the preface I indicated that I had abandoned kinship studies (Barnes 1971m: xxiv). Nevertheless several years later I was invited to contribute to a symposium on kinship systems and somewhat hesitantly accepted. The invitation gave me an opportunity to discuss some aspects of the complex, and sometimes deliberately obscured, relation between the disciplines of social anthropology and sociology. This was, I thought, appropriate for a conference held in India where this relation has taken a distinctive form. As in Chapter 1,1 used Schneider's book on American kinship (1968) as a convenient exemplar of what, in my view, was the wrong specification of this relation.
1 Changing emphases The study of kinship has long had a special place in anthropological writing and research. Morgan's Systems of consanguinity and affinity (1870) demonstrated more clearly than any other work of nineteenth-century scholarship that anthropology had a specialised corpus of data which demanded particular skills for its analysis and which could not be comprehended merely with the help of a good classical education and some practical experience of exotic peoples. In Malinowski's (1930: 20) view, 'kinship is really the most difficult subject of social anthropology'. I remember as a brash graduate student training for field research being advised, if I encountered in the field 169
I7O
MODELS OF SOCIAL SCIENCE
missionaries and colonial administrative officers who might try to overawe me with claims to superior knowledge of the culture I was sent to study, to direct the conversation to the topic of kinship where, I was assured, I should have no difficulty in outsmarting them. In the period immediately following the end of the Second World War, when in Britain social anthropologists first began to worry about the relation of their discipline to the slowly awakening discipline of sociology and the slowly declining world of hunters and gatherers and pre-literate subsistence cultivators, kinship was seen as the one aspect of human behaviour where the expertise of anthropologists was safe from challenge and which would remain available for study even in a fully industrialized world. In the conventional division of the subject-matter of social anthropology, kinship was recognized as one of the constituents of the quadrivium, along with politics, economics and religion. In the paradigmatic product of field research, the tribal monograph, a chapter on kinship was an essential component. A significant portion of the results of an international programme of field research carried out during the 1930s was published in 1950 as African systems of kinship and marriage (Radcliffe-Brown and Forde 1950), which, with its long Introduction' by Radcliffe-Brown, had a major impact on anthropological research and teaching throughout the world. Over the intervening thirty years a good deal has changed. Kinship no longer occupies so prominent a place in anthropological studies. Articles still appear in scholarly journals setting out newly studied systems of kinship, or discussing detailed aspects of systems already well known, but in recent years kinship articles, in their pure form, have become much less frequent than hitherto. Those that purport to be about kinship, pure and simple, have changed their form and content significantly. For roughly half a century after the appearance of Morgan's magnum opus, kinship, as the term was used in the technical literature, meant mainly kinship terminology. It was with this restricted connotation in mind that Malino wski (1930: 19) made his famous complaint about the iniquities of 'kinship algebra and geometry'. Under his influence other aspects of kinship began to receive more attention than did the taxonomy of terminological systems and, with this enlarged connotation, kinship remained a legitimate focus for ethnographic reports and monographs. This change in emphasis was vindicated in several outstanding studies of kinship systems. There is, however, a hint of future uncertainties in the fact that, although Malinowski announced several times that he was preparing a comprehensive monograph on Trobriand kinship, this volume never appeared. Maybe he came to realize that once the dominant nexus with terminology had been broken, kinship was no longer a satisfactory theme around which to organize an analysis of social and cultural data (cf. Fortes 1957: 161-3). This uncertainty has been displayed in recent years
KINSHIP STUDIES
IJI
both by ethnographers with fresh data they wish to present and by editors of anthropological journals. Increasingly the analytic focus of articles and monographs is not kinship as such but instead topics such as socialization, economic relations, regional systems, social control, political arrangements, systems of domination, symbolic systems and so on. In discussing almost every one of these topics with reference to non-industrial societies, substantial account has of course to be taken of relations between kin and affines, and even of the terms they use to address and refer to one another. Nevertheless the focus of the argument, the proclaimed rationale for presenting the article or monograph, is usually not kinship but something else. A monograph specifically directed towards the analysis of a single kinship system, such as Kelly's fascinating study of Etoro social structure (1977), is now somewhat unusual. Yet though monographs on single kinship systems have become less common, there has been increased publication of journal articles and books discussing kinship in general, ranging from arguments about the origin of incest to predictions of the imminent disappearance of the family. Evidence for a shift in the salience of kinship as a centre of analytic interest was provided abundantly by the programme of the tenth International Congress of Anthropological and Ethnological Sciences held in Delhi in 1978. Apart from those devoted to archaeology, physical anthropology and linguistics, and others defined regionally, there were fifty-eight sessions and symposia. Of these, only four were designated as discussions of kinship systems. Yet despite this apparent lack of interest in kinship, there were, I am sure, frequent references to many aspects of kinship in the remaining fifty-four sessions and symposia. This shift of focus of attention can be explained adequately neither by invoking the highly contingent structure of congress symposia, nor by the perverseness of journal editors anxious to follow the latest intellectual fashion. Even those who specialize in its study advertise their lack of confidence in kinship as an heuristically useful analytic category. The bluntest expression of this denigration of kinship as a category is found in Needham's (19 71: 5) remark in the symposium on Rethinking kinship and marriage that 'there is no such thing as kinship, and it follows that there can be no such thing as kinship theory'. This is indeed rethinking. Yet though kinship, as a rubric for a distinctive and isolable set of social relations, may have become less useful for the analysis offielddata, the last twenty years or so have seen the emergence in a new and more sophisticated form of what Malinowski disparagingly called kinship algebra. The study of terminological systems, drawing heavily on concepts and techniques adapted from linguistics, and sometimes making extensive use of computers for analysis, has become a new area of specialization within the wider ambit of kinship. Genealogical mathematics, in its latest manifesta-
172
MODELS OF SOCIAL SCIENCE
tions, displays a curious mixture of algebraic complexity and sociological simplicity or naivety (Ballonoff 1974a, 1974b; Barnes 1976b). This affinity with mathematics has been reciprocated, for mathematicians themselves have found in prescriptive marriage systems neat examples of algebraic groups and other logical constructs. Whereas Crow and Omaha are, or were, household names to anthropologists interested in kinship, it is Kariera and Arunda that are now familiar names to students of finite mathematics. Thus while the so-called formal analysis of kinship systems has progressed impressively from the point at which Malinowski derided it, it has also grown a long way apart from the study of 'the set of actually existing relations at a given moment of time which link together certain human beings' (Radcliffe-Brown 1940: 4).
2 Changing content Why should the study of kinship have developed in this distinctive way during the last twenty or thirty years? In a short article I can make only rough suggestions about what the answer might be. Furthermore I can make these suggestions only as an outsider, for my professional interest has strayed far from the field of kinship and my knowledge of recent studies is flagrantly threadbare. In looking for an answer we can begin by considering two phenomena already mentioned, the shift that has occurred in the range of societies available for empirical study and the increasing impact of the discipline of sociology on anthropological theory and praxis. In his Chicago doctoral dissertation of 1933, Sol Tax (1955: 445) could still discuss what he called 'social organization' largely in terms of aspects of kinship - clans, matriliny, exogamy and kinship terminology. For despite the universalistic pretensions of its theorizing, with its claim to embrace societies and cultures at all levels of technical development, the praxis of social and cultural anthropology had come to be restricted, ever since formally organized field work began at the end of the nineteenth century, to preindustrial societies. In the colonial context in which these societies were studied it was easy to assume that social relations were largely or exclusively structured in the idiom of kinship. But as card-carrying anthropologists began to turn their attention to industrial societies, and as, at a slightly later date, peasants became legitimate targets for anthropological attention, so the assumption of the de facto, if not also de jure, primacy of kinship as an organizing principle became increasingly difficult to sustain. Other principles such as economic class and bureaucratic status had equally to be taken into account. The idiom of kinship was seen to be, in some contexts, only an idiom, and even in the domestic family, the
KINSHIP STUDIES
173
sanctum of kinship values, economic, ecological and even political factors were seen to operate. While the changing content of the potential fieldwork universe, if we may call it that, has been one reason for the decline in the professional salience of the study of kinship, an equally powerful factor in altering the way in which kinship has been studied has been changes in the relation between anthropology and sociology. This is a complex topic and it would be inappropriate to attempt a full analysis here; but some aspects of this relation bear directly on the study of kinship and we must look at them. In a paper published in 1944 Raymond Firth drew a distinction between micro-sociology, the study of small groups or of small units in larger groups, which had constituted the focus of study of anthropologists, and macro-sociology which had been the province of professed sociologists (Firth 1944: 21). This interpretation of the legitimate division of intellectual labour between anthropologists and sociologists led naturally to the studies of kinship in several parts of London with which Firth was associated during the next twenty years (Firth 1956; Firth et al. 1969). A similar view, that the proper field of study for the anthropologist in industrial society was in the interstices of the social structure was reiterated by Devons and Gluckman (1964) in Gluckman's controversial symposium, Closed systems and open minds, and was supported by several contributors to the Anglo-American anthropological review conference of 1963 (cf. Barnes 1966b). However, as we can easily see with the wisdom of hindsight, interstices are difficult niches to defend, particularly if one's banner is written in the language of one's rival. If the study of interstices is micro-sociology, one can hardly complain if card-carrying sociologists seek to explore this part of their designated domain. Still seeking an area secure from sociological competition, many anthropologists in the last decade have abandoned micro-sociology in favour of culture, as clearly proclaimed in the title of Schneider's book, American kinship: a cultural account (1968). The current state of inter-disciplinary relations in this part of the battlefield, or circus ring, is well stated by Yanagisako (1978: 3) in a recent issue of the American Ethnologist, where she says that anthropologists must 'eschew replicating the work of sociologists' and claims that 'The analysis of the symbolic and meaningful aspects of kinship and other cultural domains offers us [anthropologists] one path toward the transcendence of a descriptive sociology of America'. This endeavour to establish disjoint areas of study within the kinship systems of industrial societies is strangely at odds with trends in other branches of sociology and anthropology. In the other three components of the anthropological quadrivium, economics, politics and religion, the last twenty years have seen an impressive convergence in the theoretical frameworks and areas of empirical concern of the two disciplines. In the
174
MODELS OF SOCIAL SCIENCE
Third World the monopoly of expertise sometimes attributed to anthropologists, or claimed by them, has been shattered in these areas of study both by changes that have taken place in the field and by the arrival there, or at least in the national planning office, of a host of experts belonging to other disciplines. Economists as well as anthropologists get their shoes muddy tramping through the gardens of peasants, while in the national capital the reports of political scientists, sociologists and demographers are more likely to be read by the elite than are those written by anthropologists. At the same time anthropologists who study economics and politics, even when they focus their attention on communities located far from modern centres of communication, have tended more and more to analyse their data in the light of theories developed by sociologists, economists and political scientists. Durkheim's historic position in France as the initiator of academic sociology, rather than of anthropology or ethnology, is often overlooked in these disciplinary confrontations. It is harder to overlook the clear alignment with sociology proclaimed, for example, even in the title of Fallers' Bantu bureaucracy (1956), written in the light of Max Weber's theories. A few years later a professed political scientist, David Easton, was invited to contribute an article on 'Political anthropology' (1959) to the first issue of the Biennial Review of Anthropology. It soon caught the attention of social anthropologists in many countries; yet, from the present standpoint of kinship experts, this invitation would presumably be judged as treacherous fraternization with the rival team. These tendencies towards interdisciplinary convergence in the study of politics, economics and, to a lesser extent, in the study of religion, continued throughout the ' 70s and show no sign of being reversed. Why then has there not been a similar process of convergence in the study of kinship? Part of the answer is simply that only in non-industrial societies do we find, or appear to find, elaborate kinship systems which, because of their formal complexity, constitute attractive challenges to the would-be analyst. Hence we have the development of a technical language and a body of literature which, quite appropriately and inevitably, is incomprehensible to other social scientists, and often other anthropologists, as well as to the general public. There is nothing similar to be found in sociology, as that discipline is conventionally taught. One of the more amusing of the differences between sociology and anthropology, as these disciplines are mediated to university students, is that while kinship is seen traditionally as the hardest and most arid of all branches of anthropology, its nearest analogue in sociology, the sociology of the family, is often regarded as a soft option, characterized, like rural sociology, by far too close a link between low-level theory and instant practice (cf. Anderson 1971: 8-13). The lack of formally elaborated kinship systems in the societies conventionally used as examples in sociology teaching provides only part of the
KINSHIP STUDIES
175
answer. In part the failure of the two disciplines to converge in the analysis of kinship can be explained by reference to the special characteristics of an intellectual position towards which the study of other aspects of social life seem to be converging. While by no means the only common framework of discourse, Marxism certainly constitutes one of the most vigorous attempts to bring together into a single scheme the analysis of systems of economics, politics and religion as ideology all over the world and throughout history. There is a good deal of truth in Gouldner's remark (1974; cf. 1975-6: 9) that Marxism is the Latin of the twentieth century. But Marxism provides a framework for discussion that is more developed with reference to some historical periods than others, and to some aspects of social life than others. Marxist casuistry and exegetical scholarship have reached impressive heights, and the available varieties of Marxism are legion. Nevertheless I think it remains true that Marxism's strengths lie in the analysis of capitalism rather than of subsistence production, and in the study of infrastructure rather than of superstructure. I think also that Marxism contributes more to the understanding of similarities in social and cultural life, and less to the elucidation of differences. For these reasons, partly empirical and partly theoretical, the intellectual convergence which characterizes a good deal of the customary praxis of anthropological research and analysis has not occurred as much, or in quite the same form, in the study of kinship. Yet this state of affairs seems quite paradoxical when we recall that one of the first thinkers to take Lewis Henry Morgan seriously, and the first to popularize his views on kinship, was none other than Friedrich Engels. If the first anthropologist to make a substantial impact on Marxism was the very founder of kinship studies himself, is it not odd that it is this branch of anthropological enquiry which has remained, or has tried to remain, unsullied by other social sciences where Marxism has had a substantial influence? The reason for this paradox is to be found partly in the enthusiastic but, in retrospect, uncritical acceptance by Engels of Morgan's conjectural history and in the incorporation of Engels' Origin of the family (1884) into the canon of an orthodox Marxism which was much more fundamentalist than it usually is today. When many early twentiethcentury anthropologists rejected nineteenth-century schemes of social evolution, they also rejected the then unpopular revolutionary theory which they identified with one of these schemes (Llobera 1979: 256). Some sixty years later, anthropologists are more ready to look sympathetically at evolutionary schemes; but the concentration of Marxist attention on the broad similarities of systems of domestic exploitation throughout history and throughout the world has weakened its attraction for those scholars whose chief delight in the study of kinship derives from its complexities and diversities rather than its uniformities. Yet not all
I76
MODELS OF SOCIAL SCIENCE
anthropologists scorn the search for broad similarities, not even all specialists in kinship. It is significant that many of those whose current focus of scholarly interest, and often of practical action as well, is the ubiquitous subjection of women have found in Marxism in general, and in Engels in particular, powerful and authoritative support (cf. D. H. J. Morgan 1975: ch. 5 & 6). Even the myth of matriarchy has been brought out of the nineteenth-century cupboard and thrown into the fray (FluehrLobban 1979). Sociology is not the only discipline whose advances some anthropologists have spurned; the same coyness has been displayed towards biology. But whereas rivalry and cooperation with sociology in relation to various branches of anthropology have been as much over empirical research fields as over theoretical concerns, the arguments about the appropriate relation between social anthropology and biology have been mainly confined to theory. These arguments have impinged particularly on kinship studies because of the ostensible similarity between genetic connexion and consanguineal kin relationship. During the quarter century from 1940 onwards, in their efforts to free their discipline from its traditional association with archaeology and physical anthropology and, incidentally, to associate it more closely with sociology (Firth 1944: 22), some (mainly British) social anthropologists strove to deny that the presence or absence of genetic connexion had any relevance for the understanding of kin relations. Durkheim's plea for the autonomy of social facts was cited in support of this view. Scholastic arguments, partly a byproduct of the wider debate for and against positivism, were advanced about the basis and extent of the disjunction between social kinship and genetic links (cf. Gellner 1963; Schneider 1972; Barnes 1974a), and about the validity of maintaining a distinction between notions of procreation held by actors and those founded on scientific inquiry. The debate took a more vigorous turn with the widespread publicity given to Wilson's Sociobiology, published in 1975 and comprehensively criticized by, among others, Sahlins in the following year. Luckily, on this front there seems to be more diversity of opinion within the ranks of anthropology, a greater readiness to borrow and incorporate the findings of other disciplines, and less preoccupation with maintaining sacred and impermeable interdisciplinary boundaries. The emerging consensus seems to favour some recognition of a casual link between the so-called selfishness of genes and widespread, even if not universal, features of social relations between kin, even though, as in Fortes' (1969) work, these features are better described in moral and cultural rather than in instinctive terms. So far, however, biology, like Marxism, seems to be more successful at explaining similarities rather than at elucidating differences. In a recent article Barkow (1978: 100) proclaims that
KINSHIP STUDIES
177
Sociobiology must account for the diversity and malleability of human social behaviour, rather than content itself with pointing out its limits and central tendencies. This task has still to be achieved, and we should not assume in advance that it will prove to be possible. 3 Changing disciplinary boundaries The current state of kinship studies may therefore be summarized thus. The study of kinship terminologies and of models has become a sophisticated specialism largely divorced from consideration of the substantive content of relations between individuals. The substance of relations between kin has come to be analysed in conjunction with other systems of relations between individuals not tied to kinship, and with relations with the environment. The symbolism of kinship offers a field for those who see a division of labour between sociologists studying social action and anthropologists studying culture. The movement for the liberation of women has stimulated the analysis of domestic relations in political terms, and it is here that the impact of Marxism has been greatest. Aspiring biological reductionists have prodded anthropologists into thinking more constructively about what, if anything, is meant by the biological basis of kinship. If this analysis of the present position in kinship studies is correct, there are two important metatheoretical issues to be faced. First there is the preoccupation with disciplinary boundaries. This concern for intellectual territorial rights is easy to explain in terms of the sociology of academic life. Nevertheless it is, in my view, a perversion of the proper aim of social science, which is to increase our understanding of social life without worrying about the purity of the intellectual origins of the insights used in this task. Over ten years ago Uberoi argued that 'For us the important considerations should be to define our relevant problems and pursue their solutions, not worrying about the sources of ideas' (Abbi & Saberwal 1969: 192). Uberoi had in mind the national origins of ideas but his remark applies equally forcefully to their origins in different disciplines. Interdisciplinary intellectual cooperation affects all aspects of kinship; the other metascientific issue has particular reference to the relation between kinship and biology. The two extreme theses, that social kinship has nothing whatever to do with genetic connexion, and that social kinship is merely an epiphenomenon of genetic programming, are equally false. The compromise thesis, that the genes have some influence on the symbols and social relations of kinship but do not uniquely determine them, is certainly true but merely vacuously so. When we begin to try to specify the extent of this influence and to discover the mechanisms by
I78
MODELS OF SOCIAL SCIENCE
which it operates, it is only then that there is anything to argue about. The process of making our propositions falsifiable will almost inevitably entail making them quantified, for only when there is at least some kind of ordinal scale linking two opposed polar positions does it become possible to describe an intermediate position that can be defended intellectually. The task of constructing suitable scales that make social, as well as algebraic, sense is bound to be difficult but it also essential if we are to avoid an endless and unenlightening swing of the pendulum of intellectual fashion from one extreme to the other. To prolong the life of analytic categories that have outgrown their usefulness would be irrational, however important they may have been in the past and however much professional intellectual capital may still be invested in them. If there is no longer any such thing as kinship, we should look for other fields to conquer. I believe that the concept of kinship still has plenty of heuristic utility. The present disposition of intellectual effort in kinship studies does, however, leave largely unexplored important areas of human social activity of great theoretical and practical interest. To concentrate on the study of symbols as a way of transcending the descriptive sociology of kinship, as Yanagisako recommends, is to abandon untouched the analysis of actual sets of relations between individuals, some of them mediated by kinship and some legitimated in other ways. In order to understand how kinship systems operate in practice we need to study not only the symbolic and cultural systems through which actors perceive their situations and order their goals, and not only the enduring institutional arrangements such as marriage and the household which presumably form the subject-matter of what Yanagisako calls descriptive sociology. We need also to watch these systems in action, to study tactics and strategy, not merely the rules of the game, and to analyse why and how one individual is mobilized on one occasion and not on another, why and how one activity is carried through on one occasion by reference to kinship and on another by reference to some other system of relations. We need to look at how one family differs in its internal constitution and external relations from the family living next door and how it changes in the course of its cycle of development. We need to study individual differences as well as community-wide similarities. This is the field which Firth identified as micro-sociology and it is too interesting and important to be neglected, whatever may be the disciplinary banners borne by those who cultivate it. In fact the field of micro-sociology is an area of study which those who approach it from anthropology are, in my view, particularly well equipped to tackle. Anthropologists are more likely to have access to the currently somewhat esoteric investigative techniques of the componential analysts and genealogical mathematicians than are those scholars who approach
KINSHIP STUDIES
179
micro-sociology from sociology, whether mainline or Marxist. The formal analysis of kinship terminologies and of genealogies is one of the many fields in the social sciences where increased sophistication of technique seems to have led its practitioners to become more interested in the techniques themselves than in the real world they were originally designed to elucidate (cf. Barnes 1972b, 1978c). Maybe it is only in this kind of cloistered intellectual environment that sophisticated techniques of analysis can be developed. But once developed, there is no reason why they should be used only for solving artificially constructed intellectual conundrums and for inventing mathematical games. The micro-sociology of kinship is a field of study where early attempts at systematic analysis were not particularly enlightening. The number of possible choices between which actors have to select is enormous, so that the analyst who lacks sophisticated mathematical tools can do little more than count heads, categorize households and present a few apt anecdotes. The tools of analysis now being used by the terminologists and genealogists may require some modification before they can be used for a slightly different purpose, but something like them seems both necessary and possible for micro-sociology. Other useful tools come from recent developments in the analysis of social networks, where at last readily usable computer programmes are becoming available which deal satisfactorily with hitherto formidable problems of sampling and boundaries. It is tools such as these which, in my view, hold out the best prospects for the analysis of social life at the level of interpersonal relations. Some programmatic proposals for empirical research of this kind in traditional areas of anthropological fieldwork have already been made (e.g. Jackson 1976: 81) but the hard work of implementing them still lies largely ahead (cf. Foster 1979). One of the most influential publications in the field of kinship since the end of the Second World War has been Levi-Strauss' massive Les structures elementaires de le parente (1949). In this book he made a strong plea for concentrating on the study of what he called structure as against the study of events. During the last thirty years the study of structures, and of kinship structures in particular, has been most rewarding. But partly because Levi-Strauss posed the contrast in these terms, the structures studied have been mainly the structures of non-events, of norms and symbols. Events too have their inner logic, their regularities and trends, even if individual events remain inherently fortuitous. To neglect totally either symbols or events is perilous; we must keep both in sight. Yet science cannot afford to forgo the help of enthusiasm even if enthusiasts, at least for a decade or two at a time, narrow their sights recklessly. The singleminded pursuit of symbolic structures has gained us a better staging post from which, if we can match the enthusiasm of the first generation of structuralists, we can now advance on a different tack. In turning anew to
I8O
MODELS OF SOCIAL SCIENCE
the study of events, we can hope to transcend the particularism of Vhistoire evenementielle, and to grasp their structures, only if we make use not of the bastard mathematics which Malinowski so disliked but of newly developed mathematical tools which have already shown their utility in one restricted corner of the study of kinship, and which can be adapted to tackle tantalizing but intractable problems in other parts of that study. With luck, Firth's micro-sociology and Levi-Strauss' structuralism can form a new synthesis.
II
SOCIOLOGY IN CAMBRIDGE: AN INAUGURAL LECTURE
In Chapter 8 my discussion was directed at the differences between natural and social sciences; in this chapter I try to distinguish between the social sciences and the humanities. The University of Cambridge, in which this chapter was delivered as my inaugural lecture, had delayed so long in first appointing a professor of sociology that by the time I arrived in 1969 not only had the funds for new developments begun to dry up but also the period of world-wide student unrest had begun. Sociology students are recognized as leaders of this unrest. The decision to expand the teaching of social sciences in the university had been reached only in the face of widespread hostility, and even some of the supporters of sociology opposed the move because they feared that by treating sociology as a distinct discipline the teaching of related subjects such as history would become less sociological. In these circumstances I thought that in my inaugural I should make as strong a case as possible for sociology being a science rather than one of the humanities, but also a distinctively social, and not a natural, science. I also tried to emphasize that sociology is inherently an iconoclastic social science, and that its students must be expected to be iconoclasts. Somewhat unfairly I took the Cambridge philosopher and reformer Henry Sidgwick as my anti-hero, mainly so that I could show that the university, notorious for its caution about accepting new disciplines into its curriculum, had already had ample time to reconcile itself to the advent of sociology. When a reviewer complained of the tameness of the lecture, and remarked that 'it will hardly set the Cam on fire' (Cohen 1971: 722), I felt that I should reply that to 181
l82
MODELS OF SOCIAL SCIENCE
do so had not been my intention; I had merely hoped to explain to the university why its sociology students might try to burn its buildings. Many inaugural lectures conform to a simple paradigm. The new incumbent of a chair begins by paying tribute to the illustrious pioneers who have preceded him and then goes on to demonstrate how his discipline has advanced to a level of sophistication and complexity far beyond anything they could ever have imagined. Mercifully I am spared the temptation to follow this arrogant model. Today I do not have to mark the beginning of a new tenure of an established chair in a recognized discipline. Instead I find myself with the heavy responsibility of justifying a new chair in a discipline which in this university is still striving for full recognition. In these circumstances it seems appropriate to discuss some of the distinctive features of sociology in very broad terms but with special reference to its place in the academic tradition of Cambridge. There are two salient features which distinguish the state of sociology in Cambridge from the situation found in many other major universities. First, the decision to establish a chair of sociology was taken only very recently; and secondly, there are many members of the university, junior as well as senior, who do not regard sociology as a proper academic subject. These two facts are not merely domestic blemishes which it would be tactful to ignore on a euphoric occasion such as this. On the contrary, I am convinced that these local characteristics have to be analysed and set in context if we are to understand what sociology is about and how, or whether, it can fit into the intellectual life of this university. At first glance it may well seem odd that the establishment of a chair of sociology in this university was so long delayed. Whatever may be its pedagogic status, sociology clearly lies at the centre rather than on the periphery of social science. This university has a glorious tradition of empirical scientific inquiry, though it is true that this is predominantly in the natural rather than in the social sciences. Sociology began to be taught in some American universities in the 1880s, and here in Britain Hobhouse was appointed to the first chair of sociology at the London School of Economics in 1907. Sociology in Britain grew out of a fusion of two streams of thought and activity, one being concerned with the hard facts of poverty, housing and wages and with social action that would lead to material improvement. Even if this stream, a very direct outcome of Victorian capitalism, did not flow anywhere near nineteenth-century Cambridge, the other stream, exemplified in the grand theories of social evolution advanced by Herbert Spencer, was surely quite close to the intellectual interests of the university (cf. Abrams 1968). But although Darwin acknowledged his intellectual debt to Spencer, it seems that Victorian Cambridge was guided by Henry Sidgwick in its
SOCIOLOGY IN CAMBRIDGE
183
attitude towards the claims of sociology. In 1885 Sidgwick (1904: 198, 257) took the view that general sociology was in a Very rudimentary condition', and that though an established science of sociology was to be hoped for, it must wait until it had 'succeeded in establishing on the basis of a really scientific induction its forecasts of social evolution' before meriting professional recognition. The process of establishing a genuinely scientific evolutionary sociology, he seems to have implied, must take place somewhere outside Cambridge. Thus it was left to other universities to offer shelter to this nascent science and to assist it to realize its potentialities. Now that sociology has appeared in the very part of Cambridge that commemorates Sidgwick's memory, we may well ask whether significant progress has been made towards solid knowledge, or whether sociology earns its place in the university on other grounds. For despite Sidgwick's scepticism towards sociology, the study of society in one academic guise or another has had a long history in this university. I need mention only the work of Sir Henry Maine who, although several of his classic studies of the development of legal systems were written elsewhere, began and ended his professional career here in nineteenthcentury Cambridge. A little later Haddon organized the Cambridge anthropological expedition to Torres Straits, and his colleague W. H. R. Rivers took up the comparative study of kinship systems. Sidgwick's great pupil, Sir Frederick Maitland, despite his disapproval of mono-causal sociological explanations (Maitland 1911: 285-303), joined with Pollock to write what is indubitably a sociologically informed history of English law. Economics, under the inspiration of Marshall, early acquired that uneasy and imperfect correspondence between empirical data and theoretical propositions which is characteristic of all social science. Under Haddon the study of overseas societies blended speculation about origins and social evolution with the practical demands of colonialism. Thus although the claims of some of the practitioners of sociology may still seem as pretentious as those that displeased Sidgwick, the disipline has many close cognates well established in the university and has no reason to feel out of place. Yet there are differences between one social science and another, as Sidgwick himself saw when he advocated in place of sociology what he called 'more limited and empirical studies of society in as scientific a manner as possible' (Sidgwick 1904: 198). His misgivings seem to linger on. When, some eighteen months ago, members of the Regent House were called to vote on the proposal to establish the new Social and Political Sciences Tripos in which sociology has a prominent place, it seems that many of those who voted non placet were opposed not to the social sciences as such but only to sociology. Sociology, they asserted, is not an ordinary academic discipline. Paradoxical though it may seem, I firmly agree with
184
MODELS OF SOCIAL SCIENCE
their diagnosis, though I reject the implication they drew from it. My thesis is that these opponents of the new Tripos were quite right, mainly because there is much more to sociology than just those 'limited and empirical studies' advocated by Sidgwick. I regret the timidity of Sidgwick who abandoned to other places of learning the difficult but vital task of developing a comprehensive science of society, just as I regret the reluctance of many of my colleagues to abandon what I regard as an obsolete model of a proper academic discipline. But perhaps they, like Sidgwick before them, were more perspicacious than many of those who voted placet, in recognizing that here in sociology was something new which could not be fitted easily into the existing framework of academic assumptions and intellectual concerns. My argument, of course, is that this framework should change. What is it that makes sociology seem such an uneasy bedfellow with traditional academic disciplines? I think the answer to this question lies in the relation of sociology to its subject matter. For when we consider how the various disciplines go about their tasks, it becomes clear that sociology stands in a relation to the content of its studies that is quite different from that which prevails in the natural sciences, or in the arts, or even in some social sciences. It is this relationship which marks sociology off from even so closely linked a discipline as economics and, in the earlier epoch of colonialism, from social anthropology. What are these differences? Let me begin with the simplest contrast, that between sociology and the natural sciences. In the natural sciences there is a clear division between the scientist and the things he studies. Typically the natural scientist looks down his microscope at the object on the slide, but the object, whether it be animal, vegetable or mineral, does not look up to observe the scientist. Of course it is true, and familiar even to readers of popular science, that any act of observation has some effect on the object observed, and that sometimes this effect may be significant. Nevertheless we cannot hide behind Heisenberg's skirts to escape the fact that, by and large, empirical inquiry in the natural sciences is essentially an asymmetrical process, in which the roles of observer and observed do not overlap, and in which there is a flow of information in one way only. The scientist collects the raw data of observation and translates them into a language of science which he himself has invented and which he is free to modify without danger of protest from the atoms or algae or stars or whatever they are that generate his data. The motions of the planets continue unchanged, so we assume, regardless of whether Ptolemy or Newton or Hoyle reigns supreme on earth, while closer home the zoologist's monkeys gambol and frolic untroubled about whether their actions are interpreted in ecological or ethological terms. There is the language of science, and there are the objects of inquiry, and however fiercely the language may
SOCIOLOGY IN CAMBRIDGE
185
select and distort the data, objects and observers never meet on equal terms. At the other end of the scale, among the humanities, the position is quite different. Here the aim of the exercise, so it seems to me, is to learn the language of whatever group or period or place is being studied, using language in a very broad cultural sense. The humanist seeks to understand the civilization of the Renaissance, or the world of Jane Austen, or the nature of Tudor sovereignty, or whatever, not primarily by translating the concepts, beliefs and actions under scrutiny into some trans-cultural scientific meta-langauge but rather by entering, as far as he can, into the thoughts and perceptions of the actors and by trying to see the world through their eyes. It is of course true that when as humanists we deal with the present there is the possibility of direct contact and communication between those who study and those who are being studied, but even then the people we study are clearly dominant in the exchange. It is they, and not us, who define and elaborate the language that we have to learn. Both the natural sciences and the humanities seem to me tofiteasily into the Cambridge academic tradition. With each type of study, each in its own way, there is a simple one-way relation between the data and the person explaining them. It is possible to develop and maintain a tradition of scientific objectivity and of exact scholarship, for the data are not only given, they are autonomous. The scientist tries to understand the significance of publicly accessible facts that are generated largely independently of his actions as an observer. If it happens that his actions do have some influence on the empirical data, this is a complication that has to be minimized, not an opportunity to be exploited. Likewise the humanist seeks to understand history emphatically, rather than to create it, and I think the same may be said of the humanistic study of literature and art. Thus on the one hand the tradition of natural science is based on the assumption of a fundamental separation between the facts under scrutiny, and the scientist who observes them. The scientist, in his professional capacity, is further separated from his own private life as a citizen, his hopes and fears, his social and political views and so on. On the other hand the humanist embraces the whole man and promotes, I think, a sense of identity with some established view of the world, either contemporary or from the past, rather than any strong concern to create or to change. Both natural scientist and humanist seem to fit well into an intellectual environment where continuity through time is valued, and where a recognizable separation between the university and its immediate social environment is considered appropriate. When we turn from the natural sciences and humanities to scrutinize the social sciences, we find a more complicated situation. The social scientist has to combine the tasks of the natural scientist and the humanist
186
MODELS OF SOCIAL SCIENCE
and in so doing hefindshimself facing problems that neither of the other two have to deal with. Unlike the natural scientist, the social scientist studies objects that can talk. Atoms do not talk to each other, whereas ordinary people, and not only social scientists, have languages of their own, ideas about truth and falsity and right and wrong and about the nature of the universe. But unlike the humanist, the social scientist is not content merely to learn the language of the people he studies and to enter vicariously into their perceptions and beliefs. It is certainly an important part of his task to learn this language and to see the world through the eyes of the people whose beliefs, customs, values, sentiments and actions he is trying to understand. Without looking carefully through these culture-tinted spectacles he cannot hope to understand very much. This task he shares with the humanist. But this is only part of what he has to do. Like the natural scientist he also has a scientific analytical language which he has built up for his own use, and into which he endeavours to translate his empirical data. Patterns of belief, customs and social institutions, and sequences of action have to be shown to make sense to the actors he is studying, but they have also to make sense in terms of a body of theory which, to a greater or lesser degree, transcends the axioms of any particular social and cultural pattern. The social scientist has to handle the language of science as well as the vernaculars spoken in the field. In some instances this procedure, although more complicated than that followed by either the natural scientist or the humanist, presents no insuperable difficulties. This is the case where there is effective insulation between the field situation and the academic world of social science in which thefielddata are sorted, translated, analysed and discussed, and finally published. These conditions used to be met when sociological fieldwork was carried out among primitive peoples living under colonial conditions. Thus for example the inquiries made by Haddon and others on their famous expedition to Torres Straits in 1898 were carried on as if the islands of the Straits were laboratories that could well have been part of another planet rather than the home of fellow subjects of Queen Victoria. The islanders had no say in the arrival of the expedition. Publication of the results of thefieldworkwas not completed until thirty-seven years after the visit of the expedition (Cambridge Anthropological Expedition 1901-35), and these results were presented to the world of learning in Cambridge and other university centres rather than to the islanders themselves. Haddon, Rivers and their companions on the expedition were, no doubt, not entirely dispassionate about the scientific interpretation of thefielddata whether or not the islanders had a High God, whether their system of totemism was based on an archaic dual organization, and so on - but their commitment to one interpretation rather than another referred only to their stance as scholars and not to their position as citizens here in Britain.
SOCIOLOGY IN CAMBRIDGE
187
The spatial and cultural insulation which under colonial conditions so conveniently separated the field situation from the world of learning has now largely disappeared. The reasons for this change need not concern us here; the sorry story of 'Project Camelot' shows clearly enough how greatly the situation has altered since the expedition to Torres Straits (cf. Horowitz 1967). My point is simply that on the home front, where sociologists have concerned themselves with the study of their own societies, this insulation has never existed at all. For this reason sociology has been able to conform to the model neither of a natural science nor of the humanities. Instead, right from its inception in the early nineteenth century, it has had to develop a dialectical relation to its object of study, a task which, with the disappearance of colonial conditions, now faces social anthropology as well. For simplicity of exposition, let me stick to a discussion of language. In Western civilization there is, on the one hand, a debate about the nature of society which has lasted for centuries and which, although conducted by philosophers and other scholars, has by and large been carried on in the language of everyday life rather than in a technical vocabularly or in mathematical symbols. On the other hand, many philosophical and theological concepts, some of great subtlety, have passed over, in name if not always in content, into the language of the marketplace, the political arena and the home. The nineteenth-century sociologists and other social scientists who began more self-consciously to attempt to construct a science of society continued to use the language of ordinary speech. In this they were unlike the natural scientists who, as their theories and observational procedures developed, were forced to give specialized meanings to more adopted or invented words. Nineteenthcentury sociologists also differed from anthropologists who had to develop a specialized vocabulary to describe exotic societies and cultures. There was much less initial impetus in sociology to construct a special technical language. When in the twentieth century sociologists, professional and amateur, became sufficiently numerous to develop a private language of their own, it could not remain entirely private. The same process of continuous interchange between social scientists, other scholars and the lay world continued as before, with the same transformations and distortions. Indeed, despite the increasing professionalism of the social sciences, the interchanges have, I think, increased as the lay view of society has come to include the social sciences themselves as part of the accepted perceived pattern. The market-research girl and the interviewer in depth have become familiar figures and their activities are taken for granted, just as among some erstwhile primitive tribes the presence of a visiting anthropologist is said to be taken for granted. We have reached the stage at which
188
MODELS OF SOCIAL SCIENCE
the interaction between society and social science has undergone a qualitative change, as exemplified by the layman who, when asked how he would vote in the recent election, replied, 'I don't know, I haven't seen the result of the opinion polls'. Hence it is no surprise to find that terms like 'class', 'prestige', 'crime' and 'race' have a whole range of meanings, some technical and scientific, others lay and popular, the majority a mixture of both. Confusion is made greater partly because there is no single orthodox scheme of analysis in social science, but only a variety of rival schools whose analytical terminologies overlap, and partly because the ambient society itself is characterized by cultural heterogeneity and diversity. Under these conditions the colonial model of empirical inquiry, with one-way translation from the vernacular into the language of science, becomes quite inapplicable. Suppose, for example, that we try to analyse the notion of social class held by residents of a London suburb. We are likely to find that though some of their ideas about class are derived from direct experience and from traditionally transmitted ideas, other ideas derive, at one or more remove, from the writings of sociologists about class. Even an ugly invention like the term 'socio-economic status' may appear not only on the would-be value-free pages of a sociology textbook but also with feeling and resentment on the lips of a suburban housewife. We can then no longer regard as separate the language in which 'they' describe their world and the language in which 'we' analyse what they do and think, for we and they are to some extent the same people, speaking the same language. Because sociological concepts are often derived from ordinary speech, sociologists often disagree among themselves about how they should be defined; and even when they do agree, the scientific definitions are continually being obscured by the secondary meanings these terms acquire when they are taken over, or taken back, into ordinary speech. As Cohen (1969: 228) says of the study of politics, 'The very concepts and categories of thought which political scientists employ in their analysis are themselves part of the very political ideology which they try to understand'. It is here, I think, that social anthropologists have a special contribution to make towards sociological understanding. The chief value of an anthropological approach to sociological questions lies, in my view, not in the intensive analysis of multiplex relations, as has sometimes been suggested (Banton 1964), but rather in the clearer identification of the contrasts and connexions between the concepts used by actors and by observers, between emic concepts and etic concepts, as they are sometimes called. This simple distinction which anthropologists have adopted from structural linguistics (Sturtevant 1964: 101-3; Pike 1967: 37-72, 145) needs elaboration before it can be applied outside the colonial situation,
SOCIOLOGY IN CAMBRIDGE
189
but it is only through some analytic opposition of this kind that social science can maintain itself as something distinct from popular culture and yet not be entirely isolated from it in an impenetrable jungle of professional gobbledegook. Using this distinction between the perceptions of actors and observers we should be able to cut through much of the present debate, simultaneously philosophical and ideological, about how we come to possess knowledge of the social world and whether in the explanation of social behaviour there is such a thing as objectivity. This debate is important, for the foundations of social science should at least be exposed, even if they cannot always be entirely secure. Yet the main task of social science is not to contemplate its foundations but to discover what goes on in the world. The battle for precision and clarity in sociological analysis has continually to be fought but, for good sociological reasons, it is continually being lost. It is in this connexion that mathematical methods of analysis are particularly important, not only because quantitative indices readily provide greater precision (even though we must beware of false precision), but also because concepts which are labelled with symbols rather than words are easier to protect from unwanted connotations and secondary meanings. The ordinary world of everyday life is itself becoming increasingly quantified, and scientism, the magic of numbers and percentages and even tests of significance, forms an increasing part of popular culture. Mystification by pseudo-science, by the dazzle of numerical indices and algebraic symbols, is widespread. As yet, despite valiant efforts (e.g. Coleman 1964), the constructive use of rigorous mathematical methods in social analysis is still rudimentary, though the misuse of statistical and similar techniques is enormous. It is essential that the social scientist gains a thorough mastery of quantitative methods, partly in the hope that he may be able to make good use of them, but mainly that he may escape from being mesmerized by them and may be able to keep them in their proper subordinate place. The use of numbers and symbols in social science is often necessary and is then to be applauded. Yet we stand a greater chance of being understood if we use simple words whenever possible. For convenience, I have talked about the connexions between sociology and the ambient society in terms of language. If the complications that arise affected only language, perhaps they would not matter much. But what I have said about words applies in much the same way to the values, attitudes and assumptions for which the words are labels. If we study the social institutions and cultural values of a remote group of people it is relatively easy to dissect them in the cold light of science. When we turn to the study of our own society, this task becomes harder and its consequences more serious. We have to deal with institutions in which we
I9O
MODELS OF SOCIAL SCIENCE
ourselves are involved as citizens as much as social scientists, and to which we are likely to be strongly committed, emotionally or intellectually, either positively or negatively. It then becomes harder to achieve anything that we may call scientific objectivity, and as our work proceeds our degree of commitment is likely to alter one way or the other. I have already stressed that we cannot simply follow the humanist procedure and seek merely to understand, to feel in our bones, for example, what the world seems like through the eyes of the middle-class suburban housewife. Our aim includes understanding how the world seems to her, but we have also to translate her point of view and her actions into some cross-cultural language so that she may be compared with her working-class sisters and her cousins in other countries. This procedure of comparison may be simple to perform when, for instance, we are engaged in comparing forms of inheritance among a group of African tribes or varieties of subsection systems in Aboriginal Australia. But when the social institutions of our own society form part of the comparison, the task is more complex, for part of our task is then self-scrutiny. The consequences of sociological comparison are also more far-reaching. The necessity to translate and to analyse entails the continual acknowledgement that there is nothing invincibly unique or incomparable about the particular configurations of actions and ideas that are found in any locality or at any point of time, our own place and time included, for other solutions with alternative consequences are always conceivable. Thus built into the analytical and comparative procedures of social science is the notion of scepticism and hence of criticism. Nothing is sacred, least of all the sacred itself. Thus it is no accident that sociologists are often seen by others as inherently disturbing the status quo. Although a fuller awareness of the way society works may well lead to a renewal of support on a firmer basis for many of the customary features of a given society, the initial effort of scrutiny and analysis is bound to be unsettling. The support given by the perceptive social scientist to the imperfect institutions of his own society is bound to be more restrained and more circumspect, even if it is more firmly based, than the spontaneous and naive support given by an ordinary citizen who has never questioned the assumption that his country, or religion, or university is the best in the world. In much the same way the sophisticated theologian, however orthodox his views, must disturb the simple faith of the believer who has never known any doubts. But if sociology is thus perceived by members of the wider public as a potential threat to deeply held values and cherished beliefs, in a narrower academic context it may be seen more as a waste of time rather than as a threat. Shils (1967: 88) notes that sociology is disturbing because it is 'parochial about the universal', but equally disturbing to many academics is the fact that sociology is apparently unnecessarily pretentious about the
SOCIOLOGY IN CAMBRIDGE
commonplace. The natural scientist, if he sticks to his last, is immune from lay criticism and the humanist likewise, but the social scientist who studies his own society must expect criticism from every quarter. His friends and neighbours, including his colleagues in other disciplines, will all assume that they can criticize as experts, for they too have views on social class, on how the social services work, on how much social mobility there is, and so on. It would be a strange society indeed whose members did not hold firm convictions about its salient characteristics, whether or not any evidence is available. Every anthropologist knows how embarrassing it is when a member of some distant tribe writes to say: 'Your analysis is wrong. This is my society, not yours, so I should know best'. But when as sociologists we write about our own society we have potential critics all round us. The notion of sociology as a social science with its own hard concepts clearly defined transculturally and with its own corpus of tested and interrelated propositions along the lines of the natural science paradigm is of course the hallmark of Comte and the positivist tradition whose extravagant claims so offended Sidgwick. During the last ten years or so many younger sociologists have rejected the apparently unproductive search for universally valid generalizations, for social laws analogous to the laws of gravity and of thermodynamics in the natural sciences. Indeed, 'positivism' has become a derogatory term in a way that would have met with Sidgwick's approval. But whereas Sidgwick merely protested against the over-ambitious pretensions of nineteenth-century sociologists and looked forward to the distant day when their claims to have discovered the laws of social development might be validated, many present-day critics of positivism reject this scientific quest entirely and seek to promote a sociology that is particularist rather than generalizing, that emphasizes the emic categories of the actor rather than the etic categories of the outside observer, and that stresses the social function of criticism rather than the scientific function of explanation. It seems to me important that we should not abandon the search for a fuller understanding of the nature of order and disorder in social life merely because the natural science paradigm used by the early positivists and by most later mainline sociologists has proved to be inadequate and largely unrewarding. Without some effort to adapt the rigorous methods of scientific testing and validation to the more difficult field of social life, we have to fall back on traditional wisdom, common sense and rules of thumb, on social technology, as it were, rather than social science. Traditional wisdom, whether in social affairs or in the world of nature, may be quite adequate when the environment is stable and when there are untold generations of trial and error in which to arrive at a well-balanced society. For good or ill, these palaeolithic conditions have long ago disappeared. In the present crisis, we have to grapple with the social
192
MODELS OF SOCIAL SCIENCE
environment in the same determined spirit as our forebears tackled the physical environment. We cannot afford to trust our future to the automatic processes of social evolution, and instead must use our brains to secure our survival. At the same time it is equally important to develop the critical function of sociology. The special position of the sociologist, half inside and half outside his own society, does not necessarily make him wiser or saner. It provides the possibility for him to take a more informed and more farseeing view of the social scene than those of his fellows who have not been trained to take a detached view of their own social institutions and who lack the opportunity to compare these carefully with alternative arrangements found in other societies or advocated by social reformers and theorists. Sociology has sometimes been described as the study of the unintended consequences of social action, and no intended consequences are more hazardous than those which follow well-intentioned reforms and radical changes. The formulation and testing of theory does not preclude criticism, nor do criticism and analysis preclude the generation of wellgrounded theory. On the contrary, all are essential if the interchange between social science and society is to continue fruitfully for the benefit of both. With this perspective we can reject as obsolete the perception of sociology held by Sidgwick (1904: 251) as 'an attempt to make the study of history scientific by applying to it conceptions derived from biology'. Instead we can adopt the view that sociology is concerned with the regularities and lack of regularity in social institutions, and that there is a two-way or dialectical relation between the conceptual apparatus of the sociologist and the world view of the people whose actions, sentiments and beliefs he seeks to understand. Sociology is, as Levi-Strauss (1963b: 3 6 2 3) says, the study of societies from within, but it is a study in which the observers are continually striving to escape beyond the perspective of their own society, and never quite managing to do so. If this is so, it is not surprising that students and scholars are attracted to the study of sociology for a diversity of reasons. There are some who wish to engage in the elusive and difficult search for solid knowledge, either by studying the findings of others or by themselves engaging in research. Others are interested in applying the findings of sociology, tentative and inconclusive though these often are, either in restricted contexts like personnel management or social work, or to the solution of wider social and political problems, many of which simply will not wait for well-tested solutions. Others wish to acquire a more sophisticated, informed and sensitive view of the social world in which they live, whatever occupation they may follow as citizens. Yet others see sociology as part of the main stream of Western humanistic and scientific thought, at once explanatory
SOCIOLOGY IN CAMBRIDGE
I93
and critical (cf. Smelser and David 1969: 15 7). It is impossible with limited resources to provide adequate teaching facilities for any one of these clienteles and it is tempting, in an academic climate where hard science has high respectability, to concentrate all efforts in meeting the requirements of those who want to learn about the positive findings of sociology, and about the ways in which these findings may be enlarged. I happen to agree that this is where much of our effort should lie. This is not because hard science is the key to local academic respectability, nor because it is customary to tailor university courses to the needs of future research students, even though only a small fraction of students follow that path. On the contrary, I have tried to indicate in this lecture why I think sociology can never conform to the model of a natural science. Furthermore there are many ways other than copying the natural sciences whereby an academic discipline can present a challenge and stimulus to those who choose it (cf. Times 1967). For me the reason lies in the fact that none of the four approaches to sociology I have mentioned can be entirely neglected. Fortunately, because social science has become part of the contemporary ambient culture, it is possible to allocate professional resources economically so as to make good the weaknesses in the awareness of students rather than to attempt the hopeless task of teaching everything from scratch. If the climate of opinion in the university showed that there was little interest in problems of poverty and war and overpopulation, or that there was a tendency to view questions about, say, the status of women, the incidence of marital breakdown, and the socialization of children in strict isolation from one another, then it would be necessary to pay more attention to these matters in the programme of teaching. Luckily this is far from the case, and we are particularly fortunate here in Cambridge that students develop an informed view of the world and of society through many contexts that have no direct connexion with the Tripos. Without this, the teaching resources available to us would be even more inadequate than they are. Even so, it is clear that we can provide much less teaching than we would like to do, and much less than we should in view of the importance of the issues at stake. Fortunately the plan of the new Tripos does provide for redeployment of resources, and for a continual reappraisal of the shifting boundaries between the various social sciences. There is also limited scope in the Tripos for a theme which, at the present juncture in world affairs, seems to me even more important, the study of the interconnexions between the social and natural sciences, particularly where these impinge on the fate of man. Yet even if our efforts were to be narrowly concentrated on comprehending the results of empirical inquiry into social affairs, on examining the diverse models of society and explanations of social action which have been proposed by theoreticians, and on devising ways of bringing theory
194
MODELS OF SOCIAL SCIENCE
and data into more effective association, even then sociology would appear to be an odd subject for this university. By the criteria of natural science, we still fall short of Sidgwick's conditions for academic recognition, partly because of the difficulties I have outlined, partly because the findings of social science, whether validated or not, tend to pass rapidly into the common stock of popular culture. Sociology, like most of the social sciences, remains at what Kuhn (1970) calls the pre-paradigmatic stage of intellectual development, and hence fails to pass Sidgwick's test of professional consensus. A great deal of so-called sociological theory consists merely of orienting statements, as Homans (1967) puts it, rather than scientifically respectable sets of testable propositions. On the other hand, by the standards of the humanities, much of our material must appear trivial, commonplace and mundane. The absence of a colonial situation prevents us from seeking academic respectability through the study of esoterica, trivial or otherwise. Attempts to describe the political, domestic and ideological life around us in a private meta-language tend to be regarded as pretentious and unproductive. Lastly, our ingrained critical stance and our professional concern with everyday affairs, with local administrative arrangements, and with other people's private lives continually call into question the relation of the university to the wider society. Thus even if the amount of effort that can be given to sociology remains strictly limited, it may well have an effect on the wider traditions and academic assumptions of the university which will require careful scrutiny if the good features of these traditions are to be maintained while an adequate adjustment is made to the changed conditions of the present day. From a purely parochial point of view the limitations on teaching and research in social science in this university perhaps do not matter. The university has many other claims to put forward for the recognition of excellence. But no university, however distinct and distinguished its way of life, can exist for long in isolation from the wider society which supplies it with material support, ideas and recruits. The intellectual may be free, as Dahrendorf (1968: 256n. 1) puts it, to make his own choice of questions, to study whatever attracts him and arouses his curiosity and to neglect other topics and areas of study that look dull or disturbing. It is essential that the social scientist should remain free to devote much of his time to asking awkward and even unanswerable questions. But as well as questions the social scientist also has to face what Dahrendorf calls problems, which arise from the real world and which are urgent and demand solutions. Here the social scientist is much less free to pick and choose, for these problems arise from the link between the university and the wider society. Without this link the university would soon wither away. In the wider society at the present time we see that scene which has become so familiar that we do not always heed it - a world of endemic war,
SOCIOLOGY IN CAMBRIDGE
I95
where increasing power seems to yield declining results, where population pressure and environmental degeneration have already destroyed the notion of an indefinitely long future and where the appeal to the past and to tradition has become counter-productive. These conditions arise more from social than from physical causes and present us with pressing social problems as well as difficult sociological questions. These problems and questions cannot be resolved without sustained analysis. Any university that turns its back on these problems, or remains uninterested in these questions, runs the risk of losing the support of those, old and young, who realize that time is not on our side.
12
SOCIAL SCIENCE IN INDIA: COLONIAL IMPORT, INDIGENOUS PRODUCT OR UNIVERSAL TRUTH?
In this chapter I return to a topic discussed in Chapters 6 and 7, the colonial system that prevailed throughout much of the world until the middle of the twentieth century. This time, however, I look not at relations of power or perceptions of the past but at perceptions of social science. I had worked briefly in India before being invited to contribute to an international symposium on the differences in the praxis of anthropology between Western and non-Western countries. Its long history of intellectual endeavour made India well suited for an attempt to place social science in a context of imperial conquest and retreat. In the paper I endeavour to show that the view practitioners and onlookers take of the social science enterprise is conditioned by their political stance as well as, in some cases, by the writings of philosophers of science.
i Introduction In the past, almost all the empirical studies carried out under the banner of social anthropology were conducted by men and women who were citizens of industrial nations and whose focus of inquiry was the social life of communities with techniques simpler than their own. Even today, most anthropologicalfieldworkis of this kind. The phenomenon of indigenous anthropology in non-Western countries, and particularly in countries of the Third World, therefore poses two obvious questions. First, how, if at all, is the process of inquiry altered when anthropologists and other social scientists study people who are more like themselves or who at least belong to the same nation as themselves? And second, does the process change significantly when the national society to which a scientist belongs, and 196
SOCIAL SCIENCE IN INDIA
I97
which therefore provides the financial, logistic and professional infrastructure for his inquiry, is poor and newly independent rather than affluent and with a long history of political and intellectual autonomy? We cannot provide adequate answers to these questions just by considering indigenous anthropology and fieldwork carried out by Third World anthropologists. We have to construct a model that takes account also of the phenomena with which they are contrasted, the study of people by strangers and the activities of Western social scientists. The model must also allow for the fact that the knowledge gained by social inquiry is perceived in different ways by the different parties involved in the process of inquiry, for these perceptions in part reflect the relative power of the parties over one another and have changed from one period of history to another. In part they also reflect different views about how social science is to be seen in relation to the two other major branches of learning, natural science and the humanities. Variations in these various perceptions of knowledge have practical consequences for the conduct of social science in the Third World that are seen only in milder form, if at all, in the West. Within the limited compass of this chapter I can only present one possible model and illustrate its application with scraps of evidence from only one Third World country, India. 2 Social science as science or art We can best begin by discussing the most abstract part of the model, the determination of the tripartite contrast between the humanities, social science and natural science. Natural science typically consists of assertions that are claimed to be true or, if not true in any absolute sense, at least valid, eternally and ubiquitously. Their validity can be tested by anyone who has the appropriate technical skill, be they black or white, naked or clothed. The law of gravity ignores political boundaries and affects equally the dove of peace and the bombs of war. The speed of light is the same whether measured in Washington, Moscow or Timbuktu. In this sense the humanities are quite different from natural science. Their objects of study typically are cultural products firmly embedded in particular historical and spatial contexts. It is true that some great works of art manifest their greatness in transcending the limitations of the contexts in which they were created, but even these masterpieces retain their local garb. Shakespeare displays his membership of Elizabethan England even though critiques and commentaries on his work are produced all over the world. Furthermore, unlike international commentaries in natural science, these various interpretations of his work often differ widely from one another and display in their turn their own contexts of origin in different dramatic traditions and different schools of criticism.
I98
MODELS OF SOCIAL SCIENCE
Given this contrast between natural science and the humanities, what then is the position of social science, that label under which we conventionally group social anthropology, sociology, economics, political science and cognate branches of learning? The choice of the label, social science, rather than, say, useful arts or applied humanities, seems to indicate that these disciplines, just as much as physics, chemistry and biology, are concerned with generating universalistic propositions whose validity is independent of both the identity of the scientists who first enunciated them and whether they are utilized for good or evil. Yet though the label is used widely, those of us who are social scientists often find it difficult to convince our colleagues in natural science that sociology, for instance, is a 'proper' science. An established corpus of interconnected and validated social laws, which would be a hallmark of scientific respectability, is embarrassingly absent. We are then tempted to fall back into maintaining that disciplines are to be judged by their intentions rather than by their achievements and to predict that in a mere one or two centuries social science will have vindicated by its proven discoveries its claim to scientific status. Alternatively we can shift our ground and argue, with support from other European languages, that the English word 'science' does not necessarily imply universalism or falsifiability (cf. Machlup 1956: 165, fn. 5) and that empathy and intersubjectivity, with which several branches of social science are abundantly supplied, are as essential for understanding scientifically the culture-bound utterances and social actions of fellow human beings as for, say, interpreting the untranslatable poems of Gerard Manley Hopkins. 3 Social science and political independence Whether social science is on one side or the other of the arts/science fence, or balanced precariously on top, is a question in the taxonomic status, or propositional structure, of academic knowledge which has generated a good deal of discussion among philosophers of science. Most of this has been quite far removed from the interests of working scientists and humanists alike. But while this question may be treated appropriately as irrelevant to both routine teaching and practical research in the ivory towers and scientific laboratories of the West, in the Third World it is more than merely a query of idle scholasticism. It has immediate relevance to scholarly and scientific praxis. This difference in relevance arises because when political independence was achieved in most Third World countries, universities and research institutes, if they existed at all, were institutions that had been established during the colonial rather than precolonial era. Political independence having been achieved, were universities therefore to be closed down so that the newly free nations could demonstrate that
SOCIAL SCIENCE IN INDIA
I99
they had gained cultural as well as political autonomy? Or were they to be retained and expanded, to show that, released from colonial constraints, these nations could modernize and begin to challenge their former rulers as equals? Or were institutions of learning to be modified, purged of colonial influences, and restored to their original indigenous purposes? The alternatives chosen varied from country to country and also within countries, depending on many factors, among them the extent to which the new ruling elite was itself a product of these institutions during the colonial period. Responses varied also from one branch of learning to another, and it is because of this variation that the problematic metascientific status of social science acquires practical significance. By and large, under postcolonial regimes natural science, as a corpus of theory and a mode of inquiry as well as a subject for instruction, was accepted as being universally valid. I do not know of any attempt to constitute a national physics, nor even, with the example of Lysenko in mind, a national genetics, anywhere in the Third World. All that happened in natural science was a shift of emphasis from one kind of technology to another. Even the treatment of physical illness, where, in all countries of all three worlds, natural science merges into non-scientific tradition and fad, indigenous therapies were perceived as supplementing or surpassing, but not refuting, Western medicine. On the other hand the humanities were often seen as heavily tainted with colonial values. The literature studied by students under the new regimes has had to be indigenous rather than Western; local languages have replaced irrelevant Greek and Latin or even, as in Indonesia, the former colonial language, while history has had to be taught and written as the story of national liberation, not of metropolitan growth and imperial expansion. In the long run this transformation of education and research could be achieved only under the leadership of indigenous intellectuals, rather than of expatriate experts, however well-intentioned they might be. 4 Social science in India: seven standpoints Social science can be perceived broadly in either of these perspectives, as a universalistic science or as an art tainted by its colonial origins. India provides us with a good example of a diversified response to the intellectual implications of the ending of colonial rule, for it has a long history of interest in the study of society. Mukherjee (1977) has recently surveyed the development of social theory and social philosophy in India from Vedic times onwards. While his analysis of present-day Indian trends in sociology and social anthropology may be queried, the evidence he adduces for the complexity of the intellectual currents among social scientists and their forerunners from the eighteenth century onwards is indisputable.
2OO
MODELS OF SOCIAL SCIENCE
Throughout the nineteenth century there was public debate about the extent to which Indian traditions and institutions had to be preserved and revivified if social order was to be maintained and contrariwise, the extent to which the subcontinent had to be Anglicized if it were ever to progress (Stokes 1959; Ashby 1966: 48-54; Sarkar 1970). The development of university departments of social science was one small part of the process of Anglicization, even though several of the scholars who were appointed to them drew much of their inspiration from ancient Hindu traditions rather than from modern Western writings. Mukherjee identifies five questions that may be asked of a social phenomenon: What is it? How is it? Why is it? What will it be? And what should it be? The relative emphasis given to these five questions will depend on whether the phenomenon is being studied ideographically or nomothetically and on whether the study is being made with social action as an objective or not. Mukherjee shows convincingly that the five questions have appeared in a variety of combinations in studies carried out at different periods in India. His analysis does not, however, focus on the contrast posed in this chapter. For our purposes it is more convenient to build on an analysis by Saran (1958) in which we can identify seven different responses to the growth of teaching and research in social science, responses which had begun to manifest themselves in India even before the attainment of political independence in 1947 (cf. Clinard and Elder 1965): 1.
2.
3.
Some writers rejected totally the values of Western civilization which, they maintained, social science seemed to take for granted. They advocated a return to a traditional Arcadian harmonious social order, and found it easy to quote from Western writers in support of their aim. Indeed, a few scholars in this group advocated a universal return to traditional preindustrial society, in the West as well as in the East. At the other end of the spectrum were those who believed that Indian society should be modernized and that modernization could be achieved by the adoption of Western modes of policy formation. Radhakamal Mukherjee, professor of sociology in Lucknow and Calcutta, supported meliorist policies to overcome poverty and inequality, including the abolition of caste, based on empirical inquiries and implemented through social work institutions modelled on those of the United States. He and others with similar views supported their arguments with carefully selected quotations from ancient Hindu texts. Thus, whereas the rejectionists in group 1 appealed to anti-industrial writers in the West, the modernizers in group 2 cited traditional Eastern scriptures. A third and related type of response was displayed by those who
SOCIAL SCIENCE IN INDIA
4.
5.
6.
2OI
supported policies similar to those favoured by group 2, but who denied that these were aimed at modernizing Indian society; their aim was a return to an earlier social and intellectual condition from which India had unfortunately strayed during the last couple of millennia or so. Thus, for example, Damle (1967) claims that modern American sociological concepts represent merely belated rediscoveries of concepts expounded long ago in the traditional corpus of Hindu literature. Similarly Prabhu (1963) tries to demonstrate that traditional Hindu social organization was in conformity with the norms for social harmony established by modern social psychology. Saran (1958: 1025), on whose critique I am basing my classification, notes Prabhu's caustic remark that 'some critics considered him to be inadequately loyal to tradition while others thought that he had betrayed modernism'. All three groups of writers tended to see the impact of British colonialism and Western industrial civilization as a major event in Indian history, despite their varied responses to this event. A fourth group, on the other hand, argued that India and its social institutions had been subject to alien intrusions through several millennia and that its culture and institutions had developed and adapted successfully to the challenges posed by these invasions. It would do so again and history should be allowed to take its course. Despite their differences, all four groups tended to see social theory as essentially normative, as providing a yardstick, if not necessarily always a blueprint, for social policy. A fifth group, whose mentors were mainly American, was constituted by those who saw in social science a set of techniques for the analysis of social behaviour. Social science as such was value-free and not linked necessarily to any given set of policy goals. The West provided an abundance of expressions of this metatheoretical viewpoint, and its Indian advocates adopted largely unchanged its operational procedures of inquiry and categories of analysis. They therefore engaged in empirical inquiries into topics such as social class, work satisfaction and juvenile delinquency, using without modification, as did those in group 2, modes of research that had been developed in American and European contexts but denying that these were tied to any one set of social policies (cf. Mukherjee 1977: 63). The Marxists constituted a small group, seeing themselves as equipped with a theoretical framework fully adequate for the analysis of Indian, and indeed of any, society. The Western origin of Marxism was not an accidental or irrelevant fact but constituted for them no obstacle in its application to the understanding, and eventual transformation, of Indian society.
2O2
7.
MODELS OF SOCIAL SCIENCE
Finally there was the seventh group, made up those who carried out village or community studies. These were divided in their attitudes toward science and the humanities. On the one hand were those who looked for methodological guidance to Robert Redfield with his model of little and great traditions. Redfield enunciated explicitly his vision of anthropology as one of the humanities. On the other hand M. N. Srinivas, doyen of this group, sponsored the posthumous publication of his teacher Radcliffe-Brown's most uncompromisingly universalistic collection of papers, Method in social anthropology (1958). In their monographs members of this group tended to adopt what has been called a common-sense empiricist standpoint: in their theorizing they viewed their discipline as tolerant of all systems of values and universalistic in intention, even if it remained for the time being mainly ideographic in practice.
The seventh group remained happily on the fence but the first four groups saw social science as closer to the humanities than to natural science. For them social science was not value-free. Social science implied a set of social norms which was related, either sympathetically or antagonistically, to traditional Hindu norms or to some postulated set of enlightened universal human values. The fifth group regarded social science as simply another science and it ignored the specificity of the Western context in which it had developed. The limitations of many of the inquiries carried out by scientists in group 5 are well stated by Miller (1976: 112-13) when he contrasts some recent books from this group with the work of N. K. Bose (1975), whom we may classify as belonging to group 3. Miller says of the former that they 'write not as creative individuals with any particular identity but as students who have gained from their masters not wisdom and true apperception but techniques for reducing people and cultures to anonymous, colorless statistics . . . The shallow "common sense" empiricism of the village studies . . . has given way to a crude scientism in which scientific method is seen as automatically yielding objective, rational truth'. The scientific status claimed by this fifth group contrasted sharply with the perception of science held by the Marxists of the sixth group. They saw social science as inherently value-laden but claimed that Marxism was scientific, based firmly on empirical evidence, and that its findings subsumed and superseded the biased findings and interpretations of bourgeois social science. All seven interpretations of the metatheoretical basis of social science continued in India after Independence. The activities of the fifth group, the technicians as we might call them, expanded to meet the needs of the central government's five-year plans and with the diffuse movement for national modernization. Saran (1958: 1029) writes: I n pre-independent
SOCIAL SCIENCE IN INDIA
2O3
India, no sociologist could feel wholly at ease in accepting Western civilization for it was associated with our subjection to Britain. And hence there was a bias in favor of specifically Eastern and Indian values . . . After independence the danger of strengthening foreign political rule through acceptance of foreign culture being over, the major problem for most of the Indian sociologists is how to effect the transition of India from a backward to a fully industrialized, "open" society.' 5 Objections to research by foreigners The policy of trying to catch up with the West on Western terms and of treating social science as a value-free set of techniques that scientists, whether local or foreign, applied impartially, did not go unchallenged for long. By the mid 1950s, after a decade of independence, two new trends emerged. The first was simply an objection to foreign social scientists working in India. In 1955 six sociologists signed a letter to the press complaining about the activities of American and other foreign social scientists. They asserted that 'most of the modern research methods and tools which these foreign experts are using have been implicitly designed to serve commercial purposes and are now being adapted to military objectives. This is why all American research projects in India are concentrating on Communication, Culture Contact, Social Stratification and Public Administration. These fields are strategic for Commercial and military purposes, an awareness of which is reflected in their questionnaires' (National Herald 1955). During the twenty years after this letter was published, fear of expatriate social scientists acting as agents for foreign commercial interests or as spies for foreign governments continued in India, as it did in most other Third World countries. The central government in Delhi became sensitive to criticism of its policies, or, more often, of its failure to implement its declared policies, long before the declaration of the Emergency in 1975. This sensitivity was greatest when the criticisms were made by foreigners. The government therefore established a variety of administrative controls aimed at preventing social research by foreigners on matters where the publication of data on social conditions might possibly lead to criticism of its efficiency. India was here following the example of many other countries, not all of them in the Third World. Empirical research by outsiders in communist countries is usually very difficult to arrange, and there are often substantial impediments to research by foreigners in the West; in South Africa strict controls on research among Blacks have been in force since the 1930s. In several instances these controls have been considerably more draconian and probably more effective than they have been in India. While the research activities of foreigners have been
2O4
MODELS OF SOCIAL SCIENCE
influenced by the necessity of obtaining local administrative approval, indigenous social scientists have been affected both by the consequent deflection of expatriate activity and by the cautious attitude of the administration toward research even by local scientists. 6 Relevance in research The objection by governments to research by expatriates cannot be explained merely as being caused by xenophobia - certainly not in the case of India. It is better understood as a consequence of the view, held by many leading academics as well as by government officials, that there is little room in India, and in other poor countries, for free-ranging disinterested empirical inquiry carried out in the pursuit of knowledge for its own sake. In light of the persistent attacks, even in the relatively affluent West, on the alleged iniquity of wasting public funds on 'useless' research, the Indian attitude is scarcely surprising. Even when there is no suggestion of foreign espionage, or when research proposals are put forward by Indian social scientists, government officials may refuse to provide assistance and information to facilitate research projects that they regard as unlikely to yield any benefit locally, whether this be socially or individually defined. Thus S. C. Dube (1973: 21-2) writes: 'Researches that do not relate to matters of social consequence and public issues of social structure are best dropped or accorded low priority . . . Social sciences cannot be and should not be practised in isolation from what is happening in the society and from what is sought to be done in it.' These remarks are aimed at the activities of Indian social scientists as much as at those of their expatriate colleagues. The tendency to permit research only on topics approved or selected by government, irrespective of whether they are executed under local or foreign scientific leadership, is found widely in the Third World, in state socialist societies, and to some extent in the liberal democracies of the West. Where the government acts as financial sponsor for research by its own nationals it can call the tune and confine their attention to topics it has chosen for them. When the research is to be carried out by expatriate scientists with funds from elsewhere, the local government, as gatekeeper, can still exercise censorship on what topics are selected.
7 Who can be a competent fieldworker? In India more has been at issue than simply the government's definition of relevance and its ability to control the choice of topics for investigation. The choice of who should conduct the inquiry has been treated not just as a matter of administrative and political expediency but has been con-
SOCIAL SCIENCE IN INDIA
2O5
strained by the positions taken in a debate on the objectivity and subjectivity of social science. To some extent this Indian argument about the metatheoretical characteristics of social science must be seen as a local manifestation of a wider controversy among social scientists for and against positivism and Verstehen; in India, however, it took a form which seems not to have parallels in other Third World countries; in the West the controversy does not have the practical consequences it has had in India. The argument took a distinctive form in India because of that country's long and culturally specific tradition of scholarly writing and disputation stretching back even further in time than corresponding traditions in the West (cf. Shils 1961: 20-8). In the West this debate, like the associated debate on the taxonomic status of social science, is conducted with only long-term implications, if any, for practical and political action. In their Indian context, both debates have had immediate consequences for the conduct of social inquiry. We can link the seven different perceptions of the validity of social science listed earlier with views about objectivity and subjectivity in social science and the appropriate mode for empirical inquiry. Those social scientists who belonged to the fifth group, the technicians, saw social science not only as value-free but also as positivistic, concerned with the analysis of objectively identifiable social facts. Hence, as in physics, what mattered was not the national affiliation or personal code of values of the scientist but his or her professional competence. Those who supported the second view, that India should be modernized in accordance with the enlightenment inherent in social science, were also not concerned with the personal and private values of the scientists. The Marxist orientation adopted by the sixth group was, to Marxists, a necessary qualification for correct social analysis but it was a qualification accessible to anyone, indigenous or foreign, who saw the light. The seventh group required professional competence, including command of indigenous languages and symbolic codes, and it perceived this as being within the grasp of adequately dedicated outsiders. Indeed, the omnicurious outsider might have an advantage over the insider who was always in danger of taking his own culture and society for granted. Members of the first group rejected social science completely; it was equally abominable whether practised by Indians or Westerners. Finally, the fourth group and, more clearly, the third group tended to regard Indian institutions as distinctive, so that they could be understood fully only by scientists who had grown up in Indian culture, who spoke and thought in Indian languages, and who identified themselves politically with the future well-being of India. This last viewpoint found support in those writers in the West who argued that social action can be understood only in terms of the subjective meanings that actors attach to the social situations in which they find themselves. A
2O6
MODELS OF SOCIAL SCIENCE
science based on intersubjectivity, on being able to comprehend, with reasonable accuracy and confidence, what goes on in the minds of the actors, necessarily requires a good deal of congruence, in terms of cultural background, unspoken assumptions and symbolic meanings, between the actors or members and the scientist who studies or interacts with them.
8 Contributions to Indian sociology What seems in retrospect to have been the first shot in this Indian positivist-subjectivist debate was fired in 1957 in an editorial proclamation for a new journal, Contributions to Indian Sociology, launched jointly by two social anthropologists, one French and the other English, Louis Dumont and David Pocock. The editors maintained that one of the main intellectual tasks facing sociologists and social anthropologists working in India was to define the relation of their discipline to classical Indology, the study of the traditional texts of the Hindu canon (cf. Madan 1982). It is somewhat ironic that what was soon to become an argument mainly among Indian scholars was precipitated by two Western social scientists. They attempted to defend their right to carry on their profession in India, on the one hand without conceding a superior understanding of Indian social institutions to native Sanskritists who could not be challenged on their home ground, and yet on the other hand insisting on the necessity of taking the findings of Sanskritic scholarship into account as one factor in the scientific analysis of Indian cultural values. The title of their journal was ambiguous. It could be taken to mean: contributions to the Western or universalistic sociological analysis of Indian society and culture; or else contributions to the development of a brand of sociology that would be characterized by Indian concepts and analytical procedures as well as by data collected in India. At least one of the readers of the new journal read the title, and the donnishly cryptic editorial introduction (Dumont 1957), in the second sense and attacked the editors for their apparent betrayal of the aims of science, which he saw as essentially universalistic, comparativistic, and, I would guess, positivistic as well (Bailey 1959). The editors replied to their critic by affirming the correctness of the first interpretation of the journal's title. For the editors, sociology was a universalistic discipline, and their special task was to apply its methods of analysis and interpretation to India. They wrote in i960: 'Dr Bailey has misunderstood our use of the phrase 'Indian sociology" if he imagines that, given half a chance, we would also offer an 'Indian chemistry" ' (Dumont and Pocock i960: 84). Dumont (1966: 23) later added that for him a Hindu sociology was a contradiction in terms. Yet despite these appeals for a combination of insider and outsider views which, for the editors, was 'the guarantee of real sociology', their balanced view was attacked by an Indian reviewer as
SOCIAL SCIENCE IN INDIA
2OJ
'positivistic hubris' (Saran 1962: 68). In response to this kind of criticism Dumont (1966: 25) was moved to warn those he called 'progressive Indians' that they should be particularly alert against the danger of a fake or Hindu sociology developing as a weapon of obscurantism and reaction. The debate took a new turn with the publication in 1968 of an outspoken paper, presented at a conference on urgent research in social anthropology, by Uberoi. In 'Science and swaraj' (that is, science and independence) he writes: I believe that the advancement of science in India cannot be separated from the advancement of independence. The end of our intellectual and practical apprenticeship in science is long overdue ... but we are constantly cajoled or browbeaten to prolong it... Such things are done to poor nations and peoples today, in science as in politics, in the new fashion of a false cosmopolitanism and false humanitarianism, not in the old terms of colonial tutelage or racism. The latter ideologies emphasized discontinuities and hateful ones, while nowadays all talk is of brotherhood, harmony and sweet reasonableness, but the results are not totally dissimilar. The nuclear non-proliferation treaty, in effect poised to subvert both our science and our independence, is a characteristic product of the new era, combining humbug and coercion in equal proportions. Uberoi went on to attack what he called the international system of science, and ended by saying: 'Until we can concentrate on decolonialization, learn to nationalize our problems and take our poverty seriously, we shall continue to be both colonial and unoriginal' (Uberoi 1968: 119-20, 123; cf. Madan 1982). Uberoi's paper was interpreted by some as an appeal for a distinctively Indian sociology, with its own specific concepts and procedures. This interpretation was clearly wrong, for in the discussion that followed the presentation of his paper Uberoi emphasized that he was examining only the social system of science. He advocated not a culturally unique sociology but a nationally independent one. He said, 'For us the important considerations should be to define our relevant problems and pursue their solutions, not worrying about the sources of ideas: that task can be left to be performed by our intellectual biographers' (Abbi and Saberwal 1969: 191-2). In a more recent statement of his position, Uberoi has been even more explicit: Intellectual self-reliance, in a word, is the achievement of independent home rule or swaraj in the internal and external life of national intellectual institutions; it is not a question of relying solely on home-grown or swadeshi ideas or of making ritual rules of purity and pollution in international intellectual relations ... The Indianness of an Indian science cannot properly consist in the difference of its method or its data, but in the independence of mind with which its theoretical perspective is constructed ... The sociology of India cannot possess a distinctive method, in the philosophical sense, for the simple and sufficient reason that the
2O8
MODELS OF SOCIAL SCIENCE
methods of science are truly universal and every human being knows them in principle ... We must walk on the path of self-reliance on our two legs, the theoretical and the empirical, in one coordinated movement forward. (Uberoi 1974: 136, 150). In the on-going debate about the nature of social science in India, its theoretical base and its empirical application, several issues have become entwined. First there is the taxonomic or metascientific issue, shared with scholars in the West, whether, in the words of one Indian participant, 'theory in sociology comprises a system of general propositions and concepts or whether it is a descriptive-analytic account of particular social or cultural phenomena in terms of a hierarchy of value-premises derived from some unifying principles' (Singh 1970: 140). In other words, are we dealing with something like natural science or like the humanities? Secondly, are these phenomena objective, so that any competent observer can discover and record them, or are they essentially subjective, capable of being comprehended only by those scientists who are steeped in Indian culture? Thirdly, how can the growth of social science in India best be fostered so that Indian social scientists will come to treat their Western counterparts as their intellectual equals and will no longer either accept uncritically the latest intellectual fashions of the West or reject them all equally uncritically? Fourthly, to what extent must empirical social research in India, whether carried out by Indians or by foreigners, concentrate on topics that are perceived as socially relevant? And if social relevance is a necessary condition for permitting an inquiry to proceed, who is to determine what is relevant: the bureaucrats in New Delhi, the local administration, the social scientist, his sponsor, or, though they are less often mentioned, the ordinary people who form the focus of the inquiry? The fifth and last issue is whether the first four can be discussed in isolation from one another or whether they affect each other. For instance, some commentators argue that Indian social scientists should raise their intellectual morale by no longer seeking to visit the West themselves. At the same time they should write textbooks designed specifically for Indian students, rather than continuing to rely on imported texts. These are, it is argued, often dismally parochial, being narrowly focussed on North America, and quite unsuited for the needs of untravelled Indians. This criticism applies more forcefully to texts in sociology and psychology than in social anthropology. But the same argument is applied to social science as a whole: in order to gain full benefit from closing its ranks organizationally, the profession must also isolate itself intellectually from Western scientific thought. Yet if this argument is accepted, does it not follow that social science has to be seen as ideographic rather than nomothetic, like the humanities rather than like physics and chemistry?
SOCIAL SCIENCE IN INDIA
2O9
Some Indians accept the argument and say that, whatever may happen in the long run, at least for a transitional period there should be both organizational and conceptual apartheid. It is interesting to note that Uberoi, although one of the most eloquent advocates of the importance of Indian scientists proclaiming their independence and equality, comes out firmly against this expedient. He has demonstrated what he sees as the correct implications of his paper on 'Science and swaraj' by concerning himself with the analysis of contemporary Western thought and with the development of notions of social order from classical times onwards (Uberoi 1974, 1978, 1984). A close interest in, for example, the debate between Luther and Zwingli may at first sight seem oddly matched with a lively concern for Indian intellectual self-reliance, but on closer scrutiny this apparent incongruence makes good sense. If a self-reliant Indian Weltanschauung is to be more than the ideology of an ostrich-like apartheid, it must seek to interpret the development of the West as incisively as it explains the complexities of Hinduism. 9 Three perceptions of knowledge Arguments in India and elsewhere about the place of social science among other branches of learning thus reflect not only the long-standing debate about positivism and subjectivism but also efforts to complete the process of postcolonial liberation. There is, however, a third way of looking at these arguments which sets them in an even wider perspective but which also bears directly on the possibility or necessity of empirical inquiry being carried out by members of the community or nation being studied. Consider the outcome of a typical field study. A book appears reporting that in the village of Kachikoti there are three main castes; the dominant caste is divided into two factions, led by X and Y; a few years ago, for the first time, a young widow of the dominant caste was remarried; and so on. What sort of knowledge is this? Does it increase our understanding of the human condition? Does it help us to oppress the subordinate castes more effectively, or to overthrow the dominant caste? Can we infer the crumbling of the objection to the remarriage of widows? Are we invading the privacy of the young widow if we publish the fact of her remarriage? Several different answers have been given to questions like these. Ethnographic information began to be collected in India during the colonial period, and some recent commentators seem to imply that this information was gathered, in India as in the rest of the colonial world, merely because it was thought to be useful to the British in maintaining their imperial domination. If this was so, then ethnographic knowledge was seen as a source of power. This thesis is, in my view, a very partial account of what happened, for it seems to imply a co-ordination of
210
MODELS OF SOCIAL SCIENCE
empirical inquiry which did not exist and an ability to respond to information which very few colonial administrations possessed. I accept the views of writers as different in their outlooks as Raymond Firth and Kathleen Gough who have emphasized that social anthropology, along with other social sciences, was a product of the eighteenth-century Enlightenment and that ethnographic information was perceived by those who first started to collect it as a source of enlightenment. In India the Asiatic Society of Bengal was founded in 1784 in response to an initiative taken by the Governor-General, Warren Hastings (Madan 1977: 263), though the Governor of Bengal and Bihar had stressed the importance of collecting information regarding the leading families and their customs as early as 1769 (Srinivas and Panini 1973: 181; Mukherjee 1977: 22; cf. Shils 1961: 43). At that period, and indeed throughout most of the nineteenth century as well, the ruling elites, both in the metropolitan countries and in the colonies, were seldom in doubt about their right to rule. They recognized their need for information, so that the true interests of the lower classes and dependent peoples could be discovered, for members of these social groups were thought to be incapable of articulating their needs by themselves (cf. Madan 1982). Social inquiry was indeed almost entirely directed at the inarticulate masses, the poor at home and natives abroad. It was much later, between the world wars, when notions of improvement and development began to gain currency, that the notion of knowledge as a source of enlightenment gradually gave way to an idea of knowledge as a source of power. The knowledge gained by empirical inquiry was seen as a means for discovering how to change people's lives, to make them construct latrines, dig contour ridges, spurn alcohol and prostitution, stop beating their wives, and so on. The concept of social engineering grew out of this perception of knowledge as power, and reached its climax in the ill-starred Project Camelot. The growth of government intervention in social life, linked to an exaggerated view of the extent to which social institutions and behaviour could be modified to conform to a planner's blueprint (a view for which social scientists themselves must take much of the blame), has led to the massive increase since 1945 in the number of research projects predicated on the notion that knowledge is a source of power. But another social movement came only slightly later, which has counteracted this trend. There has been what may be termed a citizens' revolt, in which individuals and previously powerless groups have begun to refuse to provide the information sought from them. This revolt has been partly fuelled by the enormous increase in the efficiency of information retrieval achieved by computers but might well have occurred without their arrival. Knowledge began to be seen as a form of property, which might sometimes be bought but which often was simply not for sale at all, or was for sale only in
SOCIAL SCIENCE IN INDIA
211
a narrowly defined market. Some communities experienced saturation effects, particularly exotic settlements relatively easy of access in areas of low population density, such as Aboriginal reserves in Australia. In the United States the refusal to answer questions in social surveys has come to be seen by citizens as 'a form of protest against some part or all of established society' (American Statistical Association 1974: 32). Most people see decent obscurity as an important part of their personal freedom (cf. Warner and Stone 1970: 114) and in the Third World some people see political independence as including freedom from being studied by Western social scientists (Barnes 1977a: ch. 3, 1980a: ch. 3 & 9). Within Western countries the spread of the notion that knowledge of personal characteristics and social practices is a form of private property has already led to increased refusal rates in social surveys, difficulties over the wording of census forms, and, more interestingly, the growth of enclaves of learning, open only to those with the right ascribed attributes, Black sociology and Women's studies being the most prominent examples. In the Third World it has led to various forms of taxes on knowledge acquired by empirical inquiry. In all countries the revolt of the citizens has forced social scientists to take account of the diacritical fact that separates them from their colleagues in the natural sciences, the fact that empirical inquiry depends on the cooperation of the people being studied. There has been a shift in the balance of power between the scientist and the citizen, so that increasingly the social scientist has to negotiate with the citizens he wishes to study about what he should study, how he should conduct his inquiry, and what he should write in his report. These constraints are likely to bear on the expatriate social scientists more severely than on his local colleague, who may be politically in a more powerful position and who may be able to continue to operate largely within what has been called the colonial paradigm offieldwork, based on the model provided by the natural sciences. But if he gains in one respect he may lose in another. The social scientist who conducts inquiries within his own national society may be better able to understand the cultural background and symbolism of the citizens he studies but he may have greater difficulties in achieving for himself a satisfactory special status within the community than his colleague from overseas. An Indian social scientist studying a village may have fewer obstacles to linguistic and symbolic competence but it may be much harder for him than for a Westerner to persuade the citizens that he stands outside the caste system and that as an outsider he can be trusted with secrets by all factions. 10 The future of social anthropology In these ways, and doubtless in many others as well, changing perceptions of knowledge about social life and of the taxonomic status of social science,
212
MODELS OF SOCIAL SCIENCE
plus alterations in the balance of power between the West and the Third World and between citizens and the state, combine to reshape what is still seen by many social scientists as quite a simple matter: asking people questions and writing about what they say. Much greater professional selfawareness will become necessary if social science is to survive. One aspect of the changing social environment in which research is conducted is that increasingly anthropologists find themselves studying communities within their own national boundaries and even collaborating with citizens in these communities who are studying themselves. This is likely to yield intellectually interesting and socially unexpected results. But we should remember that what will emerge will not be social anthropology as we now know it if, with Levi-Strauss (i966f: 126), we characterize that discipline as the study of culture as seen from the outside. Guiart (1965) once argued that Levi-Strauss' main achievement had been to rescue French anthropology from the ruins of French colonialism. Madan (1982) points out the limitations of the outsider's viewpoint. It may be necessary to rescue the disipline again, this time from Levi-Strauss' characterization, if it is to survive in a truly postcolonial world.
Part III
A MODEL OF MODELLING
MODELLING: FOR REAL OR FOR FUN?
Having looked at two classes of models, in part I at models of various aspects of the social world and in part II at models of intellectual formations associated with social science, we finally turn to look at the process of modelling itself. The stimulus to write this paper came from an invitation to talk to a conference of social network analysts in 1982. Chapter 4, which was first published in 1954, discusses social networks, but although I later published some further papers on this topic, my interests and the exigencies of academic responsibilities had increasingly led me away towards other areas of inquiry (Barnes 1987b). Long before the invitation arrived I had found myself unable to follow the burgeoning technical literature on the analysis of social networks. A paper examining my inability was obviously called for, particularly as one of the criticisms made by several sociologists and social anthropologists of the work of network analysts was that it had become unnecessarily technical. I wished to defend, and indeed applaud, the development of sophisticated analytical procedures (cf. Barnes and Harary 1983c) and at the same time to place this development in an historical context. In short I tried to construct a diachronic model of modelling in social science, with network analysis as my type specimen. As foreshadowed in Chapter 4, network analysis, or structural analysis as it is sometimes known, is a way of looking at social phenomena, particularly relations between individual persons or other entities, in terms of a set of points or nodes, and a set of ties or arcs joining all or some pairs of points. The configuration of points and lines can be subjected to mathematical analysis, using con215
2l6
A MODEL OF MODELLING
cepts drawn from graph theory, so as to show, inter alia, how relations between any specified pair of persons are influenced by the content of the relations prevailing between other individuals in their social vicinity. It is a mode of analysis that supplements the more conventional approach of seeing society in terms of its constituent social groups (Wellman 1988). As delivered at the second annual Sunbelt network conference, this address was slightly longer than the version published in the following year. I have taken the opportunity to restore some of the deleted material.
Although the literature on models in science is vast, many social scientists query its relevance. Some of us, myself included, are sufficiently oldfashioned to think that the real world should remain the main focus of scholarly attention in social science. The philosophy of the social sciences, of which these discussions of models form an important part, seems to add very little directly to our understanding of that real world. If so much has been said already, and to such little use, why then add to the jungle of words? Yet there are two features of this literature that may justify a modest attempt to add to it. First, there is little evidence that network analysts spend much time browsing through this literature. It doesn't get cited in Social networks or Connections. The word 'hermeneutics' never appears there, nor do the authors of articles in these journals speculate independently on the subtle issues that dominate the discussions of the philosophers of science and others who write professionally about models. By and large network analysts press on regardless. They produce a profusion of models, uninhibited by any doubts about how these might fit into some typology and unconcerned about the ontological and epistemological status of the connexion between the world and the model. Indeed, there is a striking and refreshing contrast between the essays that I read continually in Cambridge and the corpus of network analysis. In our student essays the words 'epistemology' and 'ontology' luxuriate, largely at the expense of any data from the real world, whereas in the writings of network analysts these words are rare birds indeed. The second relevant feature of the literature on modelling in social science is that it is almost completely classificatory, morphological and normative, and not at all sociological, even when written by card-carrying sociologists. There are learned and fascinating discussions on topics such as the difference between models and paradigms (Hesse 1976), between emic and etic models (Pike 1967), and between analogue, iconic and symbolic models (Wilier 1967). There are also historical accounts of how the key model in scientific thought has, through the centuries, shifted from
MODELLING: FOR REAL OR FUN?
217
society to the wheel, the balance, clockwork, the organism and so on (Deutsch 1951). But there is almost nothing on what we might call the sociology of modelling in the social sciences, at least not at close range. Kuhn's well-known Structure of scientific revolutions (1970), with its later modifications, and competing proposals from Popper, Lakatos, Feyerabend and others, provide large-scale accounts of how in natural science one constellation of concepts, axioms and propositions gives way to another. But these accounts cannot easily be modified to allow for the instrinic differences between the natural and the social sciences (cf. Putnam 1978: 55-65). Furthermore, their attention is directed at paradigm shifts, the major changes in scientific thinking, rather than at the waves of intellectual fashion that animate the scene between one major change and the next, during periods of so-called routine science. Certainly no one seems to have looked at network analysis from this point of view. Yet the subtitle of Sam Leinhardt's collection of articles, Social networks: a developing paradigm (1977), invites a sociological scrutiny of how routine social science, using the network paradigm or model, does actually develop. But as Harary and I have pointed out (Barnes and Harary 1983c), there are good structural reasons why network analysis, unlike graph theory, has not yet acquired an orthodox pedigree. I doubt if it ever will. All I want to do here is to throw into the ring some ideas about how modelling develops within a given intellectual and academic context, and to suggest how a knowledge of this development might affect future praxis. In doing so, I am reminded of the sentence with which Barry Wellman ends his excellent guide to network analysis: 'the capacity of network analysis to pose questions would profit from an enhanced capacity to provide answers' (1983: 180). The same cri de coeur can be made with metamodelling. We can begin with the observation that everyone seems to be in favour of models. As May Brodbeck says: 'Models are good Things'. Then she goes on to say, more questionably, 'And if models are good, "mathematical models", needless to say, are even better' (1959: 373). But she questions this latter statement herself so severely that she says we should abandon the term 'mathematical model' completely. What we have to ask is this: if models, mathematical or otherwise, are good, what are they good for? This question leads us at once into the treacherous field of definitions. There are dozens of definitions, and there is no point in adding to the roster. The best one for our present purpose is that given by Levins, a biologist, who says simply that a model is 'a reconstruction of nature for the purpose of study' (1968: 6). The main point of my thesis in this paper hinges on the word 'study' in Levins' definition. What are we studying when we use a model? Levins is quite definite about what he means. He refers to the study of nature. Nature is reconstructed and simplified in a model so that nature
2l8
A MODEL OF MODELLING
can be studied more effectively. But his definition can also be read in another way. The 'study' in question can be the study of the model rather than the nature from which it has been derived. For as an intellectual object the model has two attractions for us, and they compete with one another. Let's call them the extrovert and the introvert potentialities of the model. On the one hand we can use models to increase our understanding of the real world; on the other we can manipulate and transform them so that they tell us more about themselves. The latter point is well stated by Lave and March, in their remarkably didactic textbook, An introduction to models. They say 'Much of the power, beauty, and pleasure of models comes from inventing and elaborating them' and that 'playing with ideas is fun' (1975: vii, 3). Power, beauty, pleasure and fun: these are attractive qualities, but in general they stand in sharp contrast to those of understanding, knowledge, wisdom and utility which we expect to spring from a study of the real world. I am reminded of Einstein's remark that 'the most incomprehensible thing about the world is that it is comprehensible' (Frank 1950: 118) and would place that alongside another comment of Love and March: 'God has chosen to give the easy problems to the physicists' (1975: 2). If the world that Einstein had in mind includes the social world, and if this is truly difficult but comprehensible, then its study, whether by means of models or by any other means, is necessarily an austere and arduous task. It is therefore not surprising if those who should be struggling with that task are sometimes diverted from it by the pleasure and fun offered by the study of models. On the other hand, given the chronically sorry state of the real world, there is continual pressure political, financial, moral, and intellectual too - for social scientists to turn their attention outward from the ivory tower, to leave the introverted fun of manipulating models and turn to the extroverted task of manipulating real data, or even the real world itself. In other words, there seems to be a dialectical relation between problems posed by the real world, whether these are generated by social causes or by intellectual curiosity, and the content of the intellectual armoury we collect and elaborate within the ivory tower (cf. Dahrendorf 1968: 2 5 6 78). As with armouries of other kinds, the weapons we construct are not necessarily intended for any specified target, or indeed for any target at all. Both terms in the relation have their internal dynamic. The real world continually generates new problems without much regard to our ability to solve them, while in the ivory tower the model-makers and other builders of analytic tools beaver away, largely absorbed in their own scholastic debates. Yet the relation between the real world and activity within the ivory tower persists, and is dialectic or two-directional. Each term impinges on the other. To some extent our perception of problems in the real world is influenced by our ideas about what aspects of the world are changeable, at
MODELLING: FOR REAL OR FUN?
219
least in principle, and what is immutable. Likewise the development and elaboration of intellectual weapons is not entirely an autonomous enterprise but responds in part to pressures from outside. There is thus an oscillation of attention by social scientists, including network analysts. Sometimes the emphasis is on the real world and its troubles. We have to provide solutions, either practical or intellectual, to problems we have not chosen, making the best use we can of existing analytic tools. At other times attention is directed more towards elaborating the armoury, without bothering too much about whether the weapons might, or even could, ever be used, in anger or in earnest, on real world problems. This pleasurable activity, however, eventually attracts the critical attention of colleagues who are frustrated by their inability to understand what the model-makers are doing, or who see support for the pursuit of their own extroverted interests threatened by this flagrant introspective scholasticism. Thus we have for example a complaint directed specifically at the work of some of us here today: A considerable number of them are almost completely involved in technical problems. They are busy refining existing concepts and enlarging the arsenal; they try to make elaborate classifications, and they attempt to inject network analysis with mathematical concepts and procedures in order to give it more 'scientific' tone. Evidently, for these network 'technicians' ... there is not much time for realizing that network analysis is meant to solve anthropological problems (Bax 1978: 222). Price (1981: 304) makes a more pertinent criticism when she complains that Assumptions about the social rules and resources employed in the production and reproduction of social patterns are seldom explicitly discussed in sociological network studies ... many network studies ... take as their point of departure a view of culture as a coherent system of symbols and meanings. Apparently, a passive, adaptive, receptive view of human agency predominates ... Esoteric network analysis can be profoundly misleading when transposed into applied research agendas without specification of the assumptions on which such work is founded.
A well-aimed methodological criticism is made by Piddington in an article about irregular marriages made by Australian Aborigines. Australian Aborigines, as I am sure you know, have themselves produced models of such elegance that Levi-Strauss (1956: 143) was moved to credit them with the invention of sociology. Yet they have suffered more than most ethnic groups from mis-modelling imposed on them by outside observers. Piddington (1970: 341-2) writes that the construction of foreign models 'leads to endless discussions about hypostatized symbolic systems. Though these may provide an entertaining diversion for frustrated mathematicians
22O
A MODEL OF MODELLING
they have the fatal weakness that they do not lead back to the operations which alone could test their validity. This last requirement is an essential characteristic of a valid scientific theory'. In response to this kind of criticism, and to pressures from the real world, there is thus an irregular temporal alternation between one emphasis and the other. Not all social scientists move in synchrony, and there are plenty of reverse eddies. Nevertheless there are also discernible tides, at least on a continental scale. In periods of fervent analytic inventiveness the real world is neglected while new techniques are developed, often as ends in themselves. Some of these techniques are later found to have practical applications, but membership of this subset cannot be predicted in advance. At the other end of the pendulum's swing, in periods of coping pragmatically with urgent practical issues, existing techniques are mobilized and employed with positive effects; the world gets changed as the result of applied social research, and changed for the better, at least from some points of view and in the short run. But unforeseen consequences, and new practical problems, soon upset the applecart of social engineering. The pundits begin to call for better analytic tools, and the pendulum starts to swing back again. In the intervening periods, between tides at slack water, there is sufficient leeway for a fruitful link between empirical research and intellectual innovation. It is then possible to hoist the famous banner of the Tavistock Institute: 'No research without therapy, no therapy without research'. Thus the relation between model building and the real world can best be seen as oscillatory or cyclical, rather than as stable and static. Maybe this oscillation would be sustained by social processes even if the models we build fitted the data from the real world rather more closely than they actually do. In fact the fit, as we all know, is never as close as we would like it to be. Our dissatisfaction with our models, at least whenever the pendulum departs from the introspective end of its swing, generates an additional force for instability. Levins, the biologist I referred to earlier, puts the point well when he says 'There is no single, best all-purpose model. In particular, it is not possible to maximize simultaneously generality, realism, and precision' (1968: 7). All three qualities are desirable, but since we cannot achieve all of them, the way is open in model-building for those cyclical instabilities so beloved of political scientists in their study of multi-cornered contests (Kramer and Hertzberg 1975: 371-4). Indeed, nearer home, Anatol Rapoport (1959: 371) has cautioned social scientists that they 'should not demand realism from the mathematician's models but only pertinence'. Max Black (i960: 45) is rather more pessimistic when he writes that 'more commonly the mathematical treatment of social data leads at best to "plausible topology",... qualitative conclusions concerning distributions, of maxima, minima, and so forth. This result is
MODELLING: FOR REAL OR FUN?
221
connected with the fact that the original data are in most cases at best ordinal in character' (cf. Boulding 1952: 73). Should we accept as inevitable this continual ebb andflowof intellectual style or should we try to restructure our activities so that routine social science, or at least routine network analysis, can proceed in a more straightforward manner? Before tackling this question I would like to refer briefly to physics, that discipline whose image casts so long a shadow over the whole of scientific activity. In physics there is an accepted division of labour between the model-builders, the theoreticians as they are called, who still work with blackboards, pencils and paper, and the experimentalists who wear white coats and devour enormous research grants. This division seems to work, at least as shown by both the enhanced intellectual understanding of the physical world and the enhanced ability to manipulate it which physicists have achieved in the last hundred years or so during which this division has existed. Would a similar division of labour between theoreticians and applied social scientists yield equally spectacular results? Would each branch then be able to pursue wholeheartedly its own objectives, the introvert and extrovert aims of social science as I have labelled them, undistracted by the swing of the pendulum of intellectual fashion? I doubt it. It is always hazardous to rely on analogies with natural science in prescribing for social science, so we should in any case think twice before trying to model ourselves on the physicists. But there is another piece of evidence nearer to hand that also points against adopting this division of labour. Mathematical psychology is a specialism which provides a much more plausible picture than does physics of what might happen if network analysts were to partition themselves into pure and applied. The writings of mathematical psychologists are virtually unintelligible to most of their colleagues in other branches of psychology, so that the division of labour generates not greater productivity but merely greater segregation; the ivory tower becomes a ghetto (Barnes 1972I1: 1422). I think there is a danger that this may happen in sociology, even in network analysis. In any case, we should remember that what looks pure in one context becomes applied in another. In Cambridge, for instance, there is a department rejoicing in the name of 'Applied mathematics and theoretical physics'. The purity of even the physicists is tainted in the eyes of mathematical Brahmins. It seems to me essential that pure and applied should in social science remain a continuum along which individual scientists are encouraged to wander, moved by their curiosity and their conscience and talking to everybody they meet on the way. The evidence for the existence in social science of the oscillatory model I have sketched must, as they say, wait for another occasion, less lighthearted than this one. In brief, I think that in network analysis the initial impetus came from a desire to understand the real world. We had only
222
A MODEL OF MODELLING
very crude analytic tools. Enthusiasts then started to develop tools for their own sake, and the link with the real world was weakened. Some of these tools looked very powerful, but not until a great deal of spadework had been done on both the hardware and software of computers could they be used. My guess is that the tide is beginning to turn again, and that the next few years will see much more attention given to questions of application and falsifiability as we confront powerful analytic tools with more data from the real world. However, the forward glance I want to make is directed not only at what should be our policy for the next five or ten years but also at what consequences there might be for policy in the longer term if the oscillatory metamodel of social science has any validity. If we realize that there is this alternation of intellectual fashion, should we continue as before, naively letting the pendulum swing to and fro but pretending not to notice the periodicity of its movement? Or, following in the footsteps of Maynard Keynes, should we aim to influence the course of events by trying to damp down the swings, by swimming against the tide? In periods of heightened introspective activity should we exhort our colleagues to leave the fun and games and start collecting data from the real world? And, in periods when there are urgent calls for prescriptions on how to deal with, for example, declining morale in inner cities, or excessive use of energy, or inefficiency in the use of social services, should we then demand time to elaborate our intellectual and analytic armoury rather than base our policy recommendations on findings we know to be inconclusive? I suppose there is also a third possibility to consider, of trying to increase the amplitude of the pendulum's swing, but I am unable to think of arguments that would support this kind of strategy. The other two possibilities, however, are strategies rather than tactics, and we would be well advised to have our plans ready for action at both levels if, at the strategic level, we decide after due deliberation to content ourselves with allowing nature or history to take its course without trying to change it. We should also be aware that there are other long-term movements in science that are likely to impinge on our praxis as social scientists in ways that are difficult to predict. One of these movements particularly relevant to network analysis is that from positivism to epistemological popularism (Barnes 1981c). Almost everything published in the field of network analysis is written from a point of view that many of my sociological colleagues, and I think many social anthropologists as well, regard as naively positivist. Network analysts confidently propose to model social behaviour in all its details, adopting the natural science paradigm without reservation and undeterred by any of the warnings sounded by philosophers of science against such superhuman, hubris-risking, ambitions (cf. Putnam 1978: 65; Barnes 1980a: 25-35).
MODELLING: FOR REAL OR FUN?
223
Boldness, naivety and even brashness have a necessary place in the quest for scientific enlightenment, and I think that, at the tactical level, we should probably continue to press on as before. Positivism still has plenty of life left. But we should at least realize that much of the resistance to network analysis springs from a philosophical objection to what is seen as the reification of relations between individuals, and as an attempt to count or measure qualities which are inherently problematic, contestable or negotiable, and therefore uncountable. Maybe at the strategic level we ought to explore the possibility of developing a model of social action that is less a direct import from the natural sciences and which takes adequate account of the reflexive and self-correcting qualities of human behaviour. There are also other processes that we have to treat as unidirectional and which ensure that the pendulum does not simply swing to and fro endlessly in the same trajectory. Mathematical reasoning is pre-eminently cumulative, so that the mathematically inspired models we construct become steadily more sophisticated with each swing. There is a good example of how an advance in mathematical thinking leads to an advance in model building in an unpublished paper by Seidman and Foster (n. d.). They propose a new way of looking at what they call, somewhat oddly, 'social events' and 'pseudo-events'; these are partially overlapping sets of neighbours in social space that are mobilized for a variety of local tasks. They give a discursive account of their analytic procedure and then say I n a more formal treatment, we adopt the concepts, terminology and notation of the mathematical theory of hypergraphs . . . In fact, the hypotheses that were developed in the preceding paragraphs would have been very difficult to conceive and state without the hypergraph formalism'. As you know, it is only fairly recently that Berge's work on hypergraphs (1973) has become available in English. What I found interesting is that, although they claim to be dependent on the formalism of hypergraphs, Seidman and Foster are able to present a lucid description of their analysis without once invoking the technical terms hypergraph and hyperedge (cf. Foster and Seidman 1982: 182-5). We should defend our right to use jargon when necessary, while remembering that we preach to our colleagues, numerate and innumerate alike, not to mystify them but to enlighten them. As Fraser Darling (1947: 77) once defined it, 'good research is orderly thinking plainly said'. That so much activity in many branches of social science consists of exegesis is a sad comment on our failure to follow that precept. Technology is also, in perhaps not quite so strong a sense, a cumulative branch of knowledge and practice. We are all well aware of how formidable the impact has been on our professional activities of advances in the technology of information processing. Most of the models that have been produced in network analysis in the last ten years or so would still be
224
A MODEL OF MODELLING
mere toys, suitable for use only within the ivory tower, were it not for the computers through which alone they can be put to work. From this point of view network analysis, and a good deal of the rest of social science, characterized as they are by a combination of oscillatory and cumulative movements, show themselves to be indeed sciences; they belong to the tradition of Euclid rather than of Plato. There is a significant cumulative movement in philosophy but, at least to the outsider, the oscillations seem to outdistance the forward advance. In the creative arts the great masterpieces are never superseded. Even at this moment, classical Greek tragedies are being acted before full houses more than two thousand years after they were first performed. The plays of Aeschylus and the philosophical works of Plato remain permanently in print, an enduring source of employment for a thriving exegesis industry. The books of Euclid have disappeared long ago from the reading lists of students. One of the diacritical characteristics of science, in contrast to the humanities, is that landmark books go out of print, for they are superseded by later work in which their once new findings are incorporated and surpassed. We should remember too that we, unlike some other social scientists, follow in the steps of Galileo rather than in those of Plato's countryman, Aristotle. For it was the switch from the study of attributes, as practised by Aristotle and characteristic of the classic and mediaeval worlds, to the study of relations which led both Kurt Lewin (1933: 5-10) and Levi-Strauss (1963b: 33, 301, 307) to see in Galileo the pioneer of modern structuralism (cf. Mach i960: 168). Yet the works of Galileo, like those of Euclid, are today sadly but appropriately neglected as compared with those of Aristotle. Thus as our own modest contributions to knowledge slide into early oblivion, gathering dust unsold and uncited, we can console ourselves that this is evidence that we are indeed scientists and not litterati. In such a contrast we may surely be excused for forgetting that though instant oblivion is a necessary indicator of scientific writing, it is not also a sufficient indicator. One aspect of the oscillatory model that has been implicitly recognized by many writers on network analysis is the shift from the use of the notion of network as a metaphor to its employment as a precisely specified model. I think that most writers, myself certainly included, have seen this shift in Whiggish terms, as a forward and irreversible step in the right direction, a cumulative rather than an oscillatory movement. There is some support for this view from Max Black, when he says that 'Perhaps every science must start with metaphor and end with algebra; and perhaps without the metaphor there would never have been any algebra' (i960: 64). However, I now believe that we would be foolish to think that our specialism has become so mature that we can dispense with metaphor. The recent appearance of notions of charm and colour even in the world of particle physics should alert us to the heuristic power of metaphor. Klovdahl
MODELLING: FOR REAL OR FUN?
225
(1981), for example, has drawn attention to the importance of visual images in achieving comprehension of network phenomena and has shown how computer graphics can be put to good use to generate a variety of them. The metaphors of graph theory have been of enormous benefit to the analysis of social networks (Barnes and Harary 1983c). Nevertheless we should avoid becoming hooked inescapably even on nodes and arcs. Max Black again has something pertinent to say: 'The more persuasive the archetype [the model] the greater the danger of its becoming a self-certifying myth'. Yet he goes on to say that 'a good archetype can yield to the demands of experience', which may perhaps be read as support for the oscillatory metamodel. As Braithwaite (1953: 93) reminds us, echoing Edmund Burke, 'The price of the employment of models is eternal vigilance'. If, ready to pay this price and remaining appropriately vigilant, we accept for the moment the validity of the oscillatory metamodel of scientific activity, how should we classify it. Clearly the glaring absence of any real data in this paper places my model firmly in the introspective category; it is an exercise in metamodelling just for sheer fun and pleasure rather than for the contribution it makes to solving the troubles of the world. But if I have tried to live up to Lave and March's dictum that playing with ideas is fun, why then have I suggested that, at least at the present time, we spend too much time in play? To answer that question, we obviously have to turn to Douglas Hofstadter, whose book Godel, Escher, Bach (1979) has thrown so much light on self-referencing and allied phenomena. Hofstadter (1982: 14) makes a distinction between what he calls healthy and neurotic sentences. 'A healthy sentence is one that, so to speak, practises what is preaches, whereas a neurotic sentence is one that says one thing while doing the opposite'. The single word 'Terse' is an example of a healthy sentence, whereas a neurotic sentence is one such as this: 'Proper writing - and you've heard this a million times - avoids exaggeration'. Hofstadter's contrast between healthy and neurotic sentences can easily be generalized to larger entities, such as learned articles and even unlearned keynote addresses. Clearly what we are now reading is a neurotic address. Freud has taught us how much we can learn from neurotic material, so there is no need for me to apologize to you for not offering what used to be called clean healthy fun. Nor should we allow the presence of selfreferencing to draw us into some vortex of infinite regress, some logical black hole from which there is no rational escape. On the contrary, because the structure of my remarks has, I hope, been partly illuminated from within, it may all the more help us to understand what we are doing when we construct a model of nodes and arcs and dream up some data to feed into it. Let me in conclusion draw again on the paper by Deutsch in
226
A MODEL OF MODELLING
which he discusses models through the ages. He describes how, from classical times onwards, one of the metaphors (he calls them models) used to imply notions of process, progress and history has been that of thread, 'whether as the thread of fate, or the thread of an argument, or the thread of a human life. A web woven from these threads is then an obvious extension of this model, implying now, however, the notion of interaction'. Deutsch points out that the German word for 'reality', Wirklichkeit, is related to one of the verbs meaning 'to weave', zu wirken. He refers (1951: 247) to a passage in Goethe's Faust which, in a translation, reads Thus at Time's humming loom 'tis my hand prepares The garment of Life which the Deity wears! (1949: 43) I note, incidentally that, if Deutsch's gloss is correct, this must be one of the earliest metaphorical uses of weaving to indicate social interaction, so that we can add Goethe to the pantheon of our intellectual forebears in network analysis. Later in his article Deutsch quotes again from Faust to illustrate Goethe's statement of the principle of feedback leading to changes in goals, and then adds, in a footnote, 'Faust becomes no more trivial by our knowledge of goal changing feedbacks than a sunrise becomes trivial by our knowledge of the laws of refraction' (1951: 247). Thus we see that, for fun, I have constructed a model that generates the instruction: thou shalt model for real, at least for the next five years or so. I am therefore hoist by my own petard; but all is not lost, for I am rescued by Faust.
POSTSCRIPT: STRUCTURAL AMNESIA
The series of models presented in the foregoing chapters have successively been more abstractly constructed, and more and more remote from the physical world exemplified by the activities of copulation and birth. It may be salutary to end this collection of essays with a reference to a process working in the opposite direction, a process of deconstruction (in a simpler sense than is now fashionable) rather than construction, of forgetting rather than remembering. The following brief statement of what I called structural amnesia is taken from my first professional published paper, dealing with the collection of genealogies in an African matrilineal society. It appeared in 1947 in a journal now hard to find. It is appropriate, I suppose, that the source for a principle of amnesia should itself be forgotten, but I think that the principle is worth re-stating.
One feature that is made immediately apparent by the genealogical method is the extent of structural amnesia among the people under investigation. This phenomenon has been reported from widely separated parts of the world, and may be seen operating in our own society. As has been said 'a man tends to remember only those links in his pedigree which are socially important and which '"place" him at once in the minds of his hearers' (Culwick and Culwick 1935: 180). For example, both the Nuer (Evans-Pritchard 1940a: 199-200) and the Tallensi (Fortes 1944: 370) tend to forget the names of ancestors who do not give their names to units within the lineage structure, while in the genealogies of the strongly patrilineal British peerage, the ascending male lines are much more memorable than the associated distaff lines. The Maori can relate the names of their ancestors for some twenty generations, but each man is interested primarily only in his ascending male line, which serves to 227
228
POSTSCRIPT
determine his social status, and cannot necessarily relate the ascending lines of his fellows (Aginsky and Buck 1940: 199-200). In societies where the ties of kinship are relatively weak, the extent of an individual's knowledge of his ancestry may be very small ... Among the Lamba the matrilineal line of descent is more important than the patrilineal. It was found that whereas the ascending male lines could be traced back for only one or two generations, the ascending female lines could be traced for three to five generations, indicating the greater importance attached to matriliny. An interesting exception to this rule was the case of one headman, anxious to establish his relationship to a chief, who related an unusually long male line of ascent, with the chief as a collateral member. For similar reasons, young children, and particularly young children who have died, tend to be forgotten ...
BIBLIOGRAPHY
Published works by J. A. Barnes Cross-references to similar items are shown within square brackets unless otherwise indicated, the texts of cross-referenced items are substantially identical. Numbers in bold type at the end of an item refer to pages of this book. 1947 1948a 1948b
The collection of genealogies. Rhodes-Livingstone Journal 5:48-5 5 < 228 > Review: Brelsford, Copperbelt markets. Man 48: 46-7 Some aspects of political development among the Fort Jameson Ngoni. African Studies 7: 99-109 1948c The material culture of the Fort Jameson Ngoni. Livingstone: RhodesLivingstone Museum. 14 pp. Occasional Papers of the Rhodes-Livingstone Museum J [1975k] 1949a The village headman in British central Africa III. The Fort Jameson Ngoni. Africa 19: 100-106 [1963b] 1949b Review: Davidson, Northern Rhodesia legislative council. Rhodes-Livingstone Journal 8: 67-9 1949c Review: Kuper, Uniform of colour. Rhodes-Livingstone Journal 8: 69-72 I949d Measures of divorce frequency in simple societies. Journal of the Royal Anthropological Institute 79: 37-62 [cf. 1967c] 1950a Review: Fortes, Social structure. Oxford Magazine 68: 260, 262 1950b [unsigned] The Bamangwato. Africa 20: 151-2 1950c Review: Allan, Studies in African land usage. Africa 20: 164 Mitchell, James Clyde, and Barnes, John Arundel. I95od. The Lamba village: report of a social survey. Capetown: University of Cape Town. 69 pp. Publications from the School of African Studies, n. s. 24 I95oe Review: Duggan-Cronin, Bantu tribes, Mpondo and Mpondomise. Man 50: 112
I95of i95Og 1951a
Review: Childs, Umbundu kinship and character. Man 60: 126-8 African separatist churches [review article: Sundkler, Bantu prophets]. Rhodes-Livingstone Journal 8: 26-30 Review: Holleman, Hera kinship. African Affairs 50: 76 229
23O 1951b 19 5 ic
195id 195 ie 195 if
195ig
BIBLIOGRAPHY Review: Bullock, Mashona and Matebele. Africa 21: 154-5 The perception of history in a plural society; a study of an Ngoni group in Northern Rhodesia. Human relations 4: 295-303 [1951c, 1960m, i962d] < 120-9 = Ch. 7; vi, I 4 4 > Review: Evans-Pritchard, Kinship and marriage among the Nuer. African Affairs 50: 340-1 History in a changing society. Rhodes-Livingstone Journal 11: 1-9 [1951c] Marriage in a changing society: a study in structural change among the Fort Jameson Ngoni. Capetown: Oxford University Press, ix, 136 pp. RhodesLivingstone Papers 20 [i97on] The Fort Jameson Ngoni, In Colson, Elizabeth and Gluckman, Max, eds. Seven tribes of British central Africa. London: Oxford University Press, pp.
194-252 [i959e] 1952
1954b 1954c
Review: Evans-Pritchard, Social anthropology. British Journal of Delinquency 2: 270 Class and committees in a Norwegian island parish. Human Relations 7: 39-58 [1969c, 1977b] < 6 7 - 8 7 = Ch. 4; vi, 9O> Review: Duggan-Cronin, Bantu tribes, III:V. Man 54: 127 Review: Gluckman, Rituals of rebellion. British Journal of Psychology 45:
I954d 1954c
Review: Forde, African worlds. British Journal of Sociology 5: 288-90 Review: Richards, Economic development and tribal change. African Affairs
I954f
Politics in a changing society: a political history of the Fort Jameson Ngoni. Capetown: Oxford University Press, x, 220 pp. [cf. 1967J] Review: Contemporary civilization staff: chapters in Western civilization. Man 55: 10 Review: Trotter, Instincts of the herd. Human Relations 8: 87-8 Review: Gjessing, Mennesket og kulturen. American Anthropologist 57:
1954a
226 53: 342-4 1955a 1955b 1955c
143-4 1955d
1956b
Review: Cory, Indigenous political system of the Sukuma. International Affairs 31: 534 Race relations in the development of southern Africa. In Lind, Andrew William, ed. Race relations in world perspective. Honolulu: Hawaii University Press, pp. 167-86 < i O 3 , i o 6 > Kinship, in Encyclopaedia britannica. London. Vol. 13, pp. 403-9 [cf. 1965a] Ngoni. In Encyclopaedia britannica. London. Vol. 16, p. 405 Seven types of segmentation. Rhodes-Livingstone Journal 17: 1-22 [cf. 1967J] Land rights in two Bremnes hamlets: summary of a communication to the Institute. Man 56: 60, 100 [cf. 1957b] Review: Hanna, Beginnings ofNyasaland and north-eastern Rhodesia. Africa
1956c I956d
Review: Gluckman, Custom and conflict in Africa. Oceania 27: 65-6 Review: Debenham, Nyasaland. Africa 26: 415-16
1955e
1955f !955g 1955I1 1956a
26: 306-9
BIBLIOGRAPHY I956e 1957a 1957b 1957c 1957d 1958a 1958b 1959a 1959b 1959c i959d 19 5 9e
1960a
1960b 1960c I96od 1960c I96of I96og 19 6oh I96oi 1960k
23I
Some problems of African development. 3 pp. Geographical Society of New South Wales Papers 56/25 Anthropology and imperialism [summary]. Union Recorder (Sydney) 37: 5i Land rights and kinship in two Bremnes hamlets. Journal of the Royal Anthropological Institute 87: 31-56 [cf. 1956a] < 9 7 , 146 > Review: Sofer, Jinja transformed; Winter, Bwamba; Mitchell, Yao village. Sociological Review n. s. 5: 135-6 Review: Santos junior, Anthropologia de Mozambique. Oceania 28: 80 Social anthropology in theory and practice. Arts: the Proceedings of the Sydney University Arts Association 1:47-67 < I 3 4 > Review: Read, Ngoni of Nyasaland. Rhodes-Livingstone Journal 23: 68-70 Politics without parties. Man 59: 13-15 Anthropology after Freud. Australasian Journal of Philosophy 37: 14-27 4th South Pacific conference. Journal of the Polynesian Society 68: 144-5 Review: Thomas, Historical notes on the Bisa tribe. American Anthropologist 61: 697 The Fort Jameson Ngoni. In Colson, E. and Gluckman, M., eds. Seven tribes of British central Africa. Manchester: Manchester University Press, pp. 194-252 [i95ig] Indigenous politics and colonial administration, with special reference to Australia. Comparative Studies in Society and History 2: 133-49 [i966d] < 103-19/: Ch. 6 = v , 23 > Review: Gann, Birth of a plural society; Shepperson and Price, Independent African, Man 60: 29-30 Review: Munch, A study of cultural change. American Sociological Review 25: 149 Social anthropology, the university and the public. Australian Journal of Science 22: 371-7 Review: Read, Children of their fathers. Race 1(2): 80-1 Future developments in anthropological studies. Australian Journal of Psychology 12: 21-33 Intensive studies of small communities. Meanjin 19: 201-5 South Pacific commission [letter]. Journal of the Polynesian Society 69: 164 Anthropology in Britain before and after Darwin. Mankind 5: 369-85 Marriage and residential continuity. American Anthropologist 62: 850-66
1960m History in a changing society. In Ottenberg, Simon and Ottenberg, Phoebe, eds. Cultures and societies of Africa. New York: Random House, pp. 318-27 [1951c] 1961a Review: Low and Pratt, Buganda and British overrule. Australian Journal of Politics and History 7: 132-4 Barnes, J. A. and Epstein, Arnold Leonard. 1961b. Comment [on Eisenstadt, Anthropological studies of complex societies]. Current Anthropology 2: 210-11 1961c Review: Man, race and Darwin. Sociological Review n. s. 9: 257-8 196id Physical and social kinship. Philosophy of Science 28: 296-9 < 3 3 >
232 I96ie 196if
I96ig
1962a 1962b 1962c I962d
1963a 1963b 1963c 1963d 1963c
1963b
1964a 1964b 1964c 1965a 1965b
1965c I965d 1966a 1966b
BIBLIOGRAPHY Review: Cunnison, Luapula peoples. Man 61: 165-6 African models in the New Guinea highlands [summary]. In Tenth Pacific Science Congress. Abstracts of symposium papers. Honolulu: Pacific Science Association, pp. 111-12 [cf. 1962a] Law as politically active: an anthropological view. In Sawer, Geoffrey, ed. Studies in the sociology of law. Canberra: Australian National University, pp. 167-96 [cf. 1969^ African models in the New Guinea highlands. Man 62: 5-9 [cf. 196if, 1966c 1971I1] Review: Hanna, Story of the Rhodesias and Nyasaland. American Anthropologist 64: 195-6 Rethinking and rejoining: Leach, Fortes and filiation [review: Leach, Rethinking anthropology]. Journal of the Polynesian Society 71: 403-10 The perception of history in a plural society: a study of an Ngoni group in Northern Rhodesia. Indianapolis: Bobbs-Merrill. 9 pp. Reprint Series in the Social Sciences A-6 [1951c] Review: Falkenberg, Kin and totem. Man 63: 63 Some ethical problems in modern fieldwork. British Journal of Sociology 14: 118-34 [1970k; cf. 1967I1] Bibliographies of Australia and New Guinea [letter]. Man 63: 131 Review: Edwards, Ovimbundu. Africa 33: 377-8 Review: Klausen, Kultur og diffusion. British Journal of Sociology 14: 386 Introduction. In Malinowski, Bronislaw. The family among the Australian aborigines: a sociological study. New edn. New York: Schocken Books, pp. xi-xxx < 3 9 > Social organization: limits of contemporary studies. In Sheils, Helen, ed. Australian Aboriginal Studies. Melbourne; Oxford University Press, pp. 197-210 The village headman in British central Africa: III. The Fort Jameson Ngoni. In Gluckman, M. Order and rebellion in tribal Africa. London: Cohen and West, pp. 160-70 [1949a] Review: Schneider and Gough, Matrilineal kinship. Man 64: 59-60 Physical and social facts in anthropology. Philosophy of Science 31: 294-7 Review: Greenway, Bibliography. Man 64: 196 Review: Schmitz, Grundformer der Verwandtschaft. Journal of the Polynesian Society 74: 242-4 Foreword. In Hiatt, Lester Richard. Kinship and conflict: a study of an aboriginal community in northern Arnhem Land. Canberra: Australian National University, pp. vii-xi Foreword. In Madan, Triloki Nath. Family and kinship: a study of the Pandits of rural Kashmir. Bombay: Asia Publishing House, pp. vii-x Kinship. In Encyclopaedia britannica. London. Vol. 13, pp. 371-7 [minor revision of 1955f; cf. 1974b] Durkheim's Division of labour in society. Man n. s. 1: 158-75 Review: Banton, Social anthropology of complex societies. Man n. s. 1: 407
BIBLIOGRAPHY 1966c 1966c!
I966e
1967a 1967b 1967c
233
Comment [on Henry, Role of the fleldworker]. Current Anthropology 7: 554 Indigenous politics and colonial administration with special reference to Australia. In Wallerstein, Immanuel, ed. Social change: the colonial situation. New York: Wiley, pp. 214-31 [1960a; cf. 1979c] African models in the New Guinea highlands. In Hogbin, Ian and Hiatt, L. R. eds. Readings in Australian and Pacific anthropology. Melbourne: Melbourne University Press, pp. 117-29 [corporated version of 1962a: 1971I1, I 9 7 i q ] Agnation among the Enga [review article: Meggitt, Lineage system of the Mae-Enga]. Oceania 38: 33-43 < 4 i , 56, 61, 64, I 5 9 > Feedback and real time in social enquiry. Australian and New Zealand Journal of Sociology 3: 78-92 < 133-49 = Ch. 8; 32, i 6 7 > Review: Barth, Models of social organization. Anthropological Forum 2:
123-6 ig67d 196 7e
Review: Pradhan, Political system of the Jats. Anthropological Forum 2: 128-9 The frequency of divorce. In Epstein, Arnold Leonard, ed. The craft of social anthropology. London: Tavistock. Social Science Paperbacks 22 pp. 47-99 [revised version of I949d; 1978b: cf. 1967^ Genealogies, in Epstein, A. L., ed. The craft of social anthropology, pp. 1 0 1 27 [cf. 1967c; 1978c] Inquest of the Murngin. London; Royal Anthropological Institute. 50 pp. Occasional paper 26
1967I1
Some ethical problems in modern field work. In Jongmans, Douwe Geert and Gutkind, Peter Claus Wolfgang, eds. Anthropologists in the field. Assen: van Gorcum, Samenlevingen buiten Europa 6, pp. 193-213 [revised version of 1963b] < 134, 143 > 1967J Politics in a changing society. 2nd edn. Manchester: Manchester University Press, x, 245 pp. [i945f, I955h and new preface] < I 2 8 , I 4 4 > 1968a Australian Aboriginals? Or Aboriginal Australians? New Guinea and Australia, the Pacific and South-East Asia 3(1): 43-7 < i o 8 > Barnes, J. A. and Epstein, A. L. 1968b [unsigned] Concepts and problems in institutional comparisons. In Swartz, Marc Jerome, ed. Local-level politics: social and cultural perspectives. Chicago: Aldine, pp. 47-51 [cf. ig68d] 1968c Foreword. In Glasse, Robert Marshall. Huh of Papua: a cognatic descent system. Paris: Mouton. Cahiers de Vhomme n. s. 8, pp. 3-5 I968d Networks and political process. In Swartz, M. J., ed. Local-level politics, pp. IO 7~33 [1969c cf. 1968b] i968e Preface, In Department of Anthropology and Sociology, Australian National University. An ethnographic bibliography of New Guinea. Vol. 1. Author index. Canberra: Australian National University Press, pp. v-vi 1969a Politics, permits, and professional interests: the Rose case. Australian Quarterly 41 (1): 17-31 [1978I] < 2 O > 1969b Graph theory and social networks: a technical comment on connectedness and connectivity. Sociology 3: 215-32 [cf. 1972k]
234 1969c
BIBLIOGRAPHY
1970b
Class and committees in a Norwegian island parish. In French, Robert Mills, ed. The community: a comparative perspective. Itasca, Illinois: Peacock, pp, 122-39 [ I 954 a l Foreword. In Epstein, Arnold Leonard. Matupit: land, politics, and change among the Tolai of New Britain. Canberra: Australian National University Press, pp. vii-x. Networks and political process. In Mitchell, James Clyde, ed. Social networks in urban situations: analyses of personal relationships in central African towns. Manchester: Manchester University Press, pp. 51-76 [i968d] The politics of law. In Douglas, Mary and Kaberry, Phyllis Mary, eds. Man in Africa. London: Tavistock, pp. 99-118 [revision of I96ig; 1971J] Doctors, patients and the social sciences. Murmur: Cambridge University Medical Journal 4(2): 9-12 [1970b] The social sciences in medicine. Synapse: Journal of the Edinburgh Medical
1970c I97od I97oe
Review: Lane, Structuralism. Sociological Review n. s. 18: 283-4 [unsigned] Book note: Litterer, Organizations. Economic Journal 80: 810 Review: Cohen, Modern social theory. American Anthropologist 72: 1099-
I969d
1969c
I969f 1970a
School 20(3): 17-19 [1970a]
100
I97of
Review: Manners and Kaplan, Theory in anthropology. British Journal for the Philosophy of Science 21: 399-401 i97Og Review: Firth et al., Families and their relatives. Man. n. s. 5: 712-13 1970I1 Foreword. In Jain, Ravindra Kumar. South Indians on the plantation frontier. New Haven: Yale University Press, pp. vii-x 19 70J The righthand and lefthand kingdoms of God: a dilemma of Pietist politics [summary]. In International Congress of Anthropological and Ethnological Sciences. Eighth congress, 1968, Tokyo and Kyoto. Proceedings. Vol. 3. Ethnology and archaeology. Tokyo: Science Council of Japan, p. 306 [cf. 1971k] 1970k Some ethical problems of modern fieldwork. In Filstead, William C, ed. Readings in qualitative methodology: firsthand involvement in the social world.
Chicago: Markham, pp. 235-51 [1963b] 1970m Sociology in Cambridge: an inaugural lecture. Cambridge: Cambridge
University Press. 27 pp. < 180-95 = Ch. 11; 12,32 > 19 7on
1970P 1971a 1971b 1971c
Marriage in a changing society: a study in structural change among the Fort Jameson Ngoni. Manchester: Manchester University Press. 136 pp. RhodesLivingstone Papers 20 [195 if] [Contribution to discussion] Report... on the needs for accommodation of the arts and social sciences. Cambridge University Reporter 100:1303-4 [unsigned] Book note: Encel, Equality and authority. Economic Journal 81: 443 Comment [on Jorgensen and Adams, Toward an ethics for anthropologists]. Current Anthropology 12: 342 Review: Buchler and Nutini, Game theory in the behavioral sciences. Mankind 8: 71
BIBLIOGRAPHY
235
Bambrough, R. and Barnes, J. 1971c!. Nature and human nature - Renford Bambrough talks to John Barnes about sociology. Listener 86: 172-5 [cf. Studying other people: luxury or necessity? Science or Society?: CSSRS Bulletin 3: 9-11 I97if Time flies like an arrow. Man n. s. 6: 537-52 < 150-68 = Ch. 9; I 3 4 > I 9 7 i g Review: Read, Ngoni of Nyasaland. Man. n. s. 6: 711 1971I1 African models in the New Guinea highlands. Indianapolis, Indiana: BobbsMerrill. 8 pp. Reprint Series in Social Sciences A395 [1962a, 1966c] 1971J The politics of law. In Douglas, M. and Kaberry, P. M., eds. Man in Africa. Garden City, N. Y.: Doubleday, Anchor A816 pp. 100-19 [1969(1 1971k The righthand and lefthand kingdom of God: a dilemma of Pietist politics. In Beidelman, Thomas Owen, ed. The translation of culture: essays ofE. E. Evans-Pritchard. London: Tavistock, pp. 1-17 [cf. 1970J] < 88-102 = Ch. 5: I5O> 1971m Three styles in the study of kinship. London: Tavistock, and Berkeley; University of California Press, xxiv, 318 pp. [cf. 1973c] 19 7 i n Agnatic taxonomies and stochastic variation. Anthropological Forum 3: 3 12 < 5 6 - 6 6 = Ch. 3; 45 > I 9 7 I P [unsigned] Book note: White, Chains of opportunity. Economic Journal 81: 1052 I 9 7 i q African models in the New Guinea highlands. In Langness, Lewis Leroy and Weschler, John C, eds. Melanesia: readings on a culture area. Scranton: Chandler, pp. 97-107 [1966c] 19 7 i r [Contribution to discussion] Report... on the establishment of a Diploma in Latin-American studies. Cambridge University Reporter 102: 325 1972a Review: Gulliver, Neighbours and networks. Man. n. s. 7: 158-9 1972b Helping and cheating. Cambridge Review 93: 132-4 1972c Review: Abell, Model building in sociology. Economic Journal 82: 805-6 I972d Preface. In Brown, Paula. The Chimbu. Cambridge, Mass.: Schenkman, pp. vii-ix [1973b] I972e Social networks. Reading, Mass.: Addison-Wesley. 29 pp. Module in Anthropology 26 [1973J] I972f Comment [on Mukherjee, Analysis of variations in family structures]. Current Anthropology 13: 435 Bambrough, R. and Barnes, J. I972g. Renford Bambrough talks to John Barnes about sociology. In Bambrough, R. Nature and human nature. Tokyo: Shinozaki Shorin, pp. 68-87, 99-102 [i97id, with notes] 1972I1 Review: Blalock, Causal models in the social sciences. Economic Journal 82: 1420-3 1972J [unsigned] Book note: Horowitz, Use and abuse of social science. Economic Journal 82: 1130 1972k Structural analysis of networks and graphs. In Australian UNESCO Seminar. Mathematics in the social sciences in Australia: University of Sydney, May 1968. Canberra; Australian Government Publishing Service, pp. 33-35 [1969b, with discussion] I97ie
236 1972m
BIBLIOGRAPHY [Contribution to discussion] Report . . . on the assignment of the Pitt
professorship . . . Cambridge University Reporter 102: 986 I972n
[Contribution to discussion] Report . . . on the regulations for the Econ-
omics Tripos. Cambridge University Reporter 103: 248-9 1973a Review: Liu, Murngin. Current Anthropology 14: 105 1973b Foreword. In Brown, P. The Chimbu. London: Routledge & Kegan Paul, pp. viii-ix [i972d] 1973c Three styles in the study of kinship. Paperback edn. Berkeley: University of California Press, xxiv, 318 pp. Campus 97 [minor corrections to 1971m] 1973d Review: Anderson and Wadel, North Atlantic fisherman. Man. n. s. 8: 330-1 1973c Review: Andreski, Social sciences as sorcery. Economic Journal 83: 348 I973f Review: Sahlins, Stone age economics. Economic Journal 83: 674 T 973g Genetrix : genitor :: nature : culture? Goody, Jack, ed. The character of kinship. Cambridge: Cambridge University Press, pp. 61-73 [cf. 1973I1] < 29-43 = Ch. 1: 150, 169 > 1973I1 Bibliography of the writings of Meyer Fortes. In Goody, J., ed. The character of kinship, pp. 231-5 [cf. I973g] X 973J Social networks. Module in Anthropology 26. In Current topics in anthropology: theory, methods, and content. Reading, Mass.: Addison-Wesley. 29 pp. [i972e] 1974a Positivism in the womb. Cambridge Anthropology 1(2): 48-52 < 30, 150, I76> 1974b Kinship. In Encyclopaedia britannica: macropaedia. Vol. 10, pp. 477-85 [major revision of I965d] 1974c Review: Burr, Theory construction and the sociology of the family. Man n. s. 9: 324-5 I974d Review: Spear, Zwangendaba's Ngoni. Africa 44: 95 I974e Review: Strathern, Rope of moka. Modern Asian Studies 8: 429-32 I974f Review: Epstein and Penny, Opportunity and response. Economic Journal 84: 459 I974g Review: Boissevain and Mitchell, Network analysis. Man n. s. 9: 497-9 1974I1 Review: Ramsoy, Norwegian society. Scandinavica 13: 145-7 [cf. 1974J] 1974J The welfare state of Norway [review: Ramsoy, Norwegian society]. Times Literary Supplement 1974: 1304 [cf. 1974b] 19 74k Review: de Kadt and Williams, Sociology and development. Economic Journal 84: 1071 1974m Review: Swedish Council for Social Science Research, Social science research in Sweden. Economic Journal 84: 1101 I974n Review: Hudson, Padju Epat. Modern Asian Studies 8: 559-60 X 974P [Contribution to discussion] Report . . . on the establishment of a Professorship and Department of community medicine. Cambridge University Reporter 105: 359-60 1975a On the cross-cultural survey method. Current Anthropology 16: 296-7 1975b Foreword. In Singarimbun, Masri. Kinship, descent and alliance among the Karo Batak. Berkeley: University of California Press, pp. ix-xii
BIBLIOGRAPHY 1975c 1975c!
237
Review: Horton and Finnegan, Modes of thought. Sociology 9: 165-6 Review: Pausewang, Methods and concepts of social research. Africa 45:
104-5 i975e I975f
Review: Herbst, Socio-technical design. Economic Journal 85: 221 [unsigned] Book note: Boulding, Collected papers, vol. 4. Economic Journal 85:476 I 9 7 5 8 [unsigned] Book note: Oxaal et alM Beyond the sociology of development. Economic Journal 85: 712 19 75h [untitled letter] Network: Newsletter of the British Sociological Association 3: 3-4 1975J Comment [on Strauss and Orans, Mighty sifts]. Current Anthropology 16: 585 1975k The material culture of the Fort Jameson Ngoni. In Rhodes-Livingstone Museum. The occasional papers of the Rhodes-Livingstone Museum. Manchester: Manchester University Press, pp. 1-20 [1948c] 1975m Durkheim's Division of labour in society. In Jesser, Clinton Joyce, ed. Social theory revisited. Hinsdale, Illinois: Dryden, pp. 241-61 [1966a] I 9 7 5 n [Contribution to discussion] Report... on the regulations for Part IA and Part IB of the Medical Sciences Tripos. Cambridge University Reporter 105: 816 1976a The anthropologist's calling [review: Burridge, Encountering Aborigines]. Times Literary Supplement 1976: 440 1976b Review: Ballonoff, Genealogical mathematics. Man n. s. 11: 129-30 1976c Comment [on Chilungu, Issues in the ethics of research method]. Current Anthropology 17: 467 1977a The ethics of inquiry in social science: three lectures. Delhi: Oxford University
1977b
Press. 67 pp. < 2 i i > Class and committees in a Norwegian island parish. In Leinhardt,
1977c
Press, pp. 2 3 3 - 5 2 [ 1 9 5 4 a ] [Contribution to discussion] Report . . . on the establishment . . . of
Samuel, ed. Social networks: a developing paradigm. New York: Academic
Professorships and Readerships ad hominem. Cambridge University
Reporter 107: 632-3 1978a 1978b
1978c 19 78d
1978c
Review: Boissevain, Friends of friends. American Journal of Sociology 83: 1542-4 The frequency of divorce. In Epstein, A. L., ed. The craft of social anthropology. Delhi: Hindustan Publishing Corporation. Studies in Sociology and Social Anthropology 2, pp. 47-99 [1967c; cf. 1978c] Genealogies. In Epstein, A. L., ed. The craft of social anthropology, pp. 1 0 1 27 [1967ft cf. 1978b] Neither peasants nor townsmen: a critique of a segment of the folk-urban continuum. In Barth, Fredrik, ed. Scale and social organization. Oslo: Universitetsforlaget, pp. 13-40 < 6 8 > Review: Lorrain, Reseaux sociaux et classifications sociales. American Anthropologist 80: 953-4 < I 7 9 >
238 I978f
BIBLIOGRAPHY Politics, permits, and professional interests: the Rose case. In Appell, George Nathan. Ethical dilemmas in anthropological inquiry: a case book.
Waltham, Mass.: Crossroads Press, pp. 174-83 [1969a] 1979a
Who should know what? Social science, privacy and ethics. Harmondsworth:
Penguin, 232 pp. [cf. 1980a] 1979b
Review: Owusu, Colonialism and change. British Journal of Sociology 30:
1979c
Culture contact and political change. In Deaking University, Open Campus Program, School of Social Sciences, ed. Imperialism: clash of cultures. Deakin University, Victoria: Deakin University. SS 10j. Reader A. Weeks 1-7, p. 41 [part of I966d] On the survival of communities. Current Anthropology 20: 612-13 Network analysis: orienting notion, rigorous technique or substantive field of study? In Holland, Paul W. and Leinhardt, S., eds. Perspectives on social network research. New York: Academic Press, pp. 403-23. Fortes, Meyer. In Sills, David L., ed. International encyclopedia of the social sciences. Vol. 18: Biographical supplement. New York: Free Press, pp. 1 9 5 7 [Contribution to] Kemper, Steven et al. Comments on Austin's Symbols and culture. Social Analysis 3: 64-5 Who should know what? Social science, privacy and ethics. Cambridge: Cambridge University Press. 232 pp. [minor corrections to 1979a] < 5 ,
261-2
I979d 1979c
I979f
!979g 1980a
2O,
I33,
211,
222>
1980b 1980c
Kinship: nature or culture? American Anthropologist 82: 128-9 Kinship studies: some impressions of the current state of play. Man n. s. 15: 293-303 < 169-80 = Ch. 10; 3 0
I98od 1980c I98of
Review: Bulmer, Censuses, surveys and privacy. Sociology 14: 489-90 Review: Goldschmidt, Uses of anthropology. Man n.s. 15: 555-6 Numbering Norwegians (review: Sundt, Marriage]. Times Literary Supplement 1980: 1340 Laud Humphreys and Tearoom trade. Australian Anthropological Society Newsletter 10: 35-7 Review: Cherns, Using the social sciences. Knowledge: Creation Diffusion Utilization 2: 448-51 Professionalism in British sociology. In Abrams, Philip et al., eds. Practice and progress: British sociology 1950-1980. London: Allen and Unwin, pp. 13-24 < o o o , 000 > [Contribution to discussion] Report... on certain amendments of Statute D. Cambridge University Reporter 111: 626 Review: Ackroyd and Hughes, Data collection in context. Network 22: 16 Ethische en politieke compromissen in sociaal onderzoek. In Bouw, Carolien et al., eds. Hoe weet je dat? Wegen van sociaal onderzoek. Amsterdam: Arbeiderspers en Wetenschappelijke Uitgeverij, pp. 27-37 [ X 984] Social science in India: colonial import, indigenous product, or universal truth? In Fahim, Hussein, ed. Indigenous anthropology in non-Western
1981a 1981b 1981c
198 id 1982a 1982b
1982c
countries. Durham, N. C: Carolina Academic Press, pp. 19-34 < I 9 6 -
BIBLIOGRAPHY
239
212 = Ch. 12: I I >
1982c! 1983a 1983b
William Edward Hanley Stanner 1905-1981. Canberra Anthropology 5d): 85-7 Lying: a sociological view. Australian Journal of Forensic Sciences 15: 1528 Modelling: for real or for fun? Connections 6(1): 15-21 < 2 i 5 - 2 6 = Ch. 13:2,
25>
Barnes, J. A. and Harary, Frank. 1983c. Graph theory in network analysis. Social Networks 5: 235-44 < 2 i , 215, 217, 2 2 4 > 1983d Foreword. In Hage, Per and Harary, F. Structural models in anthropology. Cambridge: Cambridge University Press. Cambridge Studies in Social Anthropology 46, pp. ix-xi O'Connor, Michael E. and Barnes, J. A. 1983c Bulmer on pseudo-patients: a critique. Sociological Review n. s. 31: 753-8 Why bother about ethics? Australian Anthropological Society Newsletter 21: 33-7 Memorandum. In House of Commons. Home affairs committee, Session 1982-83. Second report: ethnic and racial questions in the census. Vol. Ill: Appendices. London: HMSO. HC33-IU, pp. 144-5 1984 Ethical and political compromises in social research. Wisconsin Sociologist 21: I O O - I I [1982b]
1985a
Foreword. In Lukhero, Matshakaza Blackson. Ngoni nc'wala ceremony. Lusaka: National Education Company of Zambia, pp. vii-viii 1985b Review: Stocking, Observers observed. American Ethnologist 12: 148-9 1985c Review: Lawrence, Garcia. Pacific History Bibliography and Comment 1985: 45-6 1986 Review: Gross and Rayner, Measuring culture. Man n. s. 21: 760-1 1987a Academy lecture and symposium. ASS A News (Canberra) 6(1): 5-7 198 7b Citation classic commentary. Current Contents {Social & behavioral sciences and Arts & humanities) 9(23): 16 1987c Educational codes of professional ethics. Nexus 30: 15-16 Barnes, J. A. and Drucker-Brown, Susan. I987d. The writings of Meyer Fortes: a supplementary list. Cambridge Anthropology 11(2): 70-7 1987c Review: Cassell and Jacobs, Handbook on ethical issues. Mankind 17: 2 3 8 40
I987f 1988a
1988b
The gentle art of necrology. Cambridge Anthropology 12(1): 52-69 Taking stock and looking forward. In Berndt, Ronald Murray and Tonkinson, Robert, eds. Social anthropology and Australian Aboriginal studies: a contemporary overview. Canberra: Aboriginal Studies Press, pp. 269-79 Edward Evan Evans-Pritchard 1902-1973. Proceedings of the British Academy 73: 447-89 < 8 8 >
24O
BIBLIOGRAPHY Other works cited
Numbers in bold type at the end of each entry refer to pages in this book. Abbi, Behari Lai and Saberwal, Satish 1969 editors. Urgent research in social anthropology: proceedings of a conference. Simla: Indian Institute of Advanced Study, x, 235 pp. Transactions 10 < I 7 7 , 2O7> Abrahams, Israel 1924. Studies in Pharisaism and the Gospels. 2nd series. Cambridge: Cambridge University Press, x, 226 pp. < 4 i > Abrams, Philip 1968. The origins of British sociology: 1834-1914. An essay with selected papers. Chicago: University of Chicago Press, x, 304 pp. < 182 > Aeschylus 1953. The Eumenides. In Oresteia, translated by Lattimore, Richmond. Chicago: University of Chicago Press, pp. 133-71 < 3 8 > Aginsky, Bernard Willard and Buck, Peter Henry 1940. Interacting forces in the Maori family. American Anthropologist 42: 195-210 < 2 2 8 > American Anthropological Association 1967. Headline aids scuttling offieldwork. AAA Fellow Newsletter 8(2): 2-3 < 139 > 1976. Court rules anthropologist defamed by media in reports of involvement in Thai research center. Anthropology Newsletter 17(2): 3 American Statistical Association 1974. Report of the ASA conference on surveys of human populations. American Statistician 28: 30-4 < 2 i i > Anderson, Michael 19 71, Introduction. Anderson, M., ed. Sociology of the family: selected readings. Harmondsworth: Penguin, pp. 7-14 < I 7 4 > Arensberg, Conrad Maynadier 1957. Anthropology as history. In Polanyi, Karl et al., eds. Trade and market in the early empires. Glencoe, 111.: Free Press, pp. 9 7 113 < I 5 8 > Aristotle 1910. Historia animalium, translated by Thompson, D'Arcy Wentworth. In Smith, J. A. and Ross, W. D., eds. The Works of Aristotle. Vol. 4. Oxford: Clarendon Press < 35 > 1912. De generatione animalium, translated by Platt, Arthur. In Smith, J. A. and Ross, W. D., eds. The works of Aristotle, vol. 5 Oxford: Clarendon Press < 3 5 , 38> Armstrong, Wallace Edwin 1928. Rossel Island: an ethnological study. Cambridge: Cambridge University Press, xxviii, 274 pp. < 72 > Ashby, Eric 1966 Universities: British, Indian, African: a study in the ecology of higher education. London: Weidenfeld and Nicolson. xiii, 588 pp. < 2 o o > Ashby, William Ross 1956. An introduction to cybernetics. London: Chapman and Hall, ix, 295 pp. < I 3 6 > Ashley-Montagu, Montague Francis 1937. Coming into being among the Australian Aborigines: a study of the procreative beliefs of the native tribes of Australia. London: Routledge. xxxv, 362 pp. < 3 4 > 1949. Embryological beliefs of primitive peoples. Ciba Symposia 10: 994-1008 Attenborough, Frederick Levi 1922 Editor. The laws of the earliest English kings. Cambridge: Cambridge University Press, xii, 256 pp. < 3 7 > Auden, Wystan Hugh 1940. Another time. London: Faber and Faber. 125 pp.
BIBLIOGRAPHY
24I
Austin, Colin Russell and Walton, Arthur i960. Fertilisation. In Parkes, Alan Sterling, ed. Marshall's Physiology of reproduction. 3rd edn. Vol. 1, part 2. London: Longmans Green, pp. 310-416 < 3 6 > Bailey, Frederick George 1959. For a sociology of India? Contributions to Indian sociology 3: 88-101 < 2 o 6 > Ballonoff, Paul Alan 1974a editor. Mathematical models of social and cognitive structures. Urbana; University of Illinois Press, xxi, 122 pp. 1947b editor. Genealogical mathematics. Paris: Mouton. 311 pp. Banton, Michael 1964. Anthropological perspectives in sociology. British Journal of Sociology 15: 95-112 < i 8 8 > Barkow, Jerome H. 1978. Social norms, the self, and sociobiology: building on the ideas of A. I. Hallowell. Current Anthropology 19: 99-118 < I 7 6 > Barnes, Stanley Barry 1969. Paradigms - scientific and social. Man n. s. 4: 94-102 Barth, Fredrik 1966. Models of social organization. London: Royal Anthropological Institute, vi, 33 pp. Occasional Papers 23 < I 3 6 > Bax, Mart 1978. Figuration analysis: a better paradigm for networkers: with an illustration from Ireland. Anthropological Quarterly 51: 221-30 < 2 i 9 > Beals, Ralph Leon and others 1967. Background information on problems of anthropological research and ethics. AAA Fellow Newsletter 8(1): 1-13 Beattie, John Hugh Marshall 1964. Kinship and social anthropology. Man 64: 101-3
Becker, Howard Saul 1964. Problems in the publication offieldstudies. In Vidich et al. 1964, pp. 266-84 < I 3 8 > 1967. Whose side are we on? Social Problems 14: 239-47 < 18 > 1983. Freshman English for graduate students: a memoir and two theories. Sociological Quarterly 24: 575-88 < 7 > Benirschke, Kurt 1970. Spontaneous chimerism in mammals: a critical review. Current Topics in Pathology 51: 1-61 < 3 6 > Berelson, Bernard and Steiner, Gary Albert 1964. Human behaviour: an inventory of scentific findings. New York: Harcourt, Brace and World, xxiii, 712 pp. < 2 4 > Berge, Claude 1973. Graphs and hypergraphs. Amsterdam: North Holland, xiv, 528 pp. North Holland Mathematical Library 6 < 2 2 3 > Bernard, Jessie and others 1965-6. Communications to the editor. American Sociologist 1(1, 2, 4 & 5) < I 3 9 > Berreman, Gerald Duane 1962. Behind many masks: ethnography and impression management in a Himalayan village. Ithaca, N. Y.: Society for Applied Anthropology. 24 pp. Monograph 4 < 1 4 0 Beteille, Andre 1965. Caste, class, and power: changing patterns of stratification in a Tanjore village. Berkeley: University of California Press. 238 pp. < I 3 8 > Black, Max i960. Models and archetypes. In Boewe, Charles Ernest and Nichols, Roy Franklin, eds. Both human and humane: the humanities and social sciences in graduate education. Philadelphia: University of Pennsylvania, pp. 39-65
Bohm, David 1957. Causality and chance in modern physics. London: Routledge and
242
BIBLIOGRAPHY
Kegan Paul xi, 170 pp. < I 3 7 > Bondevik, Kjell 1965. Kristelig folkeparti. In 0isang, Per. ed. De 7 politiske partier i Norge. Oslo: Gyldendal norsk forlag, pp. 67-75 < 9 3 > Bonjean, Charles, Hill, Richard J. and McLemore, S. Dale 1965. Continuities in measurement, 1959-1963. Social Forces 43: 532-5 < i 4 8 > Bose, Nirmal Kumar 1975. The structure of Hindu Society. New Delhi: Orient Longman. 171 pp. < 2 O 2 > Boulding, Kenneth Ewart 1952. Economics as a social science. In Minnesota University
Social Science Research Center. The social sciences at
mid-century.
Minneapolis: University of Minnesota Press, pp. 70-83 < 2 2 i > Braithwaite, Richard Beven 1 9 5 3 . Scientific explanation: a study of the function of
theory, probability and law in science. Cambridge: Cambridge University Press. xii, 378 pp. < 2 2 5 > Brodbeck, May 1959. Models, meaning, and theories. In Gross, Llewellyn Zwicker, ed. Symposium on sociological theory. New York: Harper and Row, pp. 373-403 Brown, Richard Harvey 19 77. A poetic for sociology: toward a logic of discovery for the
human sciences. Cambridge: Cambridge University Press, x, 302 pp. < n > Busia, Kofi Abrefa 1 9 5 1 . The position of the chief in the modern political system of
Ashanti. London: Oxford University Press, xii, 233 pp. < H 2 > Cambridge Anthropological Expedition to Torres Strait 1901-35. Reports. Haddon, Alfred Cort, ed. Cambridge: Cambridge University Press. 6 vols. < 186 > Chappie, Eliot Dismore and Coon, Carleton Stevens 1942. Principles of anthropology. New York: Holt, xi, 718 pp. < 7 2 > Cheng, Lucie and So, A. 1983. The reestablishment of sociology in the PRC: toward the sinification of Marxian sociology. Annual Review of Sociology 9: 471-98 Clinard, Marshall Barron and Elder, Joseph W. 1965. Sociology in India: a study in the sociology of knowledge. American Sociological Review 30: 581-7 < 2 0 0 > Cohen, Abner 1969. Political anthropology: the analysis of the symbolism of power relations. Man n. s. 4: 215-35 < i 8 8 > Cohen, Percy Saul 1971. Review: Barnes, Sociology in Cambridge. Man n. s. 6: 7 2 1 2
Cole, Francis Joseph 1930. Early theories of sexual generation. Oxford: Clarendon Press, x, 230 pp. < 3 9 > Coleman, James Samuel 1 9 6 4 . Introduction to mathematical
sociology. New York:
Free Press of Glencoe. xiv, 554 pp. < 189 > Collie, Alexander 1830. On some particulars connected with the natural history of the kangaroo. Zoological Journal 5: 238-41 Collier, John 1947. The Indians of the Americas. New York: Norton. 326 pp. Colson, Elizabeth 1948. Modern political organization of the Plateau Tonga. African Studies 7: 85-98 < H 4 > 1962. The Plateau Tonga of Northern Rhodesia: social and religious studies. Manchester: Manchester University Press, xv, 237 pp. < H 4 > Cranz, Ferdinand Edward 1 9 5 9 . An essay on the development of Luther's thought on
justice, law and society. Cambridge, Mass.: Harvard University Press, xviii, 197
BIBLIOGRAPHY
243
pp. Harvard Theological Studies 19. < 9 7 > Culwick, Arthur Theodore and Culwick, G. M. 1935. Ubena of the rivers. London: Allen and Unwin. 444 pp. < 2 2 8 > Cunninson, Ian George 1 9 5 1 . History on the Luapula: an essay on the historical
notions of a central African tribe. Cape Town: Oxford University Press, viii, 42 pp. Rhodes-Livingstone Papers 21 < i 6 3 > 1957. History and genealogies in a conquest state. American Anthropologist 59: 20-31 < i 6 3 > Dahrendorf, Ralf 1968. Essays in the theory of society. London: Routledge & Kegan Paul, xiv, 300 pp. Damle, Yashwant Bhaskar 1967. For a theory of Indian sociology. In Unnithan, Thottamon Kantankesavan Narayanan et al., eds. Sociology in India. New Delhi: Prentice-Hall, pp. 85-103 Darling, Frank Fraser 1 9 4 7 . Natural history in the Highlands and Islands. London:
Collins, xv, 303 pp. Darwin, Charles 1 8 7 5 . The variation of animals and plants under domestication. 2nd
edn. Vol. 1. London: John Murray, xiv, 473 pp. < 3 6 > Davis, Allison, Gardner, Burleigh Bradford and Gardner, Mary R. 1941. Deep South: a social anthropological study of caste and class. Chicago: University of Chicago
Press, xv, 558 pp. < 7 4 > de Levervanche, Marie M. 1967-8. Descent, residence and leadership in the New Guinea highlands. Oceania 38: 134-58, 163-89 Deare, George Russel see A Durban man Deitchman, Seymour J. 1 9 7 6 . The best-laid schemes: a tale of social research and
bureaucracy. Cambridge, Mass.: MIT Press, xv, 483 pp. < I 3 9 > Deutsch, Karl Wolfgang 1951. Mechanism, organism, and society: some models in natural and social science. Philosophy of Science 18: 230-52 < 2 i 7 , 226 > Devons, Ely 1 9 5 0 . Planning in practice: essays in aircraft planning in wartime.
Cambridge: Cambridge University Press, viii, 232 pp. < I 4 > Devons, Ely and Gluckman, Max 1964. Conclusion: modes and consequences of limiting a field of study. In Gluckman, M., ed. Closed systems and open minds. Edinburgh: Oliver & Boyd, pp. 158-261 < I 7 3 > Dobzhansky, Theodore 1970. Heredity. In Encyclopaedia britannica. London. Vol. 11, pp. 419-27 < 3 6 , 3 7 > Douglas, Mary 1967. The meaning of myth, with special reference to 'La geste d'Asdiwal'. In Leach 1967c, pp. 49-69 < I 5 7 , 1 6 2 > 1979. Virgin birth. Man n. s. 4: 133-4 < 3 4 > Dube, Shyama Charan 1 9 7 3 . Social sciences in changing society: D. N. Majumdar
lectures 1972. Lucknow: Ethnographic and Folk Culture Society of U. P., Department of Anthropology, Lucknow University. 59 pp. < 2 O 4 > Dumont, Louis 1957. [unsigned] For a sociology of India. Contributions to Indian Sociology 1: 7-22 < 2 o 6 > 1966. A fundamental problem in the sociology of caste. Contributions to Indian Sociology 9: 17-32 < 2 o 6 , 207 > Dumont, Louis and Pocock, David Francis i960. For a sociology of India: a rejoinder to Dr Bailey. Contributions to Indian Sociology 4: 82-9 < 2 o 6 > A Durban man [Deare, George Russel] 1929. Eighteen months with the last of the
244
BIBLIOGRAPHY
slave raiders. Natal Weekend Advertiser 6 April - n May [pagination of typescript in National archives of Zimbabwe] < 123 > Easton, David 1959. Political anthropology. In Siegel, Bernard Joseph, ed. Biennial review of anthropology 1959. Stanford: Stanford University Press, pp. 210-62 1966. Categories for the systems analysis of politics. In Easton, D., ed. Varieties of political theory. Englewood Cliffs, N. J.: Prentice-Hall, pp. 143-54 < I 5 9 > Elkin, Adolphus Peter 1951. Reaction and interaction: a food gathering people and European settlement in Australia. American Anthropologist 53: 164-86 <no> Engels, Friedrich 1884. Der Ursprung de Familie, des Privateigenthums und des Staats.
Hottingen-Ziirich: Verlag der Schweizerischen Volksbuchhandlung. 146 pp. Evans-Pritchard, Edward Evan 1932. Heredity and gestation as the Azande see them. Sociologus 8: 400-14 1934. Levy-Bruhl's theory of primitive mentality. Bulletin of the Faculty of Arts (University of Egypt) 2: 1-36 1937. Witchcraft, oracles and magic among the Azande. Oxford: Clarendon Press.
xxv, 558 pp. < i 5 i > 1939. Nuer time reckoning. Africa 12: 189-216 < i 5 i > 1940a. The Nuer. Oxford: Clarendon Press, xiii, 271 pp. 1940b. The Nuer of the Southern Sudan. In Fortes and Evans-Pritchard 1940b, pp. 272-96 1945. Some aspects of marriage and the family among the Nuer. Livingstone:
Rhodes-Livingstone Institute. 65 pp. Rhodes-Livingstone Papers 11 1950. Social anthropology: past and present: the Marett lecture, 1950 Man 50: 118-24
1951. Social anthropology. London: Cohen and West, vii, 134 pp. < 11 > 1956. Nuer religion. Oxford: Clarendon Press, xii, 336 pp. 1961. Anthropology and history. Manchester: Manchester University Press. 22 pp. < i 5 i , I57> 1962. Essays in social anthropology. London: Faber and Faber. 233 pp. Fallers, Lloyd Ashton 1956. Bantu bureaucracy: a study of integration and conflict in the political institutions of an east African people. Cambridge: Heffer. xiv, 282 pp.
< H 2 , I74> Fay, Brian and Moon, J. Donald 1977. What would an adequate philosophy of social science look like? Philosophy of the Social Sciences 7: 209-27 < 2 i , 24 > Feil, Daryl Keith 1988. The evolution of highland Papua New Guinea societies.
Cambridge: Cambridge University Press. 286 pp. Firth, Raymond William 1944. The future of social anthropology. Man 44: 19-22 1956 editor. Two studies of kinship in London. London: Athlone Press. 94 pp. London School of Economics Monographs on Social Anthropology J 5 < 173 >
Firth, R., Hubert, Jane and Forge, Anthony 1969. Families and their relatives: kinship in a middle-class sector of London: an anthropological study. London:
Routledge & Kegan Paul, xv, 476 pp. < 173 >
BIBLIOGRAPHY
245
Fluehr-Lobban, Carolyn 1979. A Marxist reappraisal of the matriarchate. Current Anthropology 20: 341-59 < I 7 6 > Ford, Clellan Stearns 1945. A comparative study of human reproduction. New Haven: Yale University Press. 111 pp. Publication in Anthropology 32 < 34 > Forde, Cyril Daryll 1951. Marriage and the family among the Yakd in south-eastern Nigeria. 2nd edn. London: Lund Humphries. 121 pp. London School of Economics Monographs on Social Anthropology 5 < 161 > Forell, George Wolfgang 1954. Faith active in love: an investigation of the principles underlying Luther's social ethics. New York: American Press. 198 pp. < 9 6 > Fortes, Meyer 1944. The significance of descent in Tale social structure. Africa 14: 362-85 < 2 2 8 > 1945. The dynamics of clanship among the Tallensi. London: Oxford University Press, xx, 270 pp. < 8 3 > 1949a. The web of kinship among the Tallensi. London: Oxford University Press. xiv, 358 pp. < 4 i , 72> 1949b. Time and social structure: an Ashanti case study. In Fortes, M., ed. Social structure. Oxford: Clarendon Press, pp. 54-84 < i 6 2 , 165 > 1953. The structure of unilineal descent groups. American Anthropologist 55: 17-41 < 3 i , 56> 1957. Malinowski and the study of kinship. In Firth, Raymond, ed. Man and culture. London: Routledge & Kegan Paul, pp. 157-88 1958a. Introduction. In Goody, Jack, ed. The developmental cycle in domestic groups. Cambridge: Cambridge University Press. Cambridge Papers in Social Anthropology 1, pp. 1-14 < I 5 9 > 1958b. Thinking about society. Cambridge Review 79: 596-8 1969. Kinship and the social order. Chicago. Chicago: Aldine, xii, 347 pp. 1970. Time and social structure and other essays. London: Athlone Press, xii, 287 pp. London School of Economics Monographs on Social Anthropology 40 < 6 4 > 1983. Rules and the emergence of society. London: Royal Anthropological Institute, iv, 52 pp. Occasional Paper 39 Fortes, M. and Evans-Pritchard, E. E. 1940a. Introduction. In Fortes and EvansPritchard 1940b, pp. 1-23 < i , i o 6 > 1940b editors. African political systems, London: Oxford University Press, xxiii, 302 pp. Foster, Brian L. 1979. Formal network studies and the anthropological perspective. Social Networks 1: 241-55 < I 7 9 > Foster, Brian L. and Seidman, Stephen B. 1982. Urban structures derived from collections of overlapping subsets. Urban Anthropology 11: 177-92 < 2 2 3 > Frank, Philipp 1951. Relativity - a richer truth. London: Cape. 124 pp. < 2 i 8 >
Freud, Sigmund 1955. Totem and taboo, and other works. London: Hogarth Press, xv, 264 pp. Works 13 < I 5 4 > Gaboriau, Marc 1963. Anthropologie structurale et histoire, Esprit 3i(ii): 579-95 Gann, Lewis Henry 1958. The birth of a plural society: the development of Northern Rhodesia under the British South Africa company 1894-1914. Manchester:
246
BIBLIOGRAPHY
Manchester University Press, xxi, 230 pp. < i o 8 > Gellner, Ernest 1963. Nature and society in social anthropology. Philosophy of Science 30: 236-51 < 4 2 , I 7 6 > 1984. The scientific status of the social sciences. International Social Science Journal 36; 567-86 < i , 8, 11 > Giddens, Anthony 1976. New rules of sociological method: a positive critique of interpretative sociologies. London: Hutchinson. 192 pp. < 4 > Glasse, Robert Marshall and Meggitt, Mervyn John 1969 editors. Pigs, pearlshells and women: marriage in the New Guinea highlands. Englewood Cliffs, N. J.: Prentice-Hall, vii, 246 pp. < 6 4 > Gluckman, Max 1940. The kingdom of the Zulu of South Africa. In Fortes and Evans-Pritchard 1940b, pp. 25-55 < H 2 > 1951. The Lozi of Barotseland in north-western Rhodesia. In Colson, E. and Gluckman, M., eds. Seven tribes of British Central Africa. London: Oxford University Press, pp. 1-93 < 112 > 1963. Order and rebellion in tribal Africa. London: Cohen and West, xii, 273 pp. Godelier, Maurice 1966. Systeme, structure et contradiction dans 'Le capital'. Temps modernes 22: 828-64 < i 6 2 > Goethe, Johann Wolfgang 1949. Faust: a tragedy, translated by Taylor, Bayard. London: Euphorion Books. 361 pp. < 2 2 6 > Goodenough, Ward Hunt 1956. Residence rules. Southwestern Journal of Anthropology 12: 22-37 < 3 2 > 1970a. Epilogue: transactions in parenthood. In Carroll, Vern, ed. Adoption in eastern Oceania. Honolulu: University of Hawaii Press, pp. 391-410 < 3 8 > 1970b. Description and comparison in cultural anthropology. Chicago: Aldine, xi, 173 PP- Gouldner, Alvin Ward 1962. Anti-Minotaur: the myth of a value-free sociology. Social Problems 9: 199-213 < i 6 , 18 > 1974. Unpublished address to the British sociological association < 175 > I 975~6 Prologue to a theory of revolutionary intellectuals. Telos 26: 3-36 Grandy, Richard E. 1973. Reference, meaning, and belief. Journal of Philosophy 70: 439-52 < 2 4 > Greenberger, Martin 1966. The uses of computers in organizations. Scientific American 215(3): 193-202 < I 3 6 > Gross, Llewellyn Zwicker 1948. The use of class concepts in sociological research. American Journal of Sociology 54: 409-21 < 74 > Gudeman, Stephen 1972. The compadrazgo as a reflection of the natural and spiritual person. Proceedings of the Royal Anthropological Institute 19 71: 4 5 - 7 1 Guiart, Jean 1965. Survivre a Levi-Strauss. Arc 26: 61-4 < 2 i 2 > Habermas, Jiirgen 1972. Knowledge and human interests. London: Heinemann. viii, 356pp. < 3 3 > Hamre, Alfred 1964. Haugianskfrukt: Sunnhordland, Hardanger og Voss indremisjonssamskipnad gjennom 100 dr 1864-1964. Bergen: Sunnhordland, Hardanger
BIBLIOGRAPHY
247
og Voss indremisjonssamskipnad. 391 pp. < 9 7 > Hancock, William Keith 1937-42. Survey of British Commonwealth affairs. London: Oxford University Press. 2 vols. < 103 > Hanna, Alexander John 1956. The beginnings of Nyasaland and North-eastern Rhodesia. Oxford: Clarendon Press, viii, 281 pp. < I O 7 > Harary, Frank 1973. On the history of the theory of graphs. In Harary, F., ed. Ann Arbor conference on graph theory. New directions in the theory of graphs. New York: Academic Press, pp. 1-17 < 3 > Hertz, Robert i960. The pre-eminence of the right hand: a study in religious polarity. In Death and The right hand. London: Cohen and West. pp. 87-113, 155-60, 172-4 < 9 7 > Hesse, Mary 1976. Models versus paradigms in the natural sciences. In Collins, Lyndhurst, ed. The use of models in the social sciences. London: Tavistock, pp. 115 < 2 i 6 > Heusch, Luc de 1958. Les voies de l'anthropologie structural: l'oeuvre de M. LeviStrauss et revolution de l'ethnologie francaise. Zaire 12: 787-818 < I 5 8 , i6o> Hey, Anthony J. G. and Walters, Patrick 1987. The quantum universe. Cambridge: Cambridge University Press, x, 180 pp. < 5 > Hiatt, Lester Richard 19 71. Secret pseudo-procreation rites among the Australian Aborigines. In Hiatt, L. R. and Jayawardena, Chandra, eds. Anthropology in Oceania: essays presented to Ian Hogbin. Sydney: Angus and Robertson, pp. 7788 Hofstadter, Douglas R. 1979. Godel, Escher, Bach: an eternal golden braid. New York: Basic Books, xxi, 777 pp. < 2 2 5 > 1982. Metamagical themas: a self-referential column about last January's column about self-reference. Scientific American 246(1): 12-17 < 2 2 5 > Hofbin, Herbert Ian 1939. Experiments in civilization. London: Routledge. xvii, 268 pp. < H 5 > 1944. Native councils and native courts in the Solomon Islands. Oceania 14: 257-83 < H 5 > Holmberg, Alan Richard 1950. Nomads of the long bow. Washington, D. C: United States Government Printing Office, iv, 104 pp. Smithsonian Institution. Institute of Social Anthropology, Publication 10 Homans, George Caspar 1950. The human group. New York: Harcourt Brace, xxvi, 484 pp. < 7 8 > 1967. The nature of social science. New York: Harcourt, Brace and World. 109 pp. Horowitz, Irving Louis 1965. The life and death of Project Camelot. Trans-action (St Louis) 3: 3 - 7 , 4 4 - 4 7 < I 3 9 > 1967 editor. The rise and fall of Project Camelot: studies in the relationship between social science and practical politics. Cambridge, Mass.: MIT Press, xii, 385 pp. Hoff, Darrell 1954. H°w to lie with statistics. New York: Norton. 142 pp. < I 4 > Hunter, Monica 1936. Reaction to conquest. London: Oxford University Press, xx, 582 pp. < H 2 >
248
BIBLIOGRAPHY
Ikenberry, Ernest 1962. Quantum mechanics for mathematicians and physicists. New York: Oxford University Press, xii, 269 pp. < 137 > Jackson, Jean E. 1976. Vaupes marriage: a network system in the northwest Amazon. In Smith, Carol Ann, ed. Regional analysis: II. Social systems. Ann Arbor: University of Michigan Press, pp. 65-93 Jakobson, Roman 1949. Principes de phonologie historique. In Troubetzkoy, Nikolai Sergieevich, ed. Principes de phonologie. Paris: Klincksieck, pp. 315-36 Jencks, Christopher and others 1972. Inequality: a reassessment of the effect offamily and schooling in America. New York: Basic Books, xii, 399 pp. < 18 > Jones, Ernest, 1951. The Madonna's conception through the ear: a contribution to the relation between aesthetics and religion. In Essays in applied psychoanalysis. Vol. 2. London: Hogarth Press, pp. 266-357 < 3 7 > Kelly, Raymond Case 1968. Demographic pressure and descent group structure in the New Guinea highlands. Oceania 39: 36-63 1977. Etoro social structure: a study in structural contradiction. Ann Arbor: University of Michigan Press, xvi, 329 pp. < i 7 i > Keul, A. G. and Freeman, Linton C. 1987. Compute your network. Social Networks 9: 283-4 < 7 > Klovdahl, Alden S. 1981. A note on images of networks. Social Networks 3: 197214 Kramer, Gerald H. and Hertzberg, Joseph 1975. Formal theory. In Greenstein, Fred Irwin and Polsby, Nelson Woolf, eds. Strategies of inquiry. Reading, Mass.: Addison-Wesley. Handbook of Political Science 7, pp. 351-403 Kroeber, Alfred Louis 1935. History and science in anthropology. American Anthropologist 37: 539-69 < I 5 6 > Kuczynski, Robert Rene 1937. Colonial population. London: Oxford University Press, xiv, 101 pp. Kuhn, Thomas Samuel 1970. The structure of scientific revolutions. 2nd edn. Chicago: University of Chicago Press, xv, 172 pp. International Encyclopaedia of Unified Science 2(2) Kunstadter, Peter and others 1963. Demographic variability and preferential marriage patterns. American Journal of Physical Anthropology n. s. 21: 511-19 Kuper, Hilda 1947. The uniform of colour. Johannesburg: Witwatersrand University Press. 160 pp. < i i 2 > Kuznick, Peter J. 1987. Beyond the laboratory: scientists as political activists in 1930s America. Chicago: University of Chicago Press, x, 363 pp. < i 8 > La Flesche, Francis 1921. The Osage tribe: rites of the chiefs: sayings of the ancient men. In Bureau of ethnology. 36th annual report 1914-1915- Washington, D. C: Government Printing Office, pp. 35-597 < i 6 6 > 1925. The Osage tribe: the rite of vigil. In Bureau of Ethnology. 39th annual report 1917-1918. Washington, D. C: Government Printing Office, pp. 31-630 Langness, Louis LeRoy 1964. Some problems in the conceptualization of highlands social structures. In Watson, James B., ed. New Guinea: the central highlands.
BIBLIOGRAPHY
249
American Anthropologist 66(4:2): 162-82 < 6 i > Lave, Charles Arthur and March, James Gardner 1975. An introduction to models in the social sciences. New York: Harper and Row. x, 421 pp. < 2 i 8 > Lawrence, Peter 1975. The ethnographic revolution. Oceania 45:253-71 < 4 4 > Leach, Edmund Ronald 1961a. Rethinking anthropology and other essays. London: Athlone Press, ix, 143 pp. London School of Economics Monographs on Social Anthropology 22 < 3 4 , 167 > 1961b. Golden bough or gilded twig? Daedalus 90: 371-87 < 3 4 > 1964. Telstar et les Aborigenes ou 'La pensee sauvage'. Annales 19: 1100-16 1965a. Frazer and Malinowski: on the 'Founding fathers'. Encounter 25(5): 2 4 36 1965b. Claude Levi-Strauss - anthropologist and philosopher. New Left Review 34: 12-27 < I 5 5 > 1967a. Virgin birth. Proceedings of the Royal Anthropological Institute 1966: 3 9 49 < 3 4 > 1967b. Introduction. In Leach 1967c, pp. vii-xix < I 5 5 > 1967c editor. The structural study of myth and totemism. London: Tavistock. xxi, 185 pp. ASA Monograph 8 Lefebvre, Henri 1966. Claude Levi-Strauss ou le nouvel eleatisme. Homme et la Societe 1:21-31 < i 6 6 > Lefort, Claude 1951, L'echange et la lutte des hommes. Temps Modernes 6: 140017 < I 5 3 > 1952. Societe 'sans histoire' et historicite. Cahiers Internationaux de Sociologie 12: 91-114 < i 5 i > Leighton, Alexander Hamilton 1959. My name is legion: foundations for a theory of man in relation to culture. New York: Basic Books, xii, 452 pp. Stirling County Study of Psychiatric Disorder and Sociocultural Environment 1 < I 5 9 > Leinhardt, Samuel 1977 editor. Social networks: a developing paradigm. New York: Academic Press, xxxiv, 465 pp. < 2 i 7 > Levins, Richard 1968. Evolution in changing environments: some theoretical explorations. Princeton, N. J.: Princeton University Press, vii, 120 pp. Monographs in Population Biology 2 < 2 i 7 , 2 2 0 Levi-Strauss, Claude 1949. Les structures elementaires de la parente. Paris: Presses Universitaires de France, xiv, 639 pp. < 158, I 7 9 > 1950. Introduction a l'oeuvre de Marcel Mauss. In Mauss, M. Sociologie et anthropologie. Paris: Presses Universitaires de France, pp. ix-lii < I 5 8 , 159, 163 > 1953. Social structure. In Kroeber, A. L., ed. International Symposium on Anthropology. Anthropology today. Chicago: University of Chicago Press, pp. 524-53 < I 5 8 > 1956. Race and history. In United Nations Educational Scientific and Cultural Organization. The race question in modern science. Paris: Unesco, pp. 123-63 i960. Anthropologie sociale (Resume des cours et travaux de Tannee scolaire 1959-1960). Annuaire de College de France 60: 191-207 < i 6 2 >
25O
BIBLIOGRAPHY
1962a La pensee sauvage. Paris: Plon. ii, 395 pp. < i 5 i , i64> 1962b. Les limites de la notion de structure en ethnologic In Bastide, Roger, ed. Sens et usage du terme structure dans les sciences humaines et sociales. The
Hague: Mouton, Janua Linguarum, Series Minor 16, pp. 40-5 < 151, 158, 159, 162> 1963a. Totemism. Boston: Beacon Press. 116 pp. < I 5 7 > 1963b. Structural anthropology. New York: Basic Books, xvi, 410 pp. < 19, 15167 passim, 192, 224 > 1964a. Mythologiques: le cru et le cuit. Paris: Plon. 402 pp. < 160, 166 > 1964b. Criteria of science in the social and human disciplines. International Social Science Journal 16: 534-52 < i 5 i , I55> 1965. Man has never been so savage as he is today. Realities 175: 48-51 1966a. The future of kinship studies. Proceedings of the Royal Anthropological
Institute 1965: 13-22 1966b. The savage mind. London: Weidenfeld and Nicolson. xii, 290 pp. 1966c. The scope of anthropology. Current Anthropology 7: 112-23 < I 6 I > I966d. Mythologiques: du miel aux cendres. Paris: Plon. 452 pp. 1966c Overture to Le cru et le cuit. Yale French Studies 36/7: 41-65 < 152-67 passim >
I966f. Anthropology: its achievements and future. Current Anthropology 7: 1247
1967a. The story of Asdiwal. In Leach 1967c, pp. 1-47 < i 6 o > 1967b. L'anthropologie sociale. Sciences 47: 115-28 < I 5 3 > Levi-Strauss, C, and others 1963. Reponses a quelques questions. Esprit 3i(ii): 628-53 < J 5 2 , 153, 158, i6o> Levy, Marion Joseph, jr 1965. Aspects of the analysis of family structure. In Coale, Ansley J. and others. Aspects of the analysis of family structure. Princeton:
Princeton University Press, pp. 1-63 Lewin, Kurt 1933. A dynamic theory of personality: selected papers. New York:
McGraw-Hill, ix, 286 pp. Liebermann, Felix 1903. Die Gesetze der Angelsachsen. Erster Band. Halle: Niemeyer. lxii, 675 pp. 1916. Die Gesetze der Angelsachsen. Dritte Band. Halle: Niemeyer. 3 56 pp. < 37 > Lienhardt, Godfrey 1974. E-P: a personal view: Sir Edward Evans-Pritchard, 19021973. Man n. s. 9: 299-304 Llobera, Josep R. 1979. Techno-economic determinism and the work of Marx on pre-capitalist societies. Man n. s. 14: 249-70 < I 7 5 > Low, David Anthony 1973. Lion rampant: essays in the study of British imperialism.
London: Frank Cass. x, 230 pp. < i i 2 > Lukhero, Matshakaza Blackson 1985. Ngoni nc'wala ceremony. Lusaka: National Education Company of Zambia, xvii, 81 pp. < i 2 i > Luther, Martin 1883-. D. Martin Luther Werke. Kritische Gesammtausgabe. Weimar:
H. Bohlau. 69 vols. = WA
BIBLIOGRAPHY
25I
I
955~- Luther's works. American edn. St Louis, Mo.: Concordia Publishing House. 55 vols. = LW McArthur, Margaret 1967. Analysis of the geneaology of a Mae-Enga clan. Oceania 37: 281-5 < 6 4 > Macdonald, Graham and Pettit, Philip 1981. Semantics and social science. London: Routledge & Kegan Paul, vi, 194 pp. < 2 4 > Mach, Ernst 1969. The science of mechanics: a critical and historical account of its development. 6th edn. Lasalle, 111.: Open Court, xxxii, 634 pp. < 2 2 4 > Machlup, Fritz 1956. The inferiority complex of the social sciences. In Sennholz, Mary Homan, ed. On freedom and free enterprise. Princeton, N. J.: Van Nostrand, pp. 161-72 < I 9 8 > MacRae, Donald Gunn 1970. Let's go beyond 'problems'. New Society 15: 359-60 Madan, Triloki Nath 1977. The quest for Hinduism. International Social Science Journal 29: 261-78 < 2 i o > 1982. Anthropology as the mutual interpretation of cultures: Indian perspectives. In Fahim, Hussein, M., ed. Indigenous anthropology in non-Western countries. Durham, N. C: Carolina Academic Press, pp. 4-18
Mair, Lucy Philip 1934. An African people in the twentieth century. London: Routledge. xvi, 300 pp. < H 2 > Maitland, Frederic William 1895. Introduction. In Whittaker 1895, pp. ix-lv 1911. The collected papers of Frederic William Maitland, Downing professor of the laws of England. Vol. 3. Cambridge: Cambridge University Press. 566 pp. Malinowski, Bronislaw 1930. Kinship. Man 30: 19-29 < i 6 9 , I7O> 1963. The family among the Australian Aborigines: a sociological study. New York: Schocken Books, xxx, 322 pp. < 3 4 > Mannheim, Karl 1936. Ideology and Utopia: an introduction to the sociology of knowledge. London: Kegan Paul, xxxi, 318 pp. < 2 4 > Matarasso, Michel 1963. Lecture de Claude Levi-Strauss: pensee sauvage et logique premiere. Revue Francaise de Sociologie 4: 195-205 < I 5 8 > Medawar, Peter Brian 1969. Induction and intuition in scientific thought. London: Methuen. ix, 62 pp. < 2 > Meggitt, Mervyn John 1957. Enga political organization. A preliminary description. Mankind 5: 133-7 < 5 7 > 1962a. Desert people: a study of the Walbiri Aborigines of central Australia. Sydney: Angus and Robertson, xx, 348 pp. 1962b. Growth and decline of agnatic descent groups among the Mae Enga of the New Guinea highlands. Ethnology 1: 158-65 1962c. Dream interpretation among the Mae Enga of New Guinea. Southwestern Journal of Anthropology 18:216-2 9 < 5 7 > 1964. Male-female relationships in the highlands of Australian New Guinea. In Watson, James B., ed. New Guinea: the central highlands. American Anthropologist 66(4:2): 204-24 < 5 7 >
252
BIBLIOGRAPHY
1965a. The lineage system of the Mae-Enga of New Guinea. Edinburgh: Oliver &
Boyd. xix, 297 pp. < 5 6 , 57, 159> 1965b. The Mae Enga of the Western Highlands. In Lawrence, Peter and Meggitt, M. J., eds. Gods, ghosts and men in Melanesia. Melbourne: Oxford University Press, pp. 105-31 < 5 7 > 1967a. Uses of literacy in New Guinea and Melanesia. Bijdragen tot de taal-, landen volkenkunde 123: 71-82 < 5 7 > 1967b. The pattern of leadership among the Mae-Enga of New Guinea. Anthropological Forum 2: 20-35 1969. Introduction. In Glasse and Meggitt 1969, pp. 1-15 < 5 7 > 1972. Understanding Australian Aboriginal society: kinship systems or cultural categories? In Reining, Priscilla, ed. Kinship studies in the Morgan centennial
year. Washington, D. C: Anthropological Society of Washington, pp. 64-87 Mehan, Hugh, and Wood, Houston 1975. The reality of ethnomethodology. New York: Wiley, xiii, 259 pp. < 4 > Mendelson, Edward Michael 1967. The 'uninvited guest': ancilla to Levi-Strauss on totemism and primitive thought. In Leach 1967c, pp. 119-39 < i 6 3 > Merleau-Ponty, Maurice 1964. Signs. Evanston: Northwestern University Press. xxxiv, 355 pp. < I 5 7 > Merton, Robert King 1981. Our sociological vernacular. Columbia: the Magazine of Columbia University November: 42-4 < 6 > Meyer, Arthur William 1939. The rise of embryology. Stanford: Stanford University Press, xv, 367 pp. < 3 4 , 36, 39 > Miller, Donald Bruce 1976. A genius-like feeling for the social stuff. South Asia 6: II2-2O
Mills, James Philip 1953. Anthropology as a hobby. Journal of the Royal Anthropological Institute 85: 1-8 < i 6 > Mitchell, James Clyde 1949. The political organization of the Yao of southern Nyasaland. African Studies 8: 141-59 < i i 5 > 1 9 5 6 . The Yao village: a study in the social structure of a Nyasaland tribe.
Manchester: Manchester University Press, xviii, 235 pp. < i i 2 > Molland, Einar 19 51. Fra Hans Nielsen Hauge til Eivind Berggrav: hovedlinjer i Norges kirkehistorie i det 19. og 20. arhundre. Oslo: Gyldendal Norsk Forlag. 111 pp.
Moreno, Jacob L. 1 9 3 4 . Who shall survive? A new approach to the problem of human
interrelations. Washington, D. C: Nervous and Mental Disease Publishing Company, xvi, 440 pp. < 72 > Morgan, David Hopcraft John 1975. Social theory and the family. London: Routledge & Kegan Paul, viii, 280 pp. < I 7 6 > Morgan, Lewis Henry 1 8 7 0 . Systems of consanguinity and affinity of the human
family. Washington, D. C: Smithsonian Institution, xii, 590 pp. Smithsonian Contributions to Knowledge 17(218) < i 6 9 > Morton, Earl of 1821. A communication of a singular fact in natural history. Philosophical Transactions of the Royal Society of London 1 8 2 1 : 2 0 - 2 2 < 3 6 > Mueller, William A. 1 9 5 4 . Church and state in Luther and Calvin: a comparative study.
BIBLIOGRAPHY
253
Nashville, Term.: Broadman Press, ix, 183 pp. < 9 6 , 97 > Mukherjee, Ramkrishna 1977. Trends in Indian sociology. Current Sociology 25(3): I-I93 < 199, 2OI, 2IO>
Mullan, Bob 1987. Sociologists on sociology. London: Croom Helm. 322 pp. < 10, II, 12, 13, 2I>
Mullen, R. J. and Whitten, Wesley Kingsley 19 71. Relationship of genotype and degree of chimerism in coat colour to sex ratios and gametogenesis in chimeric mice. Journal of Experimental Zoology 178: 165-76 < 3 6 > Muller, Robert 1986. Decide to network. In Lipnack, Jessica and Stamps, Jeffrey. The networking book: people connecting with people. New York: Routledge & Kegan Paul, p. i < 6 > Myrdal, Gunnar 1944. An American dilemma: the Negro problem and modern democracy. New York: Harper, lix. 1483 pp. < i 8 > Naroll, Raoul S. 1964. On ethnic unit classification. Current Anthropology 5: 283312 < 3 i > National Herald 1955. Letter to editor. National Herald (Lucknow) 21 March [cited in Saran 1958: 1031] Needham, Joseph 1959. A history of embryology. 2nd edn. London: Cambridge University Press. 304 pp. Needham, Rodney 19 71. Remarks on the analysis of kinship and marriage. In Needham, R., ed. Rethinking kinship and marriage. London: Tavistock. ASA Monographs 11, pp. 1-34 < i 7 i > Oettinger, Anthony Gervin 1966. The uses of computers in science. Scientific American 215(3): 161-72 < i 6 7 > Okada, Yuzuru 1955. Race relations in Formosa under the Japanese. In Lind, Andrew William, ed. Race relations in world perspective. Honolulu: Hawaii University Press, pp. 371-87 < n o > Paine, Robert 1985 editor. Advocacy and anthropology, first encounters. St John's: Memorial University of Newfoundland, Institute of Social and Economic Research, xviii, 278 pp. < i 8 > Parkes, Alan Sterling i960. The biology of spermatozoa and artificial insemination. In Parkes, A. S., ed. Marshall's physiology of reproduction. 3rd edn. Vol. 1, part 2. London: Longmans, pp. 161-263 < 3 6 > Pearson, Karl 1892. The grammar of science. London: Walter Scott, xvi, 493 pp. Perham, Margery 1956. Lugard: the years of adventure, 1858-1898. London: Collins, xv, 750 pp. < I O 7 > Peters, Emrys Lloyd i960. The proliferation of segments in the lineage of the Bedouin of Cyrenaica. Journal of the Royal Anthropological Institute 90: 29-53 1963. Aspects of rank and status among Muslims in a Lebanese village. In PittRivers, Julian Alfred Lane-Fox, ed. Mediterranean countrymen: essays in the social anthropology of the Mediterranean. Paris: Mouton, pp. 159-200 Piddington, Ralph O'Reilly 1970. Irregular marriages in Australia. Oceania 40: 329-43 < 2 i 9 >
254
BIBLIOGRAPHY
Pike, Kenneth Lee 1967. Language in relation to a unified theory of the structure of human behavior. 2nd edn. The Hague: Mouton. 762 pp. Janua Linguarum, Series
Major 24 Prabhu, Pandharinath H. 1963. Hindu social organization: a study in socialpsychological and ideological foundations. 4th edn. Bombay: Popular Prakashan.
xix, 389 pp. < 2 O i > Price, Frances V. 1981. Only connect? Issues in charting social networks. Sociological Review n. s. 29: 283-312 < 2 i 9 > Putnam, Hilary 1978. Meaning and the moral sciences. London: Routledge & Kegan Paul, ix, 145 pp. < 2 i 7 , 222> Radcliffe-Brown, Alfred Reginald 1940. On social structure. Journal of the Royal Anthropological Institute 70: 1-12 < I 7 2 > 1952. Structure and function in primitive society: essays and addresses. London:
Cohen and West, vii, 219 pp. < I 5 7 , 165 > 1957. A natural science of society. Glencoe, 111.: Free Press, xii, 156 pp. < 166 > 1958. Method in social anthropology: selected essays. Chicago: University of Chicago Press, xxi, 189 pp. Radcliffe-Brown, A. R. and Forde, Cyril Daryll 1950 editors. African systems of kinship and marriage. London: Oxford University Press, ix, 399 pp. Rapoport, Anatol 1959. Uses and limitations of mathematical models in social science. In Gross, Llewellyn Zwicker, ed. Symposium on sociological theory. Evanston, 111.: Row, Peterson, pp. 348-72 Read, Margaret 1936. Tradition and prestige among the Ngoni. Africa 9: 453-84 Redfield, Robert 1953. Relation of anthropology to the social sciences and the humanities. In Kroeber, A. L., ed. International Symposium on Anthropology. Anthropology today: an encyclopedic inventory. Chicago: University of Chicago
Press, pp. 728-38 < n > 1956. Peasant society and culture: an anthropological approach to civilization.
Chicago: University of Chicago Press, vii, 162 pp. < 6 8 > Rex, John 1975. The present state of sociology. Seminar paper, University of Alberta at Edmonton, June, [cited Mullan 1987: 32] < 2 i > Ricoeur, Paul 1963. Structure et hermeneutique. Esprit 3i(ii): 596-627 < I 5 3 , 157, 162 > Riviere, Peter 1972. Review: Strathern, One father, one blood. Journal of the Anthropological Society of Oxford 3: 158 < 4 4 > Rorty, Richard 1980. Philosophy and the mirror of nature. Princeton: Princeton University Press, xv, 402 pp. < 2 > 1981. Method, social science, and social hope. Canadian Journal of Philosophy 11: 569-88 < 6 > Rose, Deborah Bird, Michaels, Eric and Myers, Fred R. 1987. Issues of representation, responsibility and reflexive anthropology: the Pintupi case. Canberra Anthropology 10(1): 35-73 < 5 > Runciman, Walter Garrison 1965. Sociologese. Encounter 25(6): 45-7 < 7 > Sahlins, Marshall David 1976. The use and abuse of biology: an anthropological critique
BIBLIOGRAPHY
255
of sociobiology. Ann Arbor: University of Michigan Press, xv, 120 pp. Saran, Awadh Kishore 1958. India. In Roucek, Joseph Slabey, ed. Contemporary sociology. New York: Philosophical Library, pp. 1013-34 1962. Review: Contributions to Indian Sociology 4. Eastern Anthropologist 15: 5 3 68 Sarkar, Susobhan Chandra 1970. Bengal renaissance and other essays. New Delhi; People's Publishing House, xi, 285 pp. < 2 o o > Sartre, Jean-Paul i960. Critique de la raison dialectique. Vol. 1. Paris: Gallimard. 759 pp. < I 5 2 > Saussure, Ferdinand de 1959. Course in general linguistics. New York: Philosophical Library, xvi, 240 pp. < I 5 7 > Schapera, Isaac 1956. Government and politics in tribal societies. London: Watts, vii, 238 pp. < H 2 > 1962. Should anthropologists be historians? Journal of the Royal Anthropological Institute 92: 143-56 < i 6 2 > Scheffler, Harold Walter 1985. Filiation and affiliation. Man n. s. 20: 1-21 < 4 5 > Schneider, David Murray 1961. The distinctive features of matrilineal descent groups. In Schneider, D. M. and Gough, Eleanor Kathleen, eds. Matrilineal kinship. Berkeley: University of California Press, pp. 1-29 < 4 i > 1965. Kinship and biology. In Coale, Ansley J. and others. Aspects of the analysis of family structure. Princeton: Princeton University Press, pp. 83-101 < 4 2 > 1968. American kinship: a cultural account. Englewood Cliffs, N. J.: Prentice-Hall. x, 117 pp. 1972. What is kinship all about? In Reining, Priscilla, ed. Kinship studies in the Morgan centennial year. Washington, D. C: Anthropological Society of Washington, pp. 32-63 1984. A critique of the study of kinship. Ann Arbor: University of Michigan Press. xii, 208 pp. < 3 O > Schroeder, Edison B. 1966. Letter to the editor. Scientific American 215(3): 12 Scott, Peter 1987a. Higher culture and the humanities. Times Higher Education Supplement 770: 28 [7 August] The knowledge business I < I 2 > 1987b. We are all social scientists now. Times Higher Education Supplement 771: 24 [14 August] The Knowledge business II < 15 > Seeley, John Ronald 1964. Crestwood Heights: intellectual and libidinal dimensions of research. In Vidich et al. 1964, pp. 157-206 < I 4 4 > Seidman, Stephen B. and Foster, Brian L. n. d. An anthropological framework for the analysis of social networks. Mimeographed, 23 pp. < 2 2 3 > Seligmann, Charles Gabriel 1902. The medicine, surgery, and midwifery of the Sinaugolo. Journal of the Anthropological Institute 32: 297-304 Shapiro, Warren 19 71, Patri-groups, patri-categories, and sections in Australian Aboriginal social classification. Man n. s. 6: 590-600 < 4 O > 1981. Miwuyt marriage: the cultural anthropology of affinity in northeast Arnhem Land. Philadelphia: Institute for the Study of Human Issues, xii, 173 pp.
256
BIBLIOGRAPHY
Sharman, G. B. 1955. Studies on marsupial reproduction. III. Normal and delayed pregnancy in Setonix brachyurus. Australian Journal of Zoology 3: 56-70 Sharman, G. B. and Berger, P. J. 1969. Embryonic diapause in marsupials. Advances in Reproductive Physiology 4: 211-40 Shils, Edward Albert 1961. The intellectual between tradition and modernity: the Indian situation. Comparative Studies in Society and History Supplement 1. 120 pp. 1967. The ways of sociology. Encounter 26(6): 85-91 < 190 > Sidgwick, Henry 1904. Miscellaneous essays and addresses. London: Macmillan. viii, 374PP- < i 8 3 , 192> Singh, Yogendra 1970. For a sociology of India. Contributions to Indian Sociology n. s. 4: 140-4 < 2 o 8 > Smelser, Neil Joseph and Davis, James Allan 1969 editors. Sociology: the behavioral and social science survey, sociology panel. Englewood Cliffs, N. J.: Prentice-Hall. ix, 178 pp. < I 9 3 > Smith, Michael Garfield 1956. On segmentary lineage systems. Journal of the Royal Anthropological Institute 86(2): 39-80 < n 8 > 1962. History and social anthropology. Journal of the Royal Anthropological Institute 92: 73-85 < I 5 7 , 162> Sontag, Susan 1966. Against interpretation, and other essays. New York: Farrar, Straus and Giroux. x, 304 pp. < I 5 4 > Spencer, Robert Francis 1949-50. Primitive obstetrics. Ciba Symposia 11: 115788 < 3 4 > Spiro, Melford Elliot 1968. Virgin birth, parthenogenesis and physiological paternity: an essay in cultural interpretation. Man n. s. 3: 242-61 < 3 4 , 38 > Srinivas, Mysore Narasimhachar 1966. Social change in modern India. Berkeley: University of California Press, xvii, 194 pp. < I 3 8 > Srinivas, M. N. and Panini, M. N. 1973. The development of sociology and social anthropology in India. Sociological Bulletin 22: 179-215 < 2 i o > Stanner, William Edward Hanley 1956. The dreaming. In Hungerford, Thomas Arthur Guy, ed. Australian signpost. Melbourne: Cheshire, pp. 51-65 < 164 > 1967. Radcliffe-Brown, Alfred Reginald (1881-1955) [draft of article for International encyclopedia of the social sciences] (quoted with permission) Steiner, George 1966. A conversation with Claude Levi-Strauss. Encounter 26(4): 32-8 < i 6 4 > Stokes, Eric Thomas 1959. The English utilitarians and India. Oxford: Clarendon Press, xvi, 350 pp. < 2 o o > Stouffer, Samuel Andrew and others 1949. The American soldier. Princeton: Princeton University Press. 2 vols. Studies in social psychology in World War II. vols. 1-2 < 7 > Straume, Jakob i960. Kristenliv i Hardanger, Sunnhordland og Karmsund: eit sogeskrift. Bergen: A. S. Lunde (Haugesund-Sunnhordland og Hardanger krins av Norsk luthersk misjonssamband). 265 pp. < 9 7 > Struve, Otto 1950. Stellar evolution: an exploration from the laboratory. Princeton,
BIBLIOGRAPHY
257
N. J.: Princeton University Press, xiv, 266 pp. < 146 > Sturtevant, William Curtis 1964. Studies in ethnoscience. In Romney, Antone Kimball and D'Andrade, Roy Godwin, eds. Transcultural studies in cognition. American Anthropologist 66(3: 2): 99-131 < i 8 8 > Sutter, Jean and Tabah, Leon 1951. Les notions de l'isolat et de population minimum. Population 6: 481-98 < i 6 2 > Swete, Henry Barclay 1909. The Holy Spirit in the New Testament: a study in primitive Christian teaching. London: Macmillan. x, 417 pp. < 3 7 > Tarkowski, Andrzej K. 1961. Mouse chimeeras developed from fused eggs. Nature 190: 857-60 < 3 6 > Tax, Sol 1955. From Lafitau to Radcliffe-Brown: a short history of the study of social organization. In Eggan, Frederick Russell, ed. Social anthropology of north American tribes. Enlarged edn. Chicago: University of Chicago Press, pp. 4 4 5 81 Textor, Robert Bayard 1967. A cross-cultural summary. New Haven: HRAF Press. xix, 208 pp. and printout < 2 5 > THES 1988. Register puts seal of approval on psychologists. Times Higher Education Supplement 793: 4 [15 January] < I 5 > Thompson, William David James Cargill i960. The two regiments: a study of the development of the theory of the relations of church and state during the Reformation, with particular reference to England. University of Cambridge, Ph. D. dissertation. < 9 7 > Thruston, Captain 1897. Thruston to Ternan, 30 June. Entebbe secretariat archives, A4/8 [cited Low 1973: 14-15] < H 2 > Times 1967. Late vocational surprises [and correspondence arising]. Times (London) 11, 16, 18&25 August < I 9 3 > TLS 19 71. No nearer to a science [Review: Barnes, Three styles in the study of kinship]. Times Literary Supplement 1971: 1395 < I 5 O > Touraine, Alain 1965. Sociologie de Faction. Paris: Editions du Seuil. 507 pp. Troughton, Ellis Le Geyt 1965. Furred animals of Australia. 8th edn. Sydney: Angus and Robertson, xxxii, 376 pp. Turney-High, Harry Holbert 1949. Primitive war: its practice and concepts. Columbia: University of South Carolina Press, xi, 277 pp. < 106 > Uberoi, Jitendra Pal Singh 1968. Science and swaraj. Contributions to Indian Sociology n. s. 2: 119-23 1974. New outlines of structural sociology, 1945-19 70. Contributions to Indian Sociology n. s. 8: 135-52 1978. Science and culture. Delhi: Oxford University Press. 88 pp. 1984. The other mind of Europe: Goethe as a scientist. Delhi: Oxford University Press. 94 pp. Vidich, Arthur Joseph and Bensman, Joseph 1964. The Springdale case: academic bureaucrats and sensitive townspeople. In Vidich et al. 1964; pp. 313-49 Vidich, A. J. and others 1964 editors Reflections on community studies. New York: Wiley and Sons, xvi, 359 pp.
258
BIBLIOGRAPHY
von Foerster, Heinz 1952 editor. Cybernetics: circular causal and feedback mechanisms in biological and social systems (Transactions of the ninth conference). New York: Josiah Macy Jr Foundation, xx, 184 pp. < 136 > WA see Luther, Martin Walsh, John 1965. Social sciences: cancellation of Camelot after row in Chile brings research under scrutiny. Science 149: 1211-13 < I 3 9 > Wang, Kang 1981. Zhongguo shehuixue de guoqu he xianzai [The past and present of Chinese sociology]. Dousou 43: 21-8 < I 9 > Ward, Lester Frank 1883. Dynamic sociology, or applied social science, as based upon statistical sociology and the less complex sciences. Vol. 1. New York: Appleton. xx, 706 pp. < 6 > Warner, Malcolm and Stone, Michael 1970. The data bank society: organizations, computers and social freedom. London: Allen and Unwin. 244 pp. < 2 i i > Warner, William Lloyd 1958. A black civilization: a social study of an Australian tribe. Rev. edn. New York: Harper, xx, 618 pp. < 4 O > Weber, Max 1946. Science as a vocation. In Gerth, Hans Heinrich and Mills, Charles Wright, eds. From Max Weber: essays in sociology. New York: Oxford University Press, pp. 129-56 < 3 3 > Wegmann, Thomas G. 1970. Enzyme patterns in tetraparental mouse liver. Nature 225: 462-3 < 3 6 > Weismann, August 1893. The germ-plasm: a theory of heredity. London: Walter Scott, xxiii, 477 pp. < 3 6 > Wellman, Barry 1983. Network analysis: some basic principles. In Collins, Randall, ed. Sociological theory 1983. San Francisco: Jossey-Bass, pp. 155-200 1988. Structural analysis: from method and metaphor to theory and substance. In Wellman, B, and Berkowitz, Stephen D., eds. Social structures: a network approach. Cambridge: Cambridge University Press, pp. 19-61 < 2 i 6 > Wells, Herbert George 1895. The time machine: an invention. Heinemann. 152 pp. Whittaker, William Joseph 1895 editor. The mirror of justices. London: Quaritch. lv, 210 pp. Publications of the Selden Society 7 < 3 7 > Wiese, Carl 1900. Beitrage zur Geschichte der Zulu im Norden des Zambesi, namentlich der Angoni. Zeitschrift fur Ethnologie 32: 181-201 < I 2 3 > Wilier, David 1967. Scientific sociology: theory and method. Englewood Cliffs, N. J.: Prentice-Hall, xxi, 131 pp. < 2 i 6 > Wilson, Edward Osborne 1975. Sociobiology: the new synthesis. Cambridge, Mass.: Belknap Press of Harvard University Press, ix, 698 pp. < \j6> Winterbottom, John Miall 1950. Outline histories of two Northern Rhodesian tribes. Rhodes-Livingstone Journal 9: 14-25 < I 2 8 > Wolfe, Roy L. 1974. The methodology of scientific research programmes versus the sociology of knowledge. Cambridge Anthropology 1(2): 32-47 Wolin, Sheldon Sanford 1956. Politics and religion: Luther's simplistic imperative. American Political Science Review 50: 24-42 < 9 6 > Worsley, Peter Maurice 1956. The kinship system of the Tallensi: a revaluation. Journal of the Royal Anthropological Institute 86(1): 37-75 < 2 9 >
BIBLIOGRAPHY
259
1974. The state of theory and the status of theory. Sociology 8: 1-17 < I 4 > Yanagisako, Sylvia Junko 1978. Introduction. American Ethnologist 5: 1-4 Zelditch, Morris, Jr 1959. Statistical marriage preferences of the Ramah Navaho. American Anthropologist 61: 470-91 < i 6 2 > Zirkle, Conway 1935. The beginnings of plant hybridization. Philadelphia: University of Pennsylvania Press, xiii, 231 pp. 1946. The early history of the idea of the inheritance of acquired characters and of pangenesis. Transactions of the American Philosophical Society n. s. 35: 9 1 -
151
INDEX
Citations are not indexed here. Their locations are shown in the bibliography, at the end of each item. 203; Sociological Association 16; Sociological Society 16; universities 182 Aborigines, Australian 105f: and Dreaming amnesia, structural 228 164; and esoteric knowledge 119; and analogue models 216 gestation 35, 38, 40; and land claims analytical procedures 148 17; and marriage systems 158, 219; and Anaxagoras 39 saturation effect 211; seen as cultureless ancestor: cult 47; worship 122f 110; and subsections 190; and Anglicization 200 Transvaal lOOff Anglo-Saxons 37 accretion 61 anthropologists, social: and embarrassment acquaintance, ties of 72 191; and marginality 19; and administration 118 professional associations 16f; and administrative paternalism, 107, 112 salience of data-collection 13; and adultery 35 sociological understanding 188f; and advocacy research 18 specialized vocabulary 187 Aeschylus 224 anthropology, social see social Aethelberht, King 3 7 anthropology affiliation, bounded and unbounded 52 anti-Christ 91 Africa, West 107, 109 anticlericalism 92, 98 African: mirage 46; phase 60; society 228 Anzac Day ceremony 147 African political systems 46 apartheid 209 Africanists 44, 67 apes 34 agency, human 219 Apollo 38 agnatic; descent 48; dogma 47; ideology applied: humanities 198; mathematics and 67; idiom 64; lineages 44, 67; theoretical physics 221; social research taxonomies 56 220 agnation 4 1 , 48, 64 April rainers 122 alcohol 93, 100 Aranda, northern (Arunda) 156, 162, 172 algebra 224: and kinship 170f Arcadian social order 200 Althusser, Louis 13 archaeology 176 ambiguity, area of 62 archetype 225 America, North 57, 107: and social science archives 90 textbooks 208 arcs 215, 225 American: anthropologists 19; Indians Aristotle: on gestation 35, 38f; still read 141, 160; research projects in India 224
260
INDEX armoury, intellectual 218f arts 137, 197f Asdiwal 160 Ashanti 30, 106f Asiatic Society of Bengal 210 assimilation 147 Association of Social Anthropologists 16 Athenians 160 atom 5 attitudes 189 Auden, Wystan Hugh 165 Austen, Jane 185 Australia 16, 20, 107 Australian: Aboriginals see Aborigines, Australian; Anthropological Society 17; Association of Social Anthropologists 17; colonial rule 116; Institute of Aboriginal Studies 20 authority, concept of 5 Azande 34, 40 Baer, Karl Ernst von 3 5 Bailey, Frederick George 206 balance model 217 ballot, secret 80 bandits 105 Bantu 108, 111 Barnes, Frances vii Barotse, 106, 114 Bedouin, Cyrenaican 46, 56, 62 Beethoven 137 beliefs about the past 121 Bemba 122 Bengal and Bihar, Governor of 210 Berlin West African Conference 107 Bible 94, 155 big man 50 binary opposition 160 biographical history 163 biological factors 158 biology: and anthropologists 176; and blood groups 135; and kinship 31, 34, 177; supplying concepts 192 birth 29, 34, 228 black: Arcadia 123; hole 5, 225 blood 34, 37, 64: menstrual 38 Boer 111 boundaries, network 179 Bremnes vii, 68ff, 75: and Pietism 91ff, 99ff bricolage 154 bride wealth: as symbol 125; in Africa 52; in New Guinea 59; among Ngoni 124 British: Academy 11; North America 108; South Africa compan 128; village study 141 Brodbeck, Mary 217
26l
Buddhism 89 Bullpitt, Beverley vii bureaucratic status 172 Bushmen 106, 109ff calculus and diachronic models 162 calibration of model 58 California as son of Uncle Sam 4 1 , 6 3 Cambridge, university of 8, 180, 221 Camelot, Project 139, 187, 210 cancer 135 Cape of Good Hope 106f Cape Town 67 capitalism, Victorian 182 cash crops 59, 113 caste 200, 209, 211 catastrophic segmentation 53 categories, social 74 Catholics 91, 101 causual: connexions 23f; interactions 12 ceremonial exchange 51, 54, 116 Cewa 122f, 125 change, trends of 59 Chaplin, Charlie 166 charisma 6 charismatic teachers 135 charters 4 1 , 48 Charybdis 156 chauvinism, male 40 chemistry 5, 198 chiefs 113f children of God 89 Chile 139 Chimbu (Simbu) 46, 55: and agnatic model 48; and Enga 58, 65 chimeric mice 36 China 19 Chinese 110: meals 156 Chipata 120f Christian: education 93f, 99; People's Party 88f, 91ff, 101 Christianity 59, 89 chromosomes 35 chronic segmentation 53 chronicles 121 church in Lutheran theology 97 cingoni 124 citizens 185, 190: as mundane actors 13; and power 212; revolt of 210 citizenship 17 clan: agnates 63; territories 65 clans 172 class, social: 5f, 74ff, 188; network 11\ society 87; structure 4 classifications, elaborate 219 cleavage, lines of 84 clergy and laity 97
262
INDEX
clerks 113 cliques 74 clockwork 217 closure 16, 21 clusters 73 cold societies 163 collective action 68 colonial: conditions 186; context 172; elite 86; model 188; origins 199; rule 86, 120; situation 20, 32, 188; system 196 colonialism vi, 103, 183f Colson, Elizabeth 114 Columbus, Christopher 140 commandments, ten 89 committees 68, 79: membership of 80 common sense 7ff, 135 communist-led countries 3, 19, 203 community studies 139, 141, 202 comparative science of cultures 3 3 comparativistic science 206 competence: of lay citizens 9; professional 205 complementary filiation 50 componential analysis 178 computer graphics 225 computers: and citizens' revolt 210; and historians 163; and machine translation 167; and poetry 155; and real time 136 Comte, August 191 conception 34 conceptual structures 8 confirmation 94 Congress of Anthropological and Ethnological Sciences 171 Congo basin 109 Connections (journal) 216 conscious-unconscious contrast 159 consensus 78 consequences, unintended 192 conspicuous consumption 76 consultants 17 contemporary records 147 continua 161 continuity 155, 163 Contributions to Indian Sociology (journal) 206 convergence, disciplinary 174 cooperatives 80 copulation 29, 228: necessary, not sufficient 40 co-residence 49 corporate groups 67 correlations 22, 25 cosmological beliefs 89ff cosmology 163 cosmopolitanism 207 counter-prestations 160
Cratylus 165 Crestwood Heights 144 crime 188 criminals 12 criminology 134 critical: function of sociology 192; obedience 96; stance 194 criticism, social function of 191 Crockett, David 143 Crow 172 Cuisenier, Jean 159 cultural: past 62; relativism 4; rules 58 culture 31, 219: popular 194 cumulative: change 60; ideology 62; mathematical reasoning 223; patrifiliation 48 Cunnison, Ian 163f curate 82, 98 curriculum, school 94 cybernetics 136 cycle, domestic 146 cyclical: instabilities 220; processes 146; time 156 Dahrendorf, Ralf 11, 194 Dallas 63 dancing 93, 100 Danish colony 83 Darwin, Charles 35f, 182 data: analysis of 14, 22; disguise of 142 data-dredging 22, 25 daughter churches 42 death duties 77 decolonization 207 deconstruction 228 default value 11 defeat of 1898 123 Delhi 171 delimited collectivity 12 democracy 79 demographers 174 demographic: evolution 163; factors 56 demography 134 Denmark 95 density, population 64 descent 46: groups 47 descriptive: models 23; sociology 173; sociology of kinship 178 determinist models 58 developmental cycle 159, 178 Devons, Ely vii, 173 diachronic: interpretation 146; models 146, 158, 161; processes 60; structures 158 diachrony 152, 155ff, 159: and RadcliffeBrown 157 dialectical relations: and practitioners 4, 9; real world and ivory tower 218; social
INDEX
263
idiom of agnation 41; and Meggitt 56ff Engels, Frederick 154, 175f engineering, social 210 England, Elizabethan 197 Enlightenment 95, 210 enterprise, individual 54 enthusiasm 179 environmental: degeneration 195; factors 158 epistemological; populism 4f, 222; status 3, 152, 216 epistemology 2 Epstein, Arnold Leonard vii equality 74, 80 equilibrated colonialism 117 equinoxes, precession of 160 equity 147 Erastianism 95 error term 40 esoteric knwoledge 12, 119 esoterica, study of 194 espionage 204 essays 121 Easton, David 174 ethnographer 90 ecclesiastical organization 8 3 ethnographic: data 57; monograph 13; eclecticism, analytical 135 ecological differences 54 present 89, 120; record 57 ecology 184 ethnography 1 5 5 ethnology 155, 160 economic class 1 72 ethnomethodologists 4, 22, 24 economics: and anthropological ethnomethodology 2 quadrivium 169, 173; data and theory ethnoscience 34, 39 183; and elaborate techniques 21; and ethology 184 routine science 1 3; as science 11; selective recruitment 19; as social science etic 32, 216: categories 191; concepts 188 134, 198; and sociology 184; and Etoro 171 unconscious processes 159 Euclid 224 Europe, medieval 54 economists 11,15 Evans-Pritchard, Edward Evan: and education 76, 91 educational attainment 18 humanities 11; and Levi-Strauss 151; egalitarian class relations 76 and lineage systems 52, 56; and Egypt 38, 112 meteorology 33ff; and Nuer 45, 61; and Einstein, Albert 218 political philosophers 1; and religion 88 elections 80f, 148 events, study of 1 79 exchange cycles 163 elites 87, 199, 210 executive council 79 embryonic diapause 40 exegesis industry 22 3f emergency, Indian 203 exogamy 1 72 emic 32, 188, 216 experts 191, 199 emigration 99 explanations 23, 191, 193; transcendental empathy 12, 20, 198 empires, world-wide 86 29 empirical: findings 9; generalizations 24; external: relations 127; system 78 observation 155; research 18; scientific extrovert: aims 221; potentiality 218 inquiry 15, 182; sociologists 13; studies of society 183 faith and works 89 empiricism 2, 30, 202 Fallers, Lloyd Ashton 112, 1 74 empiricist tradition 153 falsifiability: and network analysis 222; and employers 13 propositions 135, 178; and science 12, Enga, Mae 45f: and agnatic model 48f; and 198
science and objects of study 192 diastole 60 diffusion 6, 21 dingo 40 Dinka 46 disciplinary boundaries 177 discontinuity 155 disguise 142 district officer 124, 142 divine right of kings 113 division of labour 2, 221: anthropology and history 151; anthropology and sociology 173, 177 doctors 12 dogma of descent 54 domestication of perceptions 137 double diversity 19 Dreaming 164 Dumont, Louis 206 Durkheim, Emile 10, 157: position in France 1 74; and social facts 1 76
264
INDEX
fatherhood 29f Faust 226 feedback 134, 136: and Crestwood Heights 144; and own social life 149; observer and actors 20, 146 fertility clinics 41 festschriften 1, 30, 150 Feyerabend, Paul K. 217 field principle 51 fields, social 70f field work universe 173 filiation 31 findings of social inquiry 7 finite mathematics 172 Firth, Raymond: and microsociology 173, 178, 180; and optation 49; and social anthropology 210 fishing 69, 78: herring vi, 70; industry 78, 82; vessels 70f fishmongers 37 fissile states 106, 111 fission 52f flexibility 58 folk tales 143 folk-urban continuum 68 food gatherers 106, 109 forces, basic 135 Forde, Daryll 52, 16If Fore 46 foreign social scientists 203 forgetting 228f formal science 33 Fort Jameson 120: Ngoni see Ngoni, Fort Jameson Fortes, Meyer 30: and demographic factors 56; and descent group 47; and festschrift 30; and field principle 51; and irreducible principle 152; and kin relations 176; and lineage segments 52; and non-cumulative model 61; and organic societies 41; and political philosophers 1; and structure 165; and Tallensi 45; and transcendental explanations 29 Foster, Brian Lee 223 Francis of Assisi 4 1 , 63 Francke, August Hermann 94 free enterprise 93 freedom from being studied 211 French: anthropology 212; colonialism 212; Revolution 84, 159, 163f fresh data 13 Freud, Sigmund 154, 225 friendship 67, 72 frontier 111: contacts 23; zone 10 fun 218, 225 fundamentalism 92: Marxist 175
funerals 126f fusion 52 Galileo 224 games, theory of 147 Ganda113 gardens 49 Geisteswissenschaft 133 genders vi genealogical: connexions 47; mathematics 171, 178; structure 47 genealogies: Bedouin 56; as charter 48; and formal analysis 179; matrilineal 228 general public 4 generality in models 220 generalizations, universally valid 191 generations of computers 42 genes, selfish 176 Genesis, Book of 36 genetic: connexion 176; programming 42 geneticist 147 genitor 38 geography 135 geophysicist 147 geophysics 145 Germany 96 gestation 34 ghetto 221 ghosts 59 Giddens, Anthony 4, 10, 13 Glacier Metal Works 142 Gluckman, Max 173 gobbledegook 6f, 189 God 40, 218: High 186 Godelier, Maurice vii Goethe, Johann Wolfgang von 226 gold rush, The (film) 166 goodness of fit 22f Goody, Jack vii Gough, Kathleen 210 government 118: by committee 82; chiefs 114 governmental embrace 19 grace 92f graph theory 3, 216f gravity, law of 197 Great Rift Valley 109 Greek 199: tragedies 224 guilds 21 Habermas, Jiirgen 13 Haddon, Alfred Cort 183, 186 Ham, Johan 35, 39 Hamlet 166 Hancock, William Keith 103 Harary, Frank 217 harder facts 22
INDEX harmonic ambiguities 155 Hartsoeker, Nikolaas 39 Hastings, Warren 210 Hauge, Hans Nielsen 91, 97: and Christian witness 93; and sectarianism 96 Hawaiian prince 117 Hawaiians 108 Hawthorn, Geoffrey vii headman, village 110, 114, 127 head-taxes 109 heaven 96 Hegel, Georg Wilhelm Friedrich 154 Heisenberg, Werner Karl 137, 159, 184: and principle of certainty 137 Heraclitean view of reality 165 hereditary 114 hermeneutics 216 herring 69f: red 44 Hertwig, Oskar 36 Hertz, Heinrich Rudolph 18 Hertzsprung diagram 146 heterogeneity 120 Hiatt, Lester Richard vii hierarchical structure 79 hierarchy 9 highlands of Papua New Guinea 44f: and managerial skills 119; and no centralized authority 106; and peaceful conquest 107 hill-dwellers 62 Hindu: canon 206; norms 202; sociology 206; texts 200; traditions 200 Hinduism 209 histoire 163 historians 2 historical: determinism 162; ethnology 152 history: against system 153; and anthropology 151; as art 136; based on documents 121; cannot be ignored 90; and computer 163; concept of 120; and ethnology 155ff; of ideas 3; many meanings of 161; myth with dates 163; of national liberation 199; on our side 164; preserved in names 127; proper 128; scientific 192; and structure 150; and thread 226; tribal 124, 128, 144; vernacular 121 history-for 163 Hobhouse, Leonard Trelawney 182 Hofstadter, Douglas R. 225 Hogbin, Ian 114f Holy Ghost 37 Homans, George Caspar 12 Home Guard 83 homeostatic models 159 homunculi 39 honour 5
265
Hopi 156 Hopkins, Gerard Manley 198 Horn, Andrew 3 7 housing 182 Houston 63 Hoyle, Fred 184 Hughes, Everett 16 Huli 46 humanist 12, 186: and empathy 12, 185 humanistic thought 192 humanitarianism 207 humanities: and anthropology 11; books in 9; and culture 197; defined 185; and natural sciences 208; and social sciences 180; and sociology 194 humanity 24 Hume, David 165 hunting 93, 124 hypergraphs 223 hypotheses 8, 135 iconic models 216 iconoclasm 180 iconoclastic discipline 19 identity, sense of 18 5 ideographic studies 200, 208 ignorance 34, 39 images, visual 225 immanent necessity 153 imperialism 196 incest 171 income differences 75 independence, political 198 independent churches 100 indexial meanings 31 India 11, 169, 196, 199 Indian: chemistry 206; civil service 16; emergency 203; government controls on research 203; modernization of 200; princely state 107; sociology 206; values 202 Indians: in Africa 123; progressive 207; see also American Indians indigenous: anthropology 196; model 113; public servants 115 indirect rule 112 indoctrination among Ngoni 123 Indology 206 Indonesia 86, 104, 199 industrial: activity 70; integration 108, 117; system 71 industrialization 87 infinite regress 225 information 140: flow of 142; processing 223 infrastructures 160 Ingold, Tim vii
266 in-jokes 7 Inner Mission 96 insider view 206 insights 8 instincts 176 insulation, effective 186 intellectual territorial rights 177 intentional phenomena 21 intentions 198 interaction 226 interdisciplinary boundaries 176 intermediate society 73 internal: relations 127; system 78 internalist standpoint 3 interpretation 9, 21 interstitial status 21 intersubjectivity 198, 206 interviewer, in-depth 187 interviews 13 interpersonal relations 179 interrelated proposotions 11 intrigues 140 introspective model 225 introvert: aims 221; potentially 218 introverted fun 218 investigation, self 149 investigator, field 138f; 145 IQ testing schedules 15 Iron Curtain 4 irreducible principles 152 irreversible: changes 60; time 155 Islam 89 isomorphism 152, 162 Israel 87 ivory tower 29, 218 Jacob 36 Janus, two-faced 151 jargon 5, 7, 223 j-curves 6 Jesuits 95 Jews 95, 117 Johsnon, Gisle 93 Joseph, Keith 22 judicial organization 83 justice 147 juvenile delinquency 201 Kachikoti 209 Kaffirs 107 Kaldor, Nicholas 21 kangaroos 40 Kant, Immanuel 154 Kantianism 153 Kariera 156, 172 Kelly, Raymond Case 56, 65 Keynes, Maynard 222
INDEX Kiel, Treaty of 98 killing for its own sake 54 King: George 122; Lear 137 kingdoms of god 89 Kinsey, Alfred Charles 144 kinship 34: algebra 171; and biology 177; cognatic 72; connexions 81; consanguineal 176; difficult subject 169; and disequilibrium 160; and material world 29; and nature 30; no such thing 171; polity 40; and poly-segmentation 60; studies 30; symposium 44; system 67; terminologies 29 knowledge 218 Kroeber, Alfred Louis 159 Kuhn, Thomas 13, 39 Kuma 46, 58 Kunda 122 Kyaka 46 La Flesche, Francis 166 Laban 36 labour: migration 113, 121; supply 108 Lakatos, Imre 30, 217 Lamba 229 land 58, 70: claims 17 language 5: cross-cultural 190; country 98; local 199; private 187; vernacular 186 Latin 142, 175, 199 Lave, Charles Arthur 225 law, English 183 laws 9 lawyers 12 lay: speech 6; view of society 187 Leach, Edmund Ronald 38, 155 leadership 54, 78, 146: from behind 67, 80 Lebanese Muslims 56, 62 Leeuwenhoek, Anton 39 legal codes 121 legends 90, 121: associated with particular groups 126; written 128 lese-majesty 37 Levins, Richard 217 Levi-Strauss, Claude 150ff, 179f: and Aborigines 219; and French colonialism 212 Levy-Bruhl, Lucien 33 libel 16, 143 liberation: national 199; postcolonial 209 lineage: segmentation 164; systems 56, 61 linguistics 135, 171: structural 188; and synchronic analysis 157; and time 167 linguists 157 literacy 94 literature 136, 199 local group, dominant 51
INDEX log linear analysis 14 London: kinship 173; School of Economics 182; suburb 188 Los Angeles 63 Louvre 137 lower class 76 Luapula valley 164 Lugard, Frederick Dealtry 112 Lukhero, Matshakaza Blackson 121 Luther, Martin 91, 96: and Melanchthon 39; and Zwingli 209 Lutheran church 69, 88, 92 Lutheranism 91, 95 Lysenko, Trofim 14, 199 macrosociology 173 macro-time 155 Mae Enga see Enga, Mae Maekali 115 magic, ministerial 22 Maine, Henry Sumner 183 Maitland, Frederic William 183 majority vote 79 Malaita 115 Malawi vii Malinowski, Bronislaw: and charter 48; and kinship 169; and kinship algebra 170f man from Mars 139 managerial skills 119 Mao Tse-Tung, quotations from Chairman 155 Maori 107f, 228 March, James Gardner 225 marine-engine factory 69, 85: and alcohol 100; hierarchical structure 82 marital breakdown 193 market: forces 17; place, language of 187; research 187 marketing organization 71 marriage 124, 162, 164 Marriage and the family among the Nuer
(book) 46 Marshall, Alfred 183 marsupials 40 Marx, Karl 22, 74, 154 Marxism 175ff Marxist: eschatology 164; physicists 4 Marxists 164, 201f, 205 Mary, Blessed Virgin 37 maternal influences 36 maternity, denial of 38 mathematical: Brahmins 221; games 179; methods 189; models 217; psychology 221; reasoning 223; sophistication 11; symbols 187 mathematicians 167: frustrated 219
267
mathematics 3, 8, 135 matriarchy 176 matricide 38 matrifiliation 40 matrilateral links 50 matrilineal: line of descent 229; societies 41,228 matriliny 158, 172 matrimonial alliances 51 mayor 8If measurement 8: scales 135 mechanical: mice 14; time 157 media 143 medicine, Western 199 Meggitt, Mervyn John vii, 45, 56: and Enga, Mae 57ff Melanchthon, Philipp 39 Melanesia 44 Melanesian: markets 2; phase 60 members 4 Mendi 46 mentrual blood 38 mental constraints 153 Menzies, Robert 147 Merton, Robert King 6, 10 mesh 73 meta-language 154, 194: scientific 185 meta-modelling 217, 225 metaphor 224, 226 meta-process 159 metascientific status of social science 199 meteorology 34f methodological principle 24 micro-sociology 173, 179f micro-time 155 middle-range theories 10, 13 Miller, Donald Bruce 202 mineral wealth 108 mining companies 17 Mirror of justices 37
mis-modelling, 219 missionaries 108 missionary societies 79, 81, 98 missions 122 mitosis 53 model: arrogant 182; beauty of 218; definition of 217; undetermined 18 model-building 152, 155f modelling 216 models 9, 22, 164: foreign 219; of society 193; realism in 220 modernization of India 200 monkeys 34, 184 monolithic orthodoxy 10 monopaternity 35 monsters 39 Mooli 41
268
INDEX
Moowamb 46, 55 moral implications 17 morale 7: in inner cities 222 mores, today's 144 Morgan, Lewis Henry 169f, 175 morphological classification of societies 84 Mosman 147 mother-child relation 42 motherhood 29, 37, 42 mother-in-law 40, 125 mother's milk 36f mother-surrogates 42 mountain-dwellers 62 moving frontier 107 muddle, as methodological assumption 24 multiple: affiliation 49; regression 13 multiplex relations 188 mundane: actors 13; style of exposition 153 Murdock, George Peter 158 Murngin 167 museum specimens 120 music 166 musical score 167 mysticisim 14 mystique of professionalism 14 mythology, world of 161 myths: of former customs 124; and history 163f; interminable 160; that think themselves out 153 naivety 172, 223 Natal 111 nations 104 native: councils 115, 119; courts 115; urban court 125 naturalists 21 natural science paradigm 133: in colonial situation 20; described 5; inapplicable 20 and network analysts 222; and positivists 191 natural sciences: books in 9; contrasted 198, 208; paradigm shifts 217 natural scientists 187 natural-social divide 10 nature 33, 217: of society 187 Navaho marriage 162 Needham, Joseph 39 negotiation 211 neolithic models 160 nepotism 113 network: analysis 21, 215, 217; analysis ancestors 226; cohesion 52 networking 6 Networking Institute 7 networks, social 6, 72, 179 neurotic sentences 225
Newport, George 35 Newton, Isaac 184 Newtonian mechancis 156 New York State, town in 140 Ngoni, Fort Jameson vii, 120ff: defeat in 1898 122; kinship 67; language 122; military conquest 107; native courts 125; primitive state 106 Ngoni of northern Malawi 107 nodes 215, 225 nomothetic study 200, 208 non-agnates 46 non-cumulative models 6If non-literate tribe 121 non-localized social groups 113 non-unilineal systems 45f normative: connotations 11; social theory 201 norms 179 North Sea vi Northern Rhodesia Regiment 124 Norway vif, 67ff, 88ff: constitution 95, 98, 101; history of 84; state church see Lutheran church; thought 74; transmission of farms 146 Nsenga 128 nuclear non-proliferation treaty 207 Nuer 45, 54: chronic segmentation 53; difficulties in modelling 61; dominant local group 51; economics and cosmology 88; manipulation of genealogies 48; no central authority 106; structural amnesia 228 Nupe 106, 118 Nyanja 123, 128 objective meanings 31 objectivity 189, 205, 208 observation 8 observer 5 occupations, ranking of 143 official reports 128 Official Secrets Act 143 oil, North Sea vi Omaha 172 omnidirectional kinship system 67 one-party states 102 one-sided wars 104, 106, 119 ontological status 216 ontology 1 opinion: polls 188; surveys 14 ordinal-level data 221 Orestes 38 organic societies 41 organism 217 oriented time 162 orienting statements 194
INDEX origin of ideas 177 Orthodox church 95 orthodoxy, lack of 188 orthogonal axes 156 Orwell, George 90 Osage of Missouri 166 oscillation 220 oscillatory meta-model 222 outsider view 206: of culture 212 ovum 35 Oxbridge colleges 42 Oxford 67
269
physiological paternity 39 physiology 34f Piddington, Ralph O'Reilly 219 Pietism 88, 91, lOOf: and theological college 98 piety 94 pigs 59 Pike, Kenneth Lee 32 Plantade, F. de (Dalenpatius) 39 planters 124 Plato 154, 223f Plato's cave 160 Platonic: G I store 152; quality of science Pacific Ocean 107 11 palaeolithic: models 160; conditions 191 play with ideas 210, 225 palm oil 162 Plaza-toro, Duke of 67 pandanus 52 plural societies 105, 119 pangenesis 36 plurality of models 10 pan-primate political order 34 Pocock, David Francis 206 Papua New Guinea 40, 103, 105, 116f: see poems 137 also highlands poetry 154 paradigms 8, 39, 216f poets 155 paramount chiefs 117, 123f Poincare, Henri 2 parish: church 83: council 79, 81, 93; and points of nodes 215 Pietism 96 police 22, 115: actions 104ff Parisian code 154 policy-making process 119 Parsons, Talcott 16, 152 political: behaviour 90f; ideology 188; participant observation 20 order 34; philosophers 1; science 11, particularist: policy 92; sociology 191 134, 198; scientists 174, 188; stance paternalism 112 19, 196 path diagrams 23 politicians 118 paths in network 73 politics 29, 118: and quadrivium 169, 173 patrifiliation 40, 64 Pollock, Frederick 183 patriliny 158 pollution 37, 207 pax: australiana 106, 117; britannica 106 polygyny 65, 123 peasants, part-time 87 polypaternalism 36 pedigrees 4 1 , 72, 217 polysegmentary societies 4 1 , 45 peerage 228 polysegmentation 52, 60 pendulum, swinging 156, 178: and polysemic attributes 6 network analysis 220, 222f Pope, the 91 Pentecostalists 101 Popper, Karl Raimund 217 perceptions: of past 90, 120; of social popular: assumptions 7; wisdom 8 science 196 population 56: pressure 195 periodic time 146 positivism 223: and epistemological personnel management 192 populism 222; and kinship 176; and pertinence in models 220 linguistics 32; and Verstehen 205 petard 226 positivist-subjectivist debate 206, 209 Peters, Emrys Lloyd 56 positivists 23 phases of conquest 106 post-colonial societies 84 philosophers 1, 187, 196 poverty 182, 207 philosophy 1, 23, 223: and political science power 5, 218: balance of 212; knowledge 11; of science 2, 216 as source of 209f physical: anthropology 176; facts 44; practice 1 illness 199 praxis 13 physicists 218 prayer and perseverance 41 physics 198, 221: forces of 135: language prayer house 79, 98 of 5; national 199; particle 224 prayer-ship 79
270 precedents 126 precision in models 220 pre-clinical medical faculty 12 predictions 6 prehistory 59 pre-industrial societies 172 pre-paradigmatic stage 194 pre-scientific cultures 38, 42 prestige 143f, 188 pretentiousness 190 priests 98 primitive: states 106, 112; warfare 105 primogeniture 147 princes 96, 101 printed treaty forms 106 private lives 194 privilege of the sceptic 165 privileged status of natural science 2 3 problems 194, 218f: easy 218; questions and 194; technical 219 process 155, 226 Procrustean categories 80 professional: associations 19; closure 7; sociologists 16; status 15 professionals 15 progress 226 progressive: Indians 207; time 155 property, knowledge as form of 210 propositions 9ff, 135: testable 194 protectorates 112 provincial: administration 81; advisory council 125, 127 pseudo-events 223 pseudo-procreation rituals 39 psychological geography 72 psychology 11, 134: and legal protection 15; partitioned 11; and routine science 13 Ptolemy 184 public: opinion polls 138; policy 6, 13 publication 138, 140 purity 207 pygmies 109 quadrivium, anthropological 169, 173 quantification 11, 42, 178 quantitative data 14 quantum physics 137 quarks 5, 224 quarrels 126f quasars 135 quasi-agnates 63 questionnaires 139 questions, refusal to answer 211 quicksand 8 race 103, 188
INDEX racism 207 Radcliffe-Brown, Alfred Reginald 157, 165f, 169 rain 33: shrine priests 115 rainfall, Arnhem Land 167 rain-making ceremony 23 rationality 24 rationalizations 23 real: social world 29, 216, 219; time 134, 136, 145, 167 reality 226 Realpolitik 11
recruitment 19 rector 82, 94 red herring 44 Redfield, Robert 68, 202 reflexive behaviour 223 reflexivity 2, 5 Reformation 95 refraction, laws of 226 regiments, two 97 regularity principle 157 rehearsals of social action 148 relations: actual set of 178; of uncertainty 158; study of 224 relative: deprivation 6; worth 80 relevance 204, 208 religion 24, 29: and quadrivium 169, 173 religious political party see Christian People's Party Renaissance 185 rennet 38 replication 143f, 148 representative: government 34; status 147 research 223 residence patterns 58 respectability 193 restrictive recruitment 16 rethinking 171 reversible time 155 revolt 105: citizens' 210 Revolution, French 147, 159, 163f Rex, John 42 right hand 97 risk 154 ritual leader 114 rival schools 10 Rivers, William Halse 183, 186 Roman rule in Britain 8 7 Romans 112 routine science 13, 217, 221 Royal: Anthropological Institute 16; Society of London lOf, 39 Samoa, eastern 107 samples 13 sampling 179
INDEX San Francisco 4 1 , 63 Sanskritists 206 Sartre, Jean-Paul 152, 163 saturation effects 211 Saussure, Ferdinand de 157, 162 scale 103, 164 scandalmonger 141 scepticism 190 Schneider, David Murry vii, 30, 53ff: and categories 4 1 ; and culture 169, 173 scholars 186 scholasticism: idle 198; introverted 219 school: of skills 94; projects 20 schools: for adolescents 91, 99; rival 188 science 134, 197f: dealing with facts 31; fiction 168; hard 193; natural 193; pure and applied 17f, 221; universalistic 199, 206 scientific: methodology 2; research programmes 30 scientism 189, 202 scientist 12, 185 Scientists against Nuclear Armaments 18 Scylla 156 seasonal cycles 163 Second World War 7 secrets 139f, 211 Seeley, John Ronald 144 segmentary: lineage systems 159; model 61 segmentation 53 Seidman, Stephen Benjamin 223 self-assessment quizzes 14 self-awareness, professional 212 self-correcting qualities 223 self-referencing 225 self-reliance, path of 208 self-scrutiny 190 semen 34, 38 sentences, healthy and neurotic 225 Sermon on the Mount 89 sets 72 settlers 107 Seventh Day Adventists 91 sex: and age distribution 147; eduation 93 sexual practices 142 Shakespeare 197 Shakespearean usage 5f shared biogenetic substance 38 sheriff 82 Siane 46, 65 Sidgwick, Henry 180, 182f, 194; and Comte 191 significance tests 189 signs 162 similarities 175 simple societies 68
271
simultaneity 155, 157 Siriono 106, 109, 111 smoking 135 snake 167 snooker players 12 snuff-box 124 social: action 58; anthropologists see anthropologists, social; engineering 220; events 223; evolution 175, 183, 192; facts, autonomy of 176; field 137; gradient 73; inquiry 14; laws 24; networks 6f, 67; norms 202; organization 165, 172; and Political Sciences Tripos 183; science foundations 189; sciences 144, 180; technology 191; theory 13f; work 192, 200 social anthropology 198: as illusion 144; as cheap science 12; and colonial conditions 184, 187; and denigration 8; and dialectical relation 187; differences in praxis 196ff; andfieldwork techniques 73; and history 151, 158; history of 16f; indigenous 196; as social science 134, 198; and sociology 169; and time 157 social class 74ff: as illusion 144; in India 201; in London 188; in Norway 67f; subjective 31; as ties of equality 72 Social Networks (journal) 216 Social Science Research Council 22, 166 social sciences, immaturity of 10 socialization 171: of children 193 socially constructed reality 22 Society for Social Responsibility in Science 18 sociobiology 176f socio-economic status 6, 188 sociography 146 Sociological Association of Australia and New Zealand 15, 134 sociologists 1 73f sociology 134, 198: and anthropology 12, 172, 176; central position of 182; of the family 174; history of 15f; invention of 219; and Levi-Strauss 155, 192; proper 146; relation to objects of study 184; of religion 88; routine science in 13; rural 174; school of 135 sociometric tests 144 solidarity, group 49, 51, 52 sorcery 20 Sotho 111 South Africa HOff, 117, 203 sovereignty, Tudor 185 Soviet Union 4 Spencer, Herbert 182 spermatozoa 35, 39 spies 203
272
INDEX
Spiro, Melford Elliot 38 sponsorship 204 Springdale 139 Stanner, William Edward Hanley vii stars 146 state 4: church 95 stateless societies 55, 113 statistical: models 161; significance 14; techniques 189; time 157 status 5 status quo, disturbing the 190 stochastic variability 56, 65 stone tools 5 7 strangers 197, 211 strategy 178 stratification, social 147 structural: amnesia 228; analysis 154, 215; dialectics 162 structuralism 179f, 224 structuralists 165 structure 155, 159, 179: and history 150 student unrest 180 students 94, 192: research 193 study of societies from within 192 studying others 19 subject matter of the social sciences 4 subjection of women 99, 176, 193 subjective: meanings 205; social class 31 subjectivism 209 subjectivity 205 subordination structures 160 subsections 156, 190 subsistence economy 104 suburban housewives 190 successiveness 155, 157 sui generis 166 suicide rates 147 Sunbelt network conference 216 Sunday newspaper supplements 6 surfies 137 survey, social 146: street 20; techniques 13 Swan Lake 137 swaraj 207 Swazi 111 Swedish tutelage 98 swings, damping down the 222 symbolic: codes 205; models 216; systems 171 symbolism of kinship 177 symbols 31, 178f synchronic: data 146; evidence 146 synchrony 152, 155ff systematic: misunderstanding 24; regularities 22
systole 60 tactics 178 Taiwan 110 tales of fighting 121 Tallensi 30, 45; dominant group among 51; and matrimonial alliances 51; no centralized authority 106; and noncumulative model 61; polysegmentary 48; and segmentary fission 53; and structural amnesia 228 Talmud 41 Tavistock Institute 220 Tax, Sol 172 taxes 81 taxonomic status 205 taxonomy, zoological 74 teaching of tribal history 124 technical: language 9; unnecessarily 215 technicians 205, 219 techniques: of inquiry 14; sophisticated 179; value-free 203 technology 223 telegony 36 telescoping 61 television 100, 143 temperance hotels 98 temps modemes, Les 154 tenses vi: ethnographic present 89, 120 terminologies, kinship 170, 172, 177; formal analysis of 179 terrorists 105 Texas 63 texts, imported 208 theologians 190 theological: college, Pietist 98; rationalism 95 theology 88 theoretical preferences 19 theoreticians and experimentalists 221 theory: construction 13; testing of 192 therapeutic investigation 142 therapies, indigenous 199 therapy and research 220 Third World 197, 211 Thurston, Captian 112 ties or arcs 215 time 150f, 165: depth 53; machine 168; not on our side 195; real see real time; the refreshing river 165; undulating 156; various kinds of 155 Tiv 45, 54: and dominant local group 51; and genealogies 48; and segmentary fission 53 Tonga, Plateau 106, 114f topology, plausible 220 Torres Straits 186f
INDEX totemism 165, 186 Toynbee, Arnold 159 traders 109 traditional: assembly 113; leadership 116 traditions, great and little 202 trade union presidents 117 Transvaal 11 Of treaties 105 treaty-making 107 trends of change 59 trial votes 79 tribal: custom 125; history 128, 144; monograph 169; political structures 104; religion 89; sequences 72 tribes, African 190 Trinity 37 Trobriand kinship 170 two-way communication 32 tyrant 115 Uberoi, Jitendra Pal Singh 177, 207, 209 Uganda 112f Ullman, Walter vii unanimity 79 uncertainty, principle of 13 7 Uncle Sam 4 1 , 63 unilineal descent systems 31 United: Kingdom 15; Nations 104; States 4, 16 unity, group 80 universalism 198 universalist policy 92 unmarked term 11 urban indigenous proletariat 108 useful arts 198 useless research 204 utility 218
village studies 202 villages 124 violence 54 virgin birth 34 virilocal residence 125 voting 80f wage labour 59 wages 182 Walbiri 40 wallaby 40 war 53, 59, 105, 116: endemic 194 watchmen 24 weapons, intellectual 219 weaving 226 web 72, 226 Weber, Max 10, 89, 174 week, day of the 145 welfare state 93 Wells, Herbert George 168 Weltanschauung 209 wet nurse 3 7 wheel 217 Whiggish view of social science 10, 224 Whitten, Wesley Kingsley vii wills 77, 125 wine monopoly 100 wirken 226 Wirklichkeit 226 Wisa 122 wisdom 218: traditional 191 women: observers 140; status of 193 women's institutes 81 work satisfaction 201 writing 120: letters 121 xenophobia 204
valuations 18 Yako marriage 161 value-free: social science 20If; sociology 18 Yale, David Eryl Corbert vii Yanagisako, Sylvia Junko 178 values 189 youth clubs 93, 101 vectorial factor 158 Vedic times 199 Zambia vii, 67 Venus deMilo 137 vernacular literature 121 Zanzibar 112 Verstehen 205 Zelditch, Morris, jr 162 Victoria (Australia) 15 Zeno216 Victoria, Queen 186: coronation of 147 Zulu 106, 111, 123 Vietnam War 20 Zwangendaba 128 vigilance 225 Zwingli, Ulrich 209 Vikings 84
273