Essays on the Sociology of Perception

Mary Douglas

• Publisher: Taylor & Francis
• Year: 2013
• Print ISBN: 9780415606660
• eBook ISBN: 9781134557509
• Edition: 1
• Topic: Social Sciences
• Subtopic: Anthropology
• Index: Social Sciences

First published in 1982, this is one of Mary Douglas' favourite books. It is based on her meetings with friends in which they attempt to apply the grip/group analysis from Natural Symbols. The essays have been important texts for preparing grid/group exercises ever since. She is still trying to improve the argument of Natural Symbols and is always hoping to find better applications to illustrate the power of the two dimensions used for accurate comparison.

Part One

PERSPECTIVES ON METHOD

INTRODUCTION

Mary Douglas

David Ostrander organized the original conference in 1978. He wrote for it an introductory paper which aimed to introduce grid/group analysis by contrast with famous attempts to typologize social experience. Ingenious and clear, perhaps he oversimplified the great nineteenth-century sociological types and present-day theoretical contrasts. But he succeeds in his mission, which was to lay to rest the nagging sense of familiarity, ‘where have I heard this before? How does it differ from what I have heard before?’ He sets the effort of this book in a historical context and introduces his persuasive insights on one-dimensional and two-dimensional typologies.

Then follows Michael Thompson’s bold improvement and solution to many of the problems which will later be encountered. He asks tricky questions which will be in the following essays: can two cosmologies co-exist in the same social context? How does sudden conversion take place? Where to place the hermit recluse? James Hampton feels that several problems which I would have tried to treat by closer control of focus could be solved by adding a third dimension of ‘activity’. The hermit would have the same grid/group position as the individualist entrepreneur, but he would lie at the extreme of the social activity dimension and the entrepreneur at the other. Pusillanimously, I prefer to leave the hermit off the map of social controls, crediting him with full escape. But see how Michael Thompson accommodates him comfortably at a zero centre of his three-dimensional cube, the third dimension measuring the possibilities of exerting power. He uses the third dimension to construct a model which could look like a plane surface laid over an unevenly-carved-out cube. It has four stable plateaux at the same four corners of the two-dimensional diagram, but he locates them at different levels in the third dimension. The impossibility or probability of sliding from one to another could be calculated if certain specified information were given. The distribution of power accounts for the pressures and barriers to change. In the middle of the cube he finds a fifth stable habitable region: it is the hermit’s cell, away from power, alone, yet a model and enticement to those in society.

I have a difficulty, common to non-numerates, in finding two dimensions rather much to handle consistently. The thought of a third boggles me. My instinct is to squeeze as much information as I can out of two. However, these suggestions are intriguing. Michael Thompson uses the methods of typology and the geometry of catastrophe to explain his development of the theory. He uses the description of cusp catastrophe as a powerful metaphor which could be turned into a set of measures for predicting sudden change. He shows that on some places on the undulating surface presented on his cube two cosmologies may be held in suspense. In any particular dilemma each may be equally appropriate, but overlapping alternatives are available as justifying resources to individuals in moral predicaments. Only at his central zero does the option to choose remain continually open. He shows what deliberate action people have to exert to keep their social relations just poised like that, with no exits closed. It is a precarious balance and at any time pressures to slide down the slope or to climb up may shift individuals out of that calm shelter. Then they will perforce choose the appropriate cosmological scheme which makes sense for where they now are. His main argument for the third dimension is that it accommodates the cases of societies which cannot be fitted into any of the four positions which I have suggested. One of the special values of this book is that Michael Thompson illustrates the limitations of a two-dimensional model by working out how to turn it into a three-dimensional one.

In the next chapter James Hampton describes how he conducted and analysed two small surveys in London in 1976. Scoffers had told us that as one descended the grid scale, as first described, one would automatically be slithering towards stronger group. James Hampton demonstrated the two constructed dimensions of grid and group to be really independent. This was a major advance. But unfortunately our samples did not include representatives of very strong group measures. The small differences between rather weak group membership do not show up anything noteworthy. The main interest of James Hampton’s work lies in the questions he raises about objectivity, psychology and mismatch between predicted cosmology and social context. He remarks that it seems to be possible for a person to occupy different positions on the map according to social context. Responses to his questionnaires showed that someone might be in a free individualist environment at work and in a compartmentalized and regulated environment at home. At first, I rejected this possibility. If someone could behave as if in a high-group context in the afternoon and a low-group context in the morning, or change his grid/group score between Tuesdays and Saturdays it seemed to undermine the whole value of the method. But reflection on his results made me modify my view. If one knows anyone who works in an intensely competitive business, where no holds are barred, the weakest goes to the wall, great prizes to the swift and so on, one has seen why such a person might try to get his home life working like clockwork so that in every detail it could be absolutely relied upon not to distract from the office jungle. Then such a person would indeed be creating two totally different contexts. Even more than his family life, his office might reflect tight regulative controls that he has imposed. The people subordinate to him are up-grid, the equals he negotiates with are his world of low-grid individualists.

As new illustrative material came in for this volume, the proper uses of this method received necessary refinement. If we are talking about grid/group values comparatively, we must compare like with like as far as possible. Then the homes of businessmen in a given country could be compared and we could ask how they deal with domestic matters when under severely competitive business pressure. We could compare the women in their domestic or work scenes. We could not justifiably jump from the work to the home as if the scale of operations or the economic or social values were the same. The art of the method is to be very delicate in matching the cases compared, usually sticking to similar ethnographic materials. We will see how Martin Rudwick compares geologists with geologists, Michael Thompson compares Sherpas with other Sherpas, David Bloor compares mathematicians with each other, Celia Bloor compares young post-doctoral industrial scientists. The richest results come when the ceteris paribus rule is most carefully protected. James Hampton’s airing of this particular problem leads to a hypothesis: the more hotly competitive the society of individualists, the more those in the front ranks of competition will tend to regulate their followers, driving them up-grid. So we would expect women, cripples and children to be strongly regulated in a strongly competitive society, expressed public sentiments to the contrary notwithstanding. There are two answers to James Hampton’s query whether multiple cosmologies may not be lodged in one person’s head. The first answer is to be very careful, minutely precise about maintaining the same scale of comparisons, both in the social and cosmological parts of the investigation. The second answer is yes, obviously a person can behave in any one day as an autocrat at the breakfast table and meek as a lamb at the office; but by tracing such cases we can discover further patterns of different parts of the diagram.

James Hampton’s other question about how to achieve objectivity in an investigation is partly answered in the later paper he wrote jointly with George Gaskell on styles in accounting. His remarks about individuals launched somehow into a new part of the diagram and facing problems of conversion or dissonance with their fellows point to desirable collaboration with psychologists.

James Hampton also fears that the majority of people surveyed will fall in ‘some central grey area of eclectic, loosely integrated cosmologies’. Again, I have my simple faith in the instrument’s capacity to be made precise. It only works where the role structure can be clearly identified. In modern industrial society it works well within distinctive professional classes, when objectives and fields of interest can be clearly shown to vary on grid/group criteria.

The most accessible of the attempts to apply grid/group analysis to the sociology of knowledge is Celia Bloor’s analysis of her interviews of young industrial scientists. The problems which she and her husband surmounted when they tried to allocate grid/group scores, on social experience and theoretical bias, to the interview records help us to understand how this method can be used. Their sensitive illustrations of how an industrial scientist is likely to think of his measurements and theories if he is situated in one social environment or another are suggestive. More than anything else in the book they raise the question for social psychology: did these types select their social niches or did they, in one year, adapt so thoroughly as to suggest a perfect match?

Chapter 1

ONE- AND TWO-DIMENSIONAL MODELS OF THE DISTRIBUTION OF BELIEFS

David Ostrander

SOCIAL ENVIRONMENTS AS BASES OF COMPARISON

One requirement for a classificatory approach to the analysis of symbolic behavior is the elimination of ‘societies’ as the units of comparison in favor of the social environments of individuals. This follows from two rather obvious, yet frequently ignored, facts concerning the social and symbolic orders. First, societies do not symbolize - people do. Whenever we treat a society as a single analytic unit, we are quick to ascribe to it a unified, disembodied symbolic order which describes it, justifies it, and prescribes behavioral norms to keep it running smoothly. But the symbolic order exists and articulates with the social order only through the minds and actions of individuals operating for their own purposes within the confines of their own social environments.

Second, even the simplest of societies has a variety of social environments. Societies must adapt to a multitude of particular material and social conditions, thus producing a congeries of social environments, each of which may generate different symbolic associations. This makes classification of societies as wholes impractical. Social environments, in contrast, are at a second order of abstraction from the multi-dimensional impact of the outside world. They can be systematically classified by relatively few dimensions which define sociality itself.

In modern western thought, the dimensions of sociality have been most passionately discussed in the work of Hobbes, Locke, Rousseau and other enlightenment philosophers. For these thinkers, the very existence of social order was a problem to be explained. Individual human beings were products of nature - of observable physical processes. The aggregation of individuals into groups, and the subordination of individual wills to a group will, was not a natural process, and (they argued) had to arise out of some mutual interest, be it profit or survival. Despite their many differences, the enlightenment philosophers agreed in underscoring the idea that sociality involves the subordination of individuals to a supra-individual pattern of interaction and therefore limits freedom of individual action.

The two most general spheres of action limited by social order are (1) whom one interacts with, and (2) how one interacts with them. In order to classify social environments, we may treat these spheres of action as two dimensions which vary according to the degree to which individual freedom is restricted. Mary Douglas (1978) refers to these dimensions as ‘group’ and ‘grid’ respectively.

The grid/group classification is intended to have the sort of general applicability necessary for analyzing the relationship of the social and symbolic orders. A few points should be made concerning the limitations on this applicability.

1 It is a relative rather than an absolute tool, constructed of continuous rather than dichotomous variables. The four cells are primarily of heuristic value; actual distinctions among social environments may be less extreme depending on the scope of the comparison.

2 As it classifies social environments, it is technically incapable of distinguishing (as it stands) whole social systems or pan-system institutions; thus, capitalism, while ideally in square A (low/low), is operationally composed of at least A and B, and probably C and D as well.

3 The grid/group classification is not a causal model; it does not explain, or seek to explain, why a social environment changes, or an individual changes environments. The sources of such changes are the exigencies of the real world to which society and individuals must continually adjust. They are external to the dimensions of sociality and not generated from within.

4 It is not the only classification possible, or extant, which links social structure to symbolic structure; in fact, almost all social classifications make this linkage, if only implicitly. The link here is explicit.

Taking up this final point, we shall turn our attention to a selection of established social classifications in order to show how existing schemes share with grid/group an underlying concern to account for the distribution of beliefs according to variation in social experience. Such a systematic comparison may also help to clarify the kinds of things which we interpret as indices of grid and group, respectively, by relating them to classificatory dimensions with which the reader may be more familiar.

For the purpose of this paper, social classifications may be grouped into one- and two-dimensional schemes. The one-dimensional schemes (being far more numerous) may be further divided into grand dichotomies, special typologies, and evolutionary states.

Grand dichotomies

By grand dichotomies I mean those schemes which divide the entire social universe into two mutually exclusive parts. Durkheim’s distinction between two types of social solidarity is an appropriate first example. Durkheim argued that society integrates its members by exploiting either their commonalities (mechanical solidarity) or their differences (organic solidarity). Mechanical solidarity demands a high degree of conformity, in behavior and belief, to the strictures of the common conscience. Deviance from these norms is regarded as a crime against society, and is met with repressive legal sanctions. Organic solidarity, in contrast, encourages individuality. Integration depends on individuals carving out their own social niches. When deviance reaches a point regarded as criminal, it is usually viewed as an interpersonal action and is met with restitutive legal sanctions.

Durkheim presented this distinction in evolutionary terms, suggestin that mechanical solidarity was characteristic of small, homogeneous, primitive societies and gave way gradually to organic solidarity as size and internal differentiation (the division of labor) increased. A broader view of ethnography than was available to Durkheim makes the evolutionary nature of his scheme untenable. Highland New Guinea for example offers hundreds of examples of primitive societies predicated on the individuality, competition and economic exchange characteristic of organic solidarity.

Stripped of its evolutionary trappings, the mechanical/organic distinction falls into a larger class of dichotomies which, while employing various criteria and terminologies, are frequently reduced to a distinction between conformism and individualism as principles of social life. Sir Henry Maine (1861), primarily through an analysis of ancient Roman law, argued that:

from a condition of society in which all the relations of Persons are summed up in the relations of Family, we seem to have steadily moved towards a phase of social order in which all these relations arise out of the free ...

Table of contents

• Cover
• Title page
• Copyright page
• Contents
• CONTRIBUTORS
• INTRODUCTION TO GRID/GROUP ANALYSIS: Mary Douglas
• Part One PERSPECTIVES ON METHOD
• INTRODUCTION: Mary Douglas
• 1 ONE- AND TWO-DIMENSIONAL MODELS OF THE DISTRIBUTION OF BELIEFS: David Ostrander
• 2 A THREE-DIMENSIONAL MODEL: Michael Thompson
• 3 GIVING THE GRID/GROUP DIMENSIONS AN OPERATIONAL DEFINITION: James Hampton
• 4 TWENTY INDUSTRIAL SCIENTISTS: A PRELIMINARY EXERCISE: Celia Bloor and David Bloor
• 5 A NOTE ON STYLES IN ACCOUNTING: George Gaskell and James Hampton
• Part Two COMPARATIVE STUDIES IN HISTORY AND THE HISTORY OF IDEAS
• INTRODUCTION: Mary Douglas
• 6 ‘LES GENS DE LETTRES’: AN INTERPRETATION: George A. Kelly
• 7 THE POLITICAL CULTURE OF WARRING STATES CHINA: Katrina C.D. McLeod
• 8 REFLEXIVITY IN FESTIVAL AND OTHER CULTURAL EVENTS: Don Handelman
• 9 POLYHEDRA AND THE ABOMINATIONS OF LEVITICUS: COGNITIVE STYLES IN MATHEMATICS: David Bloor
• 10 COGNITIVE STYLES IN GEOLOGY: Martin Rudwick
• Part Three CLOSE FOCUS ON SELECTED COSMOLOGIES
• INTRODUCTION: Mary Douglas
• 11 THE PERCEPTION OF TIME AND SPACE IN EGALITARIAN SECTS: A MILLENARIAN COSMOLOGY: Steve Rayner
• 12 SPECTRAL EVIDENCE: THE WITCHCRAFT COSMOLOGY OF SALEM VILLAGE IN 1692: Dennis E. Owen
• 13 THE PROBLEM OF THE CENTRE: AN AUTONOMOUS COSMOLOGY: Michael Thompson
• INDEX