Natural Symbols (1973)
Cultural Theory, or CT, is a joint creation of Michael Thompson and the late Aaron Wildavsky. Their work marked an unexpectedly fruitful collaboration in the history of the social sciences, being both cross-disciplinary and cross-oceanic. Wildavsky was a renowned American policy analyst famous for his work on government budgeting; Thompson is a British anthropologist renowned as the author of Rubbish Theory (1979). Their collaboration came about because I introduced them to each other, having coauthored Risk and Culture with Wildavsky in 1982, and having known Thompson as a fellow British anthropologist interested, as I was, in dirt (of which more in a moment). Moreover, frankly, I had the original idea for CT, so you can say that I started it.
Culture Theory began life in my head many years ago as a method of controlling sociological comparisons, a way of ensuring that like was being compared with like. It has come a long way since as it moved from mere method to theory. CT now claims a large band of practitioners, more than 700 published titles, and some real-world traction. Indeed, it has become a practical guide for people working together on sometimes quite important issues, a few of which I will mention below. Occasionally I think to myself, “My, isn’t it curious what an interest in dirt can lead to forty years on?”
Matter Out of Place
The story of Culture Theory properly begins in the 1960s, when anthropology was essentially ethnographic, a one-tribe-at-a-time affair. Teachers insisted that it was comparative in intent, but anthropology has a special difficulty in trying to make valid comparisons. Social sciences are usually able to honor the principle of ceteris paribus, “other things being equal”, because their fields of comparison are drawn from the same type of late industrial society. But if anthropologists want to compare two types of ancestor worship, for example, or two kinds of belief in witchcraft, the cultural differences among their cases will often be so vast as to render vain, or at least suspect, any effort at comparison.
I wrote Purity and Danger in 1966 hoping to sidestep that difficulty by proposing a universal cognitive framework that would be true of all cultures. Interested in religious doctrines of defilement, I hypothesized that such doctrines were of the same order of behavior as secular attitudes to defilement. If so, then our understanding of certain key aspects of religion would be improved if we better understood attitudes toward dirt. Using the flexible definition attributed to Lord Chesterfield that “dirt is matter out of place”, my plan was to bring them both under the same cognitive rubric. Assuming that all humans have the same negative reaction to dirt thus defined, we should expect rational beings to experience the same internal pressures to prefer regularity and to reject disorder—disorder being a more abstract case of “matter out of place.”
After publication I had the good fortune to meet Basil Bernstein, a fellow anthropologist who shared my admiration for Émile Durkheim. I was disappointed, however, to find that Bernstein rejected my central argument in Purity and Danger. He protested vigorously against my tilt at universalism. Any theory of dirt and pollution, he argued, must allow for different reactions. Before vaulting from dirt to religious defilement, from the literal to the more abstract, I should try to account for the variation in reactions to contact with snails, slime, feces, vomit and entrails. He pointed out that some people relish eating the very things which fill others with disgust. We are revolted by the idea of eating human flesh, but we have to admit that cannibals like it. What about the artist passionately concentrating on his painting, so absorbed in his work that he relieves himself in the nearby studio sink, where dirty coffee mugs are standing around? His surroundings are chaotic; general dirt and disorder don’t worry him at all; the only spaces in which he cares for cleanness are his palette and his canvas. Having actually met people like this, it was hard to disagree.
In these terms Bernstein persuaded me to differentiate contexts and cultures, strong and weak systems of classification, complex and simple ones. I needed a typology of cultures before I could get back to dirt, religion and the like—though I like to think that Purity and Danger stands up after forty years as a useful conceptual sketch of the problem. As it happened, however, there was no scheme ready-made suited to my purpose, so I had to make one up. In a 1973 book called Natural Symbols, I produced a rudimentary typology called “Grid and Group.” It was designed to trace the distribution of values in any given population.
My scheme followed Bernstein’s research on two types of English families, the first labelled “positional control”, the second “personal control.”1 The former is organized on a system of positions where the family control system depends on classifications based on age, gender and time-tabling conventions (time to rise in the morning, lunch time, bed time and so forth). The space of the house is divided according to the same regime. In the kitchen the mother is the supreme authority. Each person has an assigned place at table. The children may not negotiate their bedtimes: The youngest goes off first, the second youngest next, and so on with the eldest granted the privilege of going last. At meals, family seating is based on the same principles: No one can start to eat until all the family is seated in their rightful places, and what they eat is determined by the time of day and the calendar.2 The same goes for the distribution of housework. The eldest has the most responsibility, the boys do the heavy and the dirty work, the girls do the beds, the toilets, the laundry. The positional family keeps a tidy house, assigning available space on functional principles: dining room for eating, bedrooms for sleeping, lavatories and bathrooms for private bodily functions. These principles are honored even in the wall decorations: no pictures in the lavatory, no books in the dining room. Worked out consistently, the system establishes coherence among different spheres; it makes sense.
Not so in the house of the personal family. Under the regime of personal control there is no need for a dining room; we might eat anywhere. Some space is assigned to individual members insofar as they each have their own corner where they can do as they like. Instead of a general set of principles administered from on high, the children have the right to challenge any command. The parents encourage dialogue and the child is trained to be sensitive to the emotions of others. This is the educational style of the middle classes, who are concerned to keep or to improve their place in the social class system. The child grows up in an environment without fixed boundaries, but attains a vocabulary for expressing his feelings and for anticipating the reactions of other people.
Grid and Group
These two types of family control systems work as miniature models of two types of culture—positional and individualist—as shown on the Grid/Group table (see figure 1 on next page). The table shows two dimensions, representing two types of control. One is exerted for and by the group, involving personal control exercised by members over each other. The other describes a variety of controls that do not directly stem from or support the group, but are instead collective responses to climate, technology, work—anything that underpins the web of institutions.
This method of organizing social research exposes the machinery of cultural transmission. Sets of values and expectations are transferred along the lines of the social structure. The horizontal axis represents the strength of group pressures on individuals, from weak to strong. Somewhere near the zero point where “group” is a weak principle of organization, one may be counted as a member of a parish, for example, by simply making an appearance at one of its annual ceremonies. At the other extreme, at the point of maximum integration, we would expect to find groups that demand total commitment for life, as with a monastery.
In some social environments there are no groups, strictly speaking, but there are always conventions that control behavior. The force of these conventions is to be measured on the vertical axis, called “grid” or structure. It registers the amount of regulation tolerated, the rules of the road, for example: safety laws for houses; hygiene rules for fishmongers, pubs and cafes; law of trespass; standards of decency; conventions of etiquette. These regulations apply to everybody, without privilege or exemption, regardless of group membership.
Put together on the one zero point, we can use these same two independent dimensions to give four types of culture: positional, individualist, isolates and enclave (see figure 2 on next page). Let us make the rounds of the diagram, describing each of the four cultural types. (Note first, please, that no one insists that any two societies are identical, only that every society may be placed somewhere on the grid, making it possible to think about them in relation to one another.)
Positional: At top right, the two dimensions of the diagram arrive at their maximum: strong group with strong regulation. In positional cultures all roles will be predetermined, all behavior subject to positional rules indicated by heredity or gender or age, and by various combinations of all three. Little groups organized in this way—families, for example—may be incorporated by larger groups similarly organized. There may be several levels of groups included in a large hierarchy. Just for this to be possible, the positional culture must favor tradition and continuity and frown on competition, except with outsiders. It must encourage respect, loyalty, obedience and the well-being of the community at the expense of any individual or collection of non-associated individuals.
Positional culture subordinates the individual to the good of the whole. With its coherent structure of subordination and command, it is extremely efficient for coordination. It cannot allow individual competition to disrupt the calm repetitive cycle of the generations. It affords certainty over a large range of questions, inspires confidence and trust. Decision-making roles are clearly located at the top, and support is readily mustered. This kind of culture is capable of taking the long term into account. Compared with other cultures, its greatest advantage may well be its strength in suppressing jealousy.
It is not a culture, however, keen on equal rights. In many positional societies, justice must bend often to status. The “Punishment Matrix” of the Muluki Ain legal code of Nepal (1854), for example, concerns illegal sexual intercourse. If the male offender was of high caste and his female victim was of low caste, the fine he paid was almost negligible, but the rate went up with the caste status of the woman. If he had assaulted or seduced a women of his own caste, the penalty was very high indeed.
Individualist: The individualist cultural type corresponds to the personal family. The child is trained to stand up for him or herself, to speak up and to challenge. Leaving aside the individual and going to the corresponding social environment, individualism is in the bottom-left quadrant. It is a competitive culture. The well-being of the community does not come before or above the well-being of the individual. The admired virtues are individual courage, intelligence, perseverance and success. Power and wealth are the rewards. At the extreme point of the diagram, bottom left, the contrast with the positional culture is total. It is a tough environment; competition can be merciless; the weakest go to the wall.
Anyone who tries to assess a given community’s culture according to this scheme will quickly become aware of the pitfalls. There is the danger of distortion due to personal bias. The problem of objectivity arises. Aaron Wildavsky, the most eminent leader of work on Culture Theory, prided himself on belonging to the culture of Individualism. For him it was par excellence the culture of America, the culture of the pioneers who opened the American West, the entrepreneurs who led industry and science to great heights, who developed modern technology. But to hold to his view of the early American colonists as freedom-loving individualists he had to overlook their widespread tolerance of slavery. If one culture is strongly preferred, that preference can cast other cultures into a grey area of low esteem. Equal rights in theory can coexist with vast disparities in outcome in practice.
Isolates: We are now at the top-left corner of the diagram, where the role of the group is at a minimum and that of regulation is at a maximum. Everyone in this situation is an isolate, separated from others not by natural barriers to communication, but by the rules and regulations controlling social relations. Many urban dwellers are likely to have this experience. Conventions prevent them from joining many groups, perhaps because they lack qualifications or schooling, are the wrong color, have the wrong accent or not enough money. At the extreme tip of this quadrant we find those unable to meet the conditions for belonging to viable groups in a positional society: Perhaps the isolates are deviants, refugees or immigrants. Or they may have lost patronage and been pushed out of the culture of competitive individualism. This corner of the diagram may often be constituted by rejects from other cultures in their community.
We may well doubt whether it makes sense to speak of a culture of isolates, since culture is a collective product. If they can’t get together, how can they make a culture of their own? The answer is that in this case we are talking about a shared experience to which persons respond by developing a common set of ideas. Perhaps it would be less confusing to say that isolates share a common philosophy: fatalism. Being alone they have little or no influence and no close friends. No one has reason to consult them because their support is hardly worth having. They don’t have anyone much to talk to; conversation is limited; ideas get simplified. Conspiracy often becomes the favorite explanation for what’s wrong with the world, and the anonymous “they” is pervasive. The isolate perceives injustice and accepts privation, but thinks there is nothing to be done about it.
Enclave: Finally we come to the quadrant on the bottom right: a culture with strong groups and weak structure, giving rise to a society with strongly barred frontiers and feeble internal regulation. The purest examples of enclave cultures would be the many varieties of utopian communities that flourished in America in the 19th century, the communes of the 1970s, or sectarians such as the original Seventh Day Adventists and the Plymouth Brethren. These cultures had strict rules regulating contact with the outside, but inside the group they avoided social differentiation. Aaron Wildavsky used to draw on Walter Scott’s novel Old Mortality to imagine such a community. The novel depicts paramilitary Covenanters (Calvinists) embroiled in the religious conflicts of 18th-century Scotland. Fiercely intolerant, hard on each other and pitiless toward enemies, Scott’s descriptions give such a sinister twist to the word “sect” that it would place the Covenanters at the extreme right point of the quadrant.
This is the harsh model from which our ideas about enclave culture were first based; later fieldwork has encouraged a broader view, but as I saw it then, a sect would be a group of persons who dissented from the way of life in the dominant society and had withdrawn to live together according to their chosen principles. Sectarian thinking tends to draw the world in black and white, outsiders and insiders. To account for this dichotomous way of thinking, I reasoned that, being a dissenting minority, this group would have difficulty in preserving itself. It would be threatened from without by the society it regards as corrupted by wealth and power, and it would be threatened from within by disaffected members, making its political life insecure. Enclave group leaders would constantly fear defection, the best remedy to which was to paint non-members as thoroughly evil—a characteristic proved whenever their relations with the outside turned violent. (Field research among extreme Maoist groups in London confirmed these expectations and enriched our understanding of sectarian behavior. The leadership exploited persecution from outside by organizing disruptive protests; confrontations with the authorities then invariably rallied group commitment.3 There may be lessons here for how to deal, and how not to deal, with terrorist groups.)
Structure within enclave cultures legitimates the division of labor, economic and political. Without structure, jealousy would be rampant. One of the disadvantages of a group conforming to the egalitarian ideal is the difficulty of establishing authority and the consequent weakness of leadership. Anyone who tries to exercise influence would soon be accused of free-loading or of trying to split the group. A common solution to the sect’s leadership problem is to choose a charismatic person who can claim to be getting counsel from God direct. Another is to put decisions to aleatory processes. In any event, the immense difficulty of trying to live together without structure often motivates leaders to strengthen the barricades against the outside. This large drain on their attention and resources makes for inefficiency, however, leaving enclave cultures able to make sharp, short forays against its enemies, but unable to expand their areas of influence and bring a larger part of the population under control. Enclave cultures are good at disruption, bad at administration.
Practice and Theory
To use this quadripartite scheme for empirical research, one must choose for study a clearly defined and reasonably stable “world.” Some early examples undertaken by English scholars included 19th-century mathematics departments in German universities, longshoremen in Nova Scotia, and, as noted, an extreme Maoist group in London. Others subsequently used Culture Theory to study the perception of mental handicap in Brittany, for example.4 As noted at the outset, CT is not only about theory and academic research. Some studies have had practical utility in dealing with organized crime, harmful stigma concerning AIDS victims and their medical treatment, how to understand how groups assess risk, and more besides. These benefits have become possible because there is now a way to make meaningful comparisons among different cultural phenomena.
But doing so is not always easy or simple. At the beginning of our combined research using the Grid and Group method I already had a negative view of enclave cultures. Wildavsky did too, feeling disdain for the activists of what Lionel Trilling called the adversary culture, who moved their protests from the Vietnam War to capitalist spoliation of the environment, nuclear power plants and plenty more besides. This was awkward because we were vaunting CT as an instrument that could correct or control for investigators’ mostly unwitting cultural bias. Our negative attitude toward enclave cultures was evident in our Risk and Culture, which cost us some hostile reviews from some political activists and those who sympathized with them.
At that time, happily, every new piece of Grid and Group research brought clarification and theoretical growth. From the start, the scheme was a powerful discriminator of cultural bias, but it was static. It held no normative messages, and it was offered only as a research tool, not as a theory. That changed when Thompson and Wildavsky co-authored Cultural Theory in 1990. The convergence of their interests and talents transformed Grid and Group analysis from a modest method to a new theory founded on three assumptions. First, at the level of social organization, every kind of society was deemed to comprise all four cultures, even at the level of families and especially at the national level. Second, at the cultural level, each of the four cultures is defined by contrast with and in opposition to the others, making for a relativist conceptual scheme, not a fixed and strictly empirical one. Third, relations between cultures within a given society are inherently conflictual—which is the source of cultural dynamics.
From these assumptions, in turn, came the normative principle of mutual honor among cultures within a society, for conflict ensures variety. Though one culture may predominate over others, it will endanger the well-being of the whole society if any one of the four cultures oppresses or eliminates the others. That is because each culture serves a necessary function. A well-run community needs some positional hierarchy in the sphere of government, some enterprise on the part of individualists, some criticism from enclaves, and cannot avoid having some passive isolates. If the positional culture dominates, it will stifle necessary adaptation and make things hard for those in the lowest positions. If the individualistic culture dominates, ruthless competition will cause the weak and isolates to suffer. If the enclave culture suffers too much, moral censorship will calcify the larger social scene, and if others combine to suppress the enclave, violence will erupt.
Here, perhaps, lies a normative lesson for our times, which is that a war on terror cannot be won unless the enclave’s claim of injustice is calmed. Force and policing will be unavailing unless the perception of injustice that fuels subversive movements is addressed, and that is so regardless of where or how that perception has arisen. When U.S. or British policymakers perceive problems with terror cells abroad, it means, in Grid and Group terms, that positional cultures in one society are trying to deal with enclave cultures in other societies. From what we have already said, that observation alone casts a whole new light over what is being attempted, does it not?
This conclusion is strengthened and its possible applications expanded by some recent studies of Oriental enclaves. From his field research on Arab and Israeli fundamentalist groups, Emmanuel Sivan has given us a new slant on enclave movements.5 The main difference among fundamentalist religious sects stems from their position in mainstream society—that is, the general relation of a culture within society to the society as a whole. Unlike many contemporary fundamentalist Christian movements, which are intrinsically hostile to the surrounding society and to the established Church, Islamic and Jewish fundamentalists are not necessarily in disaccord with their fellow believers, only less lukewarm in their enthusiasms. In the contemporary Middle East, Jews are reacting to threatening Arabs and Arabs to persecuting Jews—outsiders in each case—but Jewish and Arab enclavists are not strongly withdrawn from the larger society surrounding them. Most enjoy support—some or a great deal—from public opinion, and some may even be subsidized by the state.6 Their enclave politics are not in principle discordant with government policies, but they act independently.
Instead of these groups being withdrawn from the established society around them, it is rather their governments are threatened and insecure amid the larger international scene. This makes a big difference to how the enclave culture within any society is organized and how it behaves. They would be more tolerant of ranking and authority, given their quasi-military activities. If they therefore lean more toward the positional culture in ranking, we would place them further toward the middle of the diagram, not on its extreme egalitarian edge.
That would accord with Shaul Mishal’s and Maoz Rosenthal’s typology of Arab terrorists.7 They have defined a new type of political organization which they call “dune organization” on account of its extreme fluidity. It is a form of enclavism that enjoys the support of the dominant society, so there is no need for a big wall for enclavists to separate themselves from others. So far from being hostile toward groups with similar objectives, they maintain fraternal relations and give and receive reciprocal aid. Thus the objectives of Hizballah, Hamas and the Palestinian Islamic Jihad are to liberate from the control of Israel all supposedly Palestinian territories and to eliminate Western influences in Lebanon. The objectives of al-Qaeda, more global and transnational, are to overturn the Arab chiefs who spite sharia law, destabilize the West, and liberate the Islamic world from its domination. Enclave culture is best, remember, at destabilizing and liberating, not at encompassing and administering.
The new technology of communication works in favor of the political enclave and the dune organization of its institutions. An organization with aggressive projects against its enemies needs an effective system of communication with affiliate groups. Thanks to new electronic techniques, it can now contact its allies, ask for immediate military support or for armaments. Several such groups can combine for a well-coordinated attack, after which they can disperse and disappear without trace. The new enclave is an ineffective form of organization for most purposes, but for certain objectives it is admirably well-adapted. All it needs is the steady commitment of its members, and that is secured by the glory of the action itself. Grid and Group, properly applied, can show how such enclaves interact with the rest of their society, can relate how the social position of such enclaves differs from society to society, and can suggest how societies with different mixes of cultures will react when encountering each other.
There are many other examples of CT being applied to political and international issues: Nepalese water policy, civil war in Sierra Leone, the biases of English and American jury systems, even aspects of global warming. It has also inspired reflective theses on human society, not least Benjamin Davies’ treatise, Essential Injustice (1997), on how the concept of justice changes in harmony with the dominant culture. We are justified, I think, in believing that CT is here to stay, and that eventually cultural influences will be automatically included in political and social analyses. One may hope so, for some very ugly consequences have ensued from ignoring them.
We are also justified, I think, in believing that the analysis of human cultures can be conducted systematically, that reliable comparisons can be made, that observations from fieldwork are worth the trouble it takes to go out and make them. Collaboration with colleagues within and among disciplines can, with any luck, lead to better research methods and better theories guiding and advancing one another. I think the history of CT is proof of this.
Finally, there is also something to be said for innocent obsessions, for sheer curiosity. If my interest in dirt led to Cultural Theory, who can say what may be learned in future starting from other curious places? ?
Bernstein, Class, Codes and Control, vol. I: Theoretical Studies Towards a Sociology of Language (Routledge, 1971).
See Mary Douglas, ed., Food in the Social Order (Russell Sage Foundation, 1984).
Steve Rayner, “The Perception of Time and Space in Egalitarian Sects: A Millenarian Cosmology”, ed. Mary Douglas, Essays in the Sociology of Perception (Routledge, 1982).
Full citations for these and other applications of CT are available upon request.
Sivan, “The Enclave Culture”, Fundamentalism Comprehended, Martin Marty, ed. (University of Chicago Press, 1995).
Gabriel Almond and Emmanuel Sivan, Strong Religion: The Rise of Fundamentalisms Around the World (University of Chicago Press, 2003).
Mishal and Rosenthal, “Al Kaeda as a Dune Organization: Towards a Typology of Islamic Terrorist Organizations”, unpublished paper from the Department of Political Science, Tel Aviv University, July 2004.