Working knowledge

American college students traditionally had a cultural model of a “date” (Holland and Quinn 1987; Kronenfeld 2008a, 2008b; “dating” now sounds old-fashioned, however). This model involves two people who have, or think they might have, a romantic interest in each other. They go out for “dinner and a movie” or possibly something “more creative.” During this, they probe with varying degrees of subtlety to see how much physical contact they will have and how much they like each other. There is a canonical date, and there are recognized variants of it (zoo, walk in the park…). There are thoroughly non-canonical variants still recognizably “dates” (parachuting together, stealing cars together…). There are things that are only marginally “dates,” testing and probing the boundaries of the model (group outings, set-up matchmaking dinners…). There are well-known subroutines with their own cultural rules (picking the girl up, paying for her dinner or not, kissing goodnight).

Again, cultural models give a plan; people vary it at need. Different people have slightly different ideas about dates. The models change over time, too. Young people today expect far more physicality, far earlier in the game, than people of my generation did. On the other hand, the young now are far less hopeful of permanent commitment resulting from steady dating. This represents a major cultural change, but it is not total; some of the young have “old-fashioned” attitudes, and more than a few of my generation were, shall we say, ahead of their time.

What matters is that (1) everybody knows the general cultural expectations, and (2) everybody accommodates to them at least a little bit, and usually more than a little.

Claudia Strauss (2007) has used the concept of cultural models to discuss the ways Americans assign blame for the shootings at Columbine High School, Colorado, in 1999, in which two disturbed students shot down 12 classmates, a teacher, and then themselves. Americans in general have a default tendency to look to the individual. They tend to idealize individualism, self-reliance, and independence. Thus, they explain much by independent individual agency. However, Strauss found that Americans more often explained the Columbine shootings by social factors. Many of these were suggested by her interviewees, and she found them and others in the popular media as well. Usually, each source had a particular factor to blame, but some blamed two or three social trends. Among those suggested were bullying and abuse of the two students; easy availability of guns; lack of easy availability of guns (for the students and teachers to defend themselves); glorification of violence in the mass media; breakdown of morality (Biblical, national, or otherwise); various forms of school failure, from lack of values to lack of supervision; parental failure; and poor health care (including lack of psychotherapy). Evidently, there is no shortage of cultural models blaming the system. Strauss found that the few who blamed solely the will and agency of the two students were usually quite defensive about it. They were aware that they were in a minority.

Obviously, the various explanatory models were not constructed in a vacuum. The few who blamed the students (alone) were self-conscious libertarians. Traditionally devout people blamed the breakdown of Biblical morality. Political liberals blamed lax gun laws; staunch conservatives blamed too-tight ones. Sympathetic people caring about the students’ psychological problems blamed psychological factors. Teenagers seemed very commonly to blame bullying and bad treatment of the pair. Critics of American education blamed the school.

In this case we are dealing with a definition of “cultural model” that is more or less Arthur Kleinman’s “explanatory model,” which he invoked in describing the ways that Chinese talk about illness and depression (Kleinman 1980). The explanatory model has become widely enough known in medical circles to be reduced to “EM” in the literature. It has proved an extremely valuable concept. Since Kleinman’s original work, which defined EM’s in narrow and cognitive terms, Kleinman and others have increasingly grounded EM’s in emotion and in wider cultural knowledge (see e.g. Garro 1988; Kleinman 1988). We now see them as specific representations derived from a much wider base—little peaks in an adaptive mental landscape. Also relevant is Alison Wylie’s philosophy of the standpoint (Wylie 2004). Where one stands, or sits, in this world explains much about how one explains. It also explains why we often miss things that would be obvious to an observer from another culture. As an English variant of a Chinese Taoist saying has it, “a way of looking is also a way of not looking.”

Other cultural knowledge is less obviously “modeled.” “Normal” Americans have a great deal of knowledge of “celebs,” largely media personalities. Every “real” American knows exactly who is meant by the first names and nicknames on the covers of People, In Style, and similar inhabitants of supermarket checkout zones. I usually have no idea who is meant, and spend checkout time wondering what I am supposed to do with the information that Jack is breaking up with Jill and seeing Jane.

This knowledge is clearly systematic, structured, and culturally shared (by everyone but me). For some people, it is a “model,” because they pattern their lives after the “celebs,” or, perhaps more often, use the “celebs” as cautionary models showing what to avoid. Yet, for others, it is not a “model,” because it does not inspire them to do anything. They treat it as passive, list-like knowledge—declarative rather than procedural.

Normally, a full “cultural model” or “schema” would include all the cognitive tasks noted above. Driving a car, for instance, includes the actual how-to rules, the laws of the road, the bodily motions, the mental representations of the car and its steering system, the mental map of where one is going, moral standards for judging other drivers, and a sense of rhythm (try to shift gears and manage the brake without that!). Above all, it requires constant monitoring of one’s performance and the car’s, and constant adjustment. The beginning driver is acutely conscious of every slight correction of the steering wheel, accelerator or brake pressure, and so on. The veteran does it unconsciously, unless confronted (often at a cut-rate rental joint) with a new car possessed of strange ways.

Cultural models have a strange shadowy existence. Grammar rules are put down in books, and sometimes dating rules are too, but the books never give exhaustive directions. We understand “I ain’t got none,” though it follows a grammar found in few books. We also know that “the little red corner house” is fine, but “*the corner red little house” is improper. (The asterisk is a linguists’ convention to show that a form does not really exist.) Nobody ever taught us the rules for adjective order in colloquial English. We inferred them without even realizing it.

This shows that cultural models do not exist as objectively real things that hang in space somewhere. They have a weird life, being inferred by everybody but typically codified by nobody (Kronenfeld 2008a, 2008b). Grammar rules are learned by young children through trial and error and inference. Dating rules are learned by adolescents the same way. They get instruction from peers or websites, they imitate their slightly older friends, and they make countless acutely embarrassing but highly instructive errors. Even recipes are learned more often by imitative practice than by reading, in spite of all those cookbooks. Languages and cultures exist solely in this strange interactive space—unreal and nonexistent, yet totally real and totally powerful.
Kinship systems are extremely structured and are known to most competent adults. At the other end are things like our cultural theories of mind or religion. They are vague and general. Broad principles like individualism vs. collectivism also exist at a vaguer and more passive level.

However, cultural models determine actual practice of religion and so on. An American church congregation knows to stand up, sit down, and kneel at the right times, though their theological concepts, religious beliefs, and interpretations of church ethical rules may differ widely from person to person. There will usually be a Wittgensteinian “family resemblance” in religious views, but not always. A newly immigrant, highly conservative African Episcopalian family may share very little indeed with the affluent liberal Anglo-American family in the next pew.

This shows why cultural models are important. Knowing the general principles of Christianity, even memorizing the whole New Testament, will not help you one bit in church if you have no idea of the ceremony. Having a desperate desire to do what’s right, to fit in, and to maximize your social benefits will not help either. You have to know the cultural model of a proper Episcopalian service. If you do, and you use it, nobody will care deeply what your theology or motivation are. Sects differ in this; Calvary Chapel pays more attention to the belief, less to the ceremony. But there are always cultural models mediating between the very general (belief) and the very specific (your own feelings this minute in the back pew).

In farming, driving, teaching school, peeling potatoes, anything, the same is true. Very general principles do not guide action. Very specific individual motives, goals and wishes do not coordinate action. Culture exists to tell people what to do when they want to accomplish personal goals in the light of general principles. It occupies the middle position—providing the scripts, canonical plans, and strategies that allow one to get from broad goals to actual performance.

The fuzziness of the concept of “cultural model” affects how we think about sharing and agency in relation to these models. The idea originally came from grammar and other tightly rule-bound and highly overlearned systems. Thus, there is something of an idea in modeling circles that cultural models are subconscious, fairly rigid, and highly determinative of behavior. This is certainly not true of, say, song-writing. Music is structured linear communicative behavior, like language, and musical pieces are just as rule-bound and rule-governed as linguistic utterances. But song-writers and musicians are intensely conscious of the rules, and apply them diligently, as any composition textbook or course demonstrates. Jazz and folk musicians learn the rules of “improvisation” to the point of subconscious deployment, but generally they can bring the rules to consciousness if they have to explain them to a beginner or break them in a wild riff.

Grammar rules are fairly inflexible, while music rules change fairly often. Rules for clothing fashions change even more often and even more consciously. Roland Barthes (1983) did a memorable, and refreshingly light-hearted, study of the tight structure and communicative function of those rules. Cultural models of things like driving and housebuilding have to change, consciously, every year, as lawmakers and architects invent new codes. And there are areas in which cultural models are extremely vague indeed. Kronenfeld has often contrasted formal baseball or soccer—tightly scripted, with formal and informal rules—with the sandlot versions, which in their young-boy forms may have almost no rules at all. At this point, it is questionable whether one has a “cultural model” or simply a name for a somewhat arbitrary cut of everyday activity. The degree of systematization in models and knowledge structures is extremely variable. It follows that only rather systematized ones are useful in predicting behavior. The more tight the rules, and the more they constitute an actual system with all rules linked and mutually supporting, the more prediction is possible.

Cultural knowledge is to use. At any given time, we naturally are thinking only of what we need for the task we are focused on, but this does not mean that the rest of our knowledge is permanently subconscious! We can foreground only one body of lore at a time. Quoting Hedrick again: “Even how we move is not very difficult to bring to conscious awareness if one takes dance, horseback riding, or martial arts. For that matter, how we breathe can be easily brought to conscious awareness. Arguably, that is the entire point of religions like Buddhism—that you bring into your conscious awareness the models that construct who you think you are, and thereby gain freedom from them by systematically severing your attachments to them. Only humans can do this. It makes us uniquely us- our capacity to bring to awareness the most biological, natural components of ourselves on through to cultural conditioning... and then choose otherwise” (Hedrick, email of Nov. 28, 2008).
Consider how the idea of cultural models can play into understanding our modern ecological crisis in agriculture. Most analyses of agricultural problems today blame either individual short-term income maximization or big, vague things like “capitalism” or “Judeo-Christian ideology” or self-interested assertion of power by state bureaucrats (on the last, see Scott 1998). Indeed all such things matter. However, the real problems of “developed” agriculture today are monocropping, replacement of small farms by giant estates (usually owned by absentee or state landlords and worked by landless labor), erosion of genetic diversity, and total destruction of natural landscapes to make way for those three things. Detailed knowledge is not highly valued, since the owner is more concerned with urban politics or whatever he may be doing, while the workers have no time or capability to use their knowledge.

This is a pattern set by ancient Babylonian times, and perfected by the Romans. It is not new. It made some economic and ecological sense in very ancient times, in the river valleys of Egypt and Mesopotamia. Today it is highly countereconomic and antienvironmental, but it has become established among developers and international agencies as the right way to do agriculture. It thus propagates worldwide, at enormous cost to everyone.

Very different principles guide the Maya agriculture I study in Yucatan. It involves holistic management of the entire landscape, preserving as many species as possible and finding uses for all of them. It involves multicropping, with fine-tuned mixes of species. It involves small farms that give a modest living to everyone. It is skill-intensive rather than capital-intensive, so farmers have to care about learning and knowing as much as possible.

Such broad management strategies and plans are true cultural models. They guide action without specifying exactly how to act in a particular situation. They are the means of moving from very broad considerations—political power, economic gain—to very narrow tactics (growing wheat in this particular field, irrigating from this particular well). They are the specific strategies that get us from very broad cultural principles—ideology, religion, social solidarity—to very specific individual considerations. They are the models that guide the key decisions. Without them we could not possibly coordinate action, set policy, or develop anything.

Ultimately, they were developed because they made some sort of economic and political sense, so power and economic gain are at some very distant remove the real causal factors. But there has been so much history since, so much contingency and accident, so much blind following of models and then improvising a cure after failure, that very little can be predicted about modern agriculture from looking at what is “best” in some abstract sense. The system has created a “lock-in”; we are dependent on it and have tailored our infrastructure—from shipping to food processing—to accommodate it. If we were starting over, we would surely do things differently, but models that developed in ancient Mesopotamia still lie behind our strategies today.

The contingency of long-lasting cultural models is best shown, once again, by bread. A purely accidental hybrid of wheat with the Azerbaijan subspecies of the grass Aegilops squarrosus produced, around 6000 BC, bread that would actually rise and become wonderfully fluffy and soft. Previously, bread had been fairly heavy, tough stuff. Inspired women (it must have been the women) took advantage of this, and the end result is a world of cakes, cookies, breads, shortcakes, pastries, and so on. Bread wheat is the most important and widespread crop in the world. All this depends on that crazy hybridization event, but now we are locked in; we cannot imagine life without it, could not find a substitute if we had to, and would have to change our whole dietary world profoundly if bread wheat ceased to exist. That event is not a totally unlikely possibility. A strain of wheat rust arose in Africa a couple of years ago that threatened quite seriously to end bread wheat. We live in a world of chance, and cultural models sometimes fit us for that, sometimes lock us into a dangerous dependency.

Cultural Heuristics

Cultural heuristics, notions, and metaphors form a subset of cultural models. Some are mere fads that die in a year, while others go on for centuries. No one really understands this; it remains a question for research. Many of the more elaborate and important of such notions are metaphors (Lakoff and Johnson 1980). George Lakoff, who explored the whole issue of metaphors, has applied his thinking to politics and political campaigns, in such matters as the emotions aroused by various definitions and images of “freedom” (Lakoff 2006). Other heuristics are frames (Bateson 1972): we frame a situation in a particular way and interpret it accordingly.

An example of how these heuristics play on the ground is provided by Kenneth Dodge in an analysis of youth violence (2008). The popular media have portrayed violence in American youth as due to a few incorrigible “superpredators,” or to decline of old-time religion, or to amoral rational calculus, or to basic human evil, or to other metaphors that emphasize innate cussedness. All these lead to incarceration, or boot camps and special schools that inhibit rather than teach. Dodge advocates trying public health metaphors instead: Youth crime is like smoking, like chronic disease management, or, best of all, like heart disease. The risk factors for heart disease are overwhelmingly controllable: smoking, obesity, saturated fat consumption, and so on. Youths in the inner city get plenty of exposure to the worst possible conditions, models, metaphors, and so on, but little opportunity either to learn other lessons or to get anywhere by acting on them. If the only way to survive, let alone prosper, is to sell drugs, kids will do it. The best cure, Dodge finds on the basis of evidence, is early and ongoing moral teaching. Obviously, however, social and economic justice will in the long run be necessary to any cure.
Cultural Models and Agency

Cultural models are abstractions. Individuals must form them and act on the basis of them.

This is most clearly shown by the tendency of people to get them wrong, especially at first. Learners of a language make hash of the grammar. Young teenagers make “fashion statements” that bring tears to the eyes of the more sophisticated. Fledgling politicians learn the hard way what they can and can’t criticize.

All this shapes the models. Normally, cultural ways are stable, and the young and green are ridiculed or ostracized until they conform. However, all models change over time, and this is precisely because they exist only as inferences within individuals’ heads. If enough people make a grammatical error, it becomes entrenched and becomes the new norm; the subjunctive mood has essentially disappeared from English within my own lifetime. Fashions obviously change must faster than grammar, and very often it is teenagers’ outrageous fads that do the changing. Politics changes even faster, because people consciously debate and negotiate the models, as pointed out by George Lakoff in his brilliant cognitive-linguistic riff, Whose Freedom? The Battle over America’s Most Important Idea (2006).

Appropriate to this level of study is the class of models of behavior and knowledge that has recently given us actor-network theory and agent-based modeling. Basically, social science has recognized the fact that society and culture are emergent phenomena produced by the actions of individuals, rather than being some kind of essence or “thing” or frozen crystalline crust on humanity. Current models of human knowledge and action thus tend to begin with actors who occupy spaces in networks (Latour 2005, but the point has been well made ever since Dilthey’s work in the 19^th century; Dilthey 1985). These individuals interact with each other, producing all manner of strange and complicated transactions and event sequences. Eventually, more and more frequent interaction with some people produces a group in which social rules and institutions come into being, and one has a society. That society’s members find they need to share a great deal of knowledge, and culture is born.

This view of culture allows us to understand how it changes—individuals learn and share their learning—and why it tends to be distributed cognition, with different experts having different knowledge pools within the same community.

Agent-based modeling involves computer simulation of individuals. These virtual individuals are programmed to act as much as possible like the real thing. They are then allowed to interact over time. A city’s traffic for a year can be modeled on a computer in an afternoon. The virtual individuals know traffic rules, slow down for accidents, talk illegally on cellphones, and so forth, and the resulting mess can give clues to real-world traffic managers. One has to build in realistic assumptions, though; I heard anecdotally of one simulation that failed because the planners did not take into account the tendency of people on the freeway to slow down to stare at an accident in the opposite lane of traffic. Not long after hearing this story, I actually saw an accident caused by this. Two looky-loos got so carried away slowing to stare at an accident in the opposite lane that one rear-ended the other.
Ideology?

The concept of cultural models gives us a better fix on the idea of ideology. In general, that word refers to the public beliefs espoused by politicians and ordinary people. For Marxists, it is propagated by the elite; for others, it is generally shared. In any case, it may consist of hollow words with no meaningful content, as when Iran boasts of being a democracy. Or it may express a sincere commitment that may greatly overstate reality, as in America’s commitment to freedom, individualism, civil rights, and other matters that are often compromised in practice. But, in general, it expresses strategies for managing people and situations—cultural models for dealing with political problems.

As such, it can be a collection of genuine rules for society and social interaction. Social codes, from the civil law to the basic tenets of most moral systems, are of this sort. Alternatively, it can be more like Aristotle’s idea of ethics: virtues that an individual should cultivate (Aristotle 1955). Here belong individualism, self-reliance, loyalty, and the like, which have social effects but are by nature individual traits. Many rules can be taken both ways: “Thou shalt not steal” is a social rule (property is secure; out-of-code taking of property is forbidden) but also implies that honesty is an individual virtue.

Ideology, then, becomes the public projection or representation of cultural models, belief systems, and social institutions.