International Journal of Qualitative Methods 5 (2) June 2006

Printable PDF Version

Culture: Can You Take It Anywhere?

Invited Lecture Presented at the Gevirtz Graduate School of Education,
University of California at Santa Barbara

Michael Agar


Michael Agar, Professor Emeritus at the University of Maryland and Adjunct Professor at IIQM who works independently as Ethknoworks ( near Santa Fe, New Mexico

Abstract: This is a written version of a lecture delivered at the Gevirtz Graduate School of Eduction at the University of California, Santa Barbara. The original series of lectures can be viewed at In this “culture“ lecture, the author explores the difficulties in using the concept in contemporary research with particular reference to its relational nature, its partiality, and the various problems that ethnography now addresses. In the end, the concept might have outlived its technical use, even as popular use increases.

Keywords: ethnography, culture, history of anthropology, poststructural research

Author’s note
The lecture on which this article is based was made possible by Judith Green and Janet Chrispeels at the Gevirtz Graduate School of Education at the University of California, Santa Barbara, with the indispensable and generous help of Audra Skukauskaite. Many colleagues who viewed the webcast made comments that helped refine the original.

Agar, M. (2006). Culture: Can you take it anywhere? International Journal of Qualitative Methods, 5(2), Article 11. Retrieved [date] from

In late 2005 and early 2006, colleagues at the Graduate School of Education at the University of California, Santa Barbara, invited me to give a series of four lectures on ethnography, language, culture, and complexity theory. This article is a written version of the third lecture, the one on culture. The writing is an attempt to organize some of the improvised themes while still preserving the informal tone of the event and the conversational style of the presentation.

By way of background, let me summarize a couple of key parts of the ethnography and language lectures that are relevant here. First of all, ethnographic research was described as equivalent to learning a second languaculture. Languaculture is a concept that I developed in a book called Language Shock (1995) to remind readers that actually using a language involves all manner of background knowledge and local information in addition to grammar and vocabulary. Ethnography, then, was defined as an encounter between two languacultures, abbreviated LC1 for the native languaculture of the ethnographer and the audience, and LC2 for the languaculture of the studied group.

I also argued that LC2 learning, like ethnography, is driven by rich points, a concept from another book, The Professional Stranger (1996), that I named almost 20 years ago. Rich points are those surprises, those departures from an outsider’s expectations that signal a difference between LC1 and LC2 and give direction to subsequent learning.

Like all the lectures, this one on culture was a response to a specific question that my colleagues at UC Santa Barbara asked me to address. The question was, What about the culture in languaculture? How are we to understand that concept?

The lion’s share of the article will show why the question is so hard to answer by showing how the traditional concept of culture causes headaches in our poststructural world. Slowly but surely, we then move towards an answer to the question. Culture is the ethnographic product, the result that is a translation that links the LC1 and the LC2 that defined the ethnographic encounter in the first place. The simple answer, then, is to let culture label that translation. The solution, unfortunately, unravels when we think about how the culture concept has to change to describe that translation. In the end, it might well turn out that the concept is a vinyl record in an era of digital files.

So, what about the culture in languaculture? How are we to understand the concept?

The original idea behind the culture in languaculture was to remind readers that language users draw on all kinds of things besides grammar and vocabulary—their biography, the nature of the situation they’re in, history, politics—material from pretty much every discipline that’s ever dealt with people.

So my first reaction when UC colleagues posed the question about what culture means was, it might have been a poor choice of words. My own vague, all-inclusive use of the term was the perfect example of the problem they asked me to fix. Goes around, comes around, as they say in the streets.

Culture is one of the most widely (mis)used and contentious concepts in the contemporary vocabulary. And in anthropology, it runs from a desperate clinging to the past to a political critique of the field’s colonial roots to a struggle with a concept that still works but doesn’t really, certainly not like it used to.

Culture is a mess and has been for awhile. A decade or so ago, I attended an interdisciplinary pragmatics meeting in Antwerp. So did John Gumperz, an old teacher of mine who is now an eminence grise at the very university where I gave this talk. At that meeting, John and I stood in the hall and chatted. We were interrupted frequently by Europeans who wanted to ask questions of a founder of sociolinguistics.

One colleague asked him, “How do you tell which papers are presented by anthropologists?”

“It’s easy,“ said John, “They’re the ones who never say culture.”

John wasn’t the only one among my old teachers who had their reservations. Roy D’Andrade, another teacher from college days, is quoted as saying that studying culture today is like studying snow in the middle of an avalanche.

And a presidential panel at the meetings of the American Anthropological Association discussed the culture concept and described it as colonialist, sexist, racist, classist, and God knows what else. The general conclusion that I remember was, get back to us in a decade or so and maybe we’ll have figured it out.

The culture concept is a mess in anthropology. Where did such a nice concept go wrong? In the old days, we used it to describe, to explain, and to generalize. A person did something, so it was their culture. Why did they do it? Because it was their culture. Who were they? They were members of that culture.

It just doesn’t work like that anymore. It may never have, but we pretended it did.

Consider a thought experiment. Here I am giving this lecture. I’m talking out of several languacultures. Most in the audience will understand most of what I say, though there are complications because of the multisite webcast format and the diversity of listeners on just about every social dimension you can imagine. But what if I

1. went to college in the 1980s rather than the 1960s?

2. were female rather than male?

3. grew up Jewish rather than Catholic?

4. were raised in Mississippi rather than California?

5. were a native speaker of Spanish rather than English?

6. delivered a sales pitch rather than a lecture?

etc., etc.

Would any of these differences have been noticed? Maybe created rich points where none were before? For any of the audience at all? Or some of them? Or most of them?

As I write this, it’s fun to think of my opposite in this list—a Cuban Jewish female 1980s college grad from Mississippi working as a salesperson. I’ll bet she exists. All of the characteristics in the list are markers of what people now routinely call culture—generation, gender, religion, region, language, and occupation, reading down the list from 1 to 6. I leave it to the reader to think of other so-called cultures, to think of how many different ways that the X can be filled in in the phrases X culture and culture of X. Is there any way to justify this use of the culture concept?

There has to be. It’s how the world works now. The problem is, the old concept of culture can’t handle it. The old concept of culture looked like this:

Culture consists of patterns, explicit and implicit, of and for behavior, acquired and transmitted by symbols constituting the distinctive achievement of human groups, including their embodiment in artifacts; the essential core of culture consists of traditional (i.e., historically derived and selected) ideas and especially their attached values; culture systems may, on the one hand, be considered as products of action, on the other as conditioning elements of further action. (p. 357)

The words are from two of the deities of anthropology, Alfred Kroeber and Clyde Kluckhohn, from their jointly authored book Culture, based on a report they did in 1952. Their mission was to look at the ways that culture was used in anthropology and come up with a summary definition.

The definition opens up more questions than it answers. Even more of a headache are the assumptions on which the definition rests, assumptions that are a problem in ways suggested by the thought experiment I just described.

We used to assume that culture was a closed, coherent system of meaning and action in which an individual always and only participated. You were a Navajo or a Yoruba or a Trobriand Islander, period, and all of you were pretty much the same, as far as culture went. You lived in a place that was fairly small and isolated. With the exception of a market town and some nearby communities you were linked to by marriage, your world was pretty much contained in that confined geographical box. And the box was closed in time as well, because your culture was traditional. It stayed the same from one generation to the next. What people did could be described, explained, and generalized by their membership in that single, shared culture.

This is an oversimplification, if not a parody, but not by too much.

Now compare this with the earlier thought experiment, the contrast between the Anglo-American Catholic male 1960s college grad from California delivering a lecture and the Cuban Jewish female 1980s college grad from Mississippi working as a salesperson.

None of those labels have much to do with the Kroeber and Kluckhohn (1952) way of thinking about culture. Nothing closed in space or time about any of them. No claim from any of them to cover all of who you are and everything you do. They represent labels that work inside a society as well as across different ones. In the thought experiment, both people would be members of American culture, about which more in a moment, but they would also be members of different regional, linguistic, religious, generational, gender, etc. cultures.

And those labels could mix and match in ways different from the two examples in the experiment. Just take a walk through any high school and see for yourself. Well, any public high school.

And what each label means might vary from person to person, or for a person from time to time. Take my Catholic label, for example. Somebody, please, take my Catholic label. I grew up in a parish with an old Irish priest, so we got that unenlightened 1950s rural-Irish-gloom-and-doom-and-then-off-to-hell-you-go version. By high school, I though of myself as an ex-Catholic. When I was about 30, I realized I’d never be ex. Nowadays, Mother Church is mostly a source of stories and jokes, except for the days when I feel like a defrocked Jesuit. The way that my religious culture comes and goes and fits or aggravates the flow of the moment changes from year to year, or even from moment to moment.

Each of the labels in the thought experiment tags something we would now call a culture. But they mix and change and vary in ways that make old-time anthropologists—not to mention old-time anthropologies—tear their hair out. And yet the thought experiment examples can all be part of the same national culture, like America. Wait a minute. How can we talk about American culture?

Once again, we can’t, not using the old Kroeber and Kluckhohn (1966) definition. But we can in terms of America as yet another partial culture in the mix, in America or anywhere else in the world, for that matter. America is a state. It has a border that you cross. Once you cross that border, the rules change, rules that have to do with work, health care, housing, education, taxes, media, business, and other domains that, taken together, touch the details of life frequently and in many ways. So the two people in the thought experiment, different as they are on many cultural dimensions, are also members of the same culture.

You see why I considered taking medication when my colleagues asked me to talk about culture?

Let me mention one more of my favorite complications, what I call the no good kids hypothesis. I mean the phrase—no good kids—with affection, since I was one myself.

The thought occurred to me when I was reading the results of an international youth marketing survey in the business section of a Sunday newspaper. The background to the hypothesis: Imagine two teenage kids, 15 or 16,say, chosen at random from any two urban areas in the world. And imagine that they are lower middle class or higher. The hypothesis might work for poor kids as well, but I’m assuming they weren’t included in a marketing survey of consumers.

The no good kids hypothesis then goes like this: For any two kids, drawn at random from anywhere in the world, the odds are they’ll have more in common with each other than either one will with their grandparents.

Who knows if this is right, but it certainly is plausible. And the look of shock on adult faces in Europe and Latin America resemble the ones in the U.S. when I suggest it. The shock no doubt results from the same feeling I had when the international marketing survey suggested the hypothesis. It just might be true.

What does such a hypothesis say about culture if it’s true? Basically that it’s gone in any traditional sense of the term, going fast at the national level and developing at the global level. We need to rethink this concept. Or get rid of it altogether. Maybe its day has come and gone, like the days of European colonialism, and the anthropology that grew up alongside of it.

How might we rethink culture? Let’s get back to basics. What happens that makes us think that another culture is in play?

What happens is, people do something in a situation that we don’t understand. For years I’ve called such moments rich points. Those moments of incomprehension and unmet expectations are the fuel that drives ethnographic research.

There are a lot of ways that an outsider might explain a rich point. The people might have made a mistake, an error. It might be something that is stupid or crazy. It might be that the people don’t know any better. It might be that they’re trying to pull a fast one. It might be some kind of -ism, a prejudice that that kind of person never makes any sense, so what would you expect?

These kinds of responses are all too typical. Sometimes they might be right, or some of them might be. But then again, what we don’t understand might well have something to do with this culture concept that we’re trying to make sense of. The odds, based on my lifelong record, recommend a bet that they will.

Of course a predilection for cultural interpretation can go too far into politically correct myopia. They did what? Whatever it was, it must be a cultural difference. Let us genuflect in respect.

A few years ago I helped an NGO director in Washington think through some problems she was having with a development program in China. She described in some detail the behavior of a local official and struggled desperately, with some guilt, over her lack of knowledge about that country. How, she wanted to know, could she learn to look at such cultural differences? I gave her some ideas, but first I told her, based on her description, that there was also a distinct possibility that the person was obnoxious, a universal pattern I’ve encountered in every culture I’ve ever worked in. She, of course, would have to check that evaluation with locals, but she shouldn’t think that every difference has to be cultural.

At first, she was upset that I would even suggest such a thing. Then she thought of reactions to that same person on the part of Chinese acquaintances and friends. The diagnosis looked like a plausible one once she considered how several of her local colleagues viewed him and talked about him as well. The guy might well have been a universal human jerk.

Same pattern when I worked in Mexico City, helping Anglo-Americans and Mexicanos start up a new business, I often wondered the same thing—what was cultural and what wasn’t? Sometimes differences between an American and Mexican partner looked pretty cultural. Sometimes they looked like two business guys working in the same framework, manipulating and negotiating in their own interests, sometimes using mutually understandable appeals to culture as part of the maneuvering. Where was the line? I wasn’t always sure there was one.

So how can we tell if something we don’t understand, a rich point, is cultural or not? The answer is, we can’t, not on the basis of just one occurrence. Einmal ist keinmal, once is nothing, wrote Milan Kundera (1984). All we can know at first is that something we didn’t understand caught our attention. It signaled a difference between us and what was going on at that moment. But then we need to compare and contrast across people and situations, chase the rich point, as I like to say.

Let’s say you’ve chased a rich point enough to know it’s not just characteristic of those particular persons in that one situation. And you’ve chased it enough to see that it patterns across certain kinds of persons and/or certain kinds of situations. That’s the first clue that something cultural is in play. The rich point was neither universal nor unique nor random. It warrants a tentative conclusion that there is a culture of those persons in that situation, at least for you, given your languaculture. A translation between your languaculture and the one in play is called for to explain what that rich point meant.

So far, so good. Assume that it looks like a rich point is probably cultural. What does that mean for the culture concept? It means that something is going on that enables communication and action among certain kinds of persons in certain kinds of situations, and whatever that communication and action are about is not understandable to an outsider like you. By the way, this is why participant observation is celebrated among ethnographers. If you venture out into the situations where people live and work, rich points will surface in no time and keep on coming.

Culture, then, is first of all a working assumption, an assumption that a translation is both necessary and possible to make sense of rich points. The assumption is based on observations of recurrent patterns of rich points across some common person/situation categories. The assumption is, there are shared meanings/contexts unknown to you. You have to figure out what they are. The classic ethnographic problem. Whatever we do with culture, it has to anchor there. Culture names the solution that you assume you can find. Culture is what you eventually show the world to explain meaning problems in terms of contexts.

A thread now runs back to the earlier lectures on ethnography and language. In those lectures I argued that ethnography yields a product that is a translation. Now we bring the same thread forward into this lecture by claiming that culture names the translation that ethnographers build for their audience, based on the investigation of the languaculture of the people who produced the rich points that the ethnographer experienced.

This chain of argument—from ethnographic product to translation to culture—lands us in a notion of culture whose peculiarities will be discussed in more detail in a moment. For now we’ve already gone beyond the bounds of the concept’s traditional meaning. You can see why some argue that we should consider dropping the word from our vocabulary. This will never happen, though, since anthropology only gets one vote in the crowded marketplace of global discourse, and global discourse is enamored of the term culture and isn’t about to give it up.

For now let’s continue with culture, continue to redefine the concept so that we can say “culture means this,“ even if we know what we say has the chance of the proverbial snowball in hell of ever influencing popular uses of the term.

Like a translation, culture is relational. Like a translation, culture links a source languaculture, LC2, to a target languaculture, LC1. Like a translation, it makes no sense to talk about the culture of X without saying the culture of X for Y.

Whenever we hear the term culture, we need to ask, of whom and for whom? Culture names the translation required, given contact between a particular source and a particular target. It might require a little bit of material, or it might require a lot.

Culture becomes visible only when differences appear with reference to a newcomer, an outsider who comes into contact with it. What it is that becomes visible in any particular case depends on the LC1 that the newcomer brought with them, a newcomer who might be an ethnographer, or perhaps an immigrant, or a new employee, or a tourist. Different LC1/LC2 combinations, different rich points, different translations, different cultures.

The shape that culture takes depends on the of whom/for whom boundary. Some boundaries will generate more rich points than others. This offers the intriguing possibility that we could measure, in a metaphorical way, cultural differences, but with a twist. We wouldn’t really measure difference between cultures in the old-fashioned sense of the term. Instead, we’d be measuring how much work has to go into a translation given a particular source and a particular target.

Not all translation at the boundary is created equal. Besides, whatever the nature of the boundary, different kinds of rich points will occur. For example, a major rich point is one thing—complete and total incomprehension due to a massive difference between source and target. Then there are more subtle rich points, particularly important between segments of the same society. In those cases, source/target comprehension may already exist to some extent. It is probably habitual, and more than likely it is wrong, at least in part. Then there are forced rich points, ones that would never have come up if an outsider didn’t make them happen. Twisting Shakespeare’s1 words a bit, we can say that some points are rich, some achieve richness, and some have richness thrust upon them.

When I think of my own checkered ethnographic past, the idea that languaculture becomes visible in contact with an outsider and that it depends on how similar or different the two languacultures are—it makes intuitive sense.

Here are some examples: Whatever mixed bag of cultures I represent personally, assume for a moment that I’m LC1, the target audience. The LC2 sources I’ve worked with have varied in how different they were from the LC1 I brought into the encounter.

First consider South India, the small village in which I lived for the better part of an academic year. At the beginning, the wave of rich points was overwhelming. That wave is what old-time anthropology called culture shock, a feeling of profound disorientation. Very little of what I knew how to say or do made any sense in the new place. In my experience, the village in Karnataka defines one end of my difference scale.

The other end, as far as ethnography goes, was the work I did with independent truckers in the early 1980s. I had a part-time job in high school dispatching tractor-trailers at a local quarry, so I knew a little about the industry. Even though I did the official study out of Baltimore many years later, I eventually figured out that the truckers and their families were the guys and gals I’d gone to high school with. I was very comfortable very quickly and mostly focused on rich points that had to do with the details of the work, the broad outlines of which were already familiar. The major rich point was my surprise that they disliked the popular image of the independent truck driver as the last of the cowboys. That surprise wound up as the organizing theme for the book I wrote.

Other ethnographies I’ve done fell somewhere in between, but the two ends of the scale make the point. Culture is a construction, a translation between source and target, between LC1 and LC2. The amount of material that goes into that translation, that culture, will vary, depending on the boundary between the two. Using myself as LC1, the culture of the South Indian village measures high on the scale. I only skimmed the surface. But, again using myself as LC1, the culture of independent truckers was much lower on the scale, because the boundary generated fewer rich points.

Does that mean independent trucker culture is simpler and less complicated than South Indian culture? No, not at all. The measure is comparative. That’s the point of the relational concept of culture. Simpler or less complicated for whom? My Karnataka-born-and-bred field assistant, Simeon, already knew a great deal more about that specific South Indian village than I did. The world of U.S. trucking would have thrown him into culture shock.

Culture isn’t a property of them, nor is it a property of us. It is an artificial construction built to enable translation between them and us, between source and target. It is intersubjective, as the jargon says. It needs to be elaborate enough to get the job done and no more elaborate than that. If source and target are already similar in meanings and contexts, it will take less culture to do the job than if source and target are far apart. The translation we build is the culture we describe. Culture is a lens built for LC1 that focuses on problematic meanings in LC2 and the contexts that render them understandable. Culture, as used here, is the property of no group. Instead, it is a relation between at least two of them, and the relation will vary depending on the two groups in question.

Culture is relational. There is no culture of X, only a culture of X for Y. How much and what needs to go into that cultural description depends on which X and which Y define the boundary. In that way at least, culture is the same as translation.

The complications get worse. Culture is not only relational. It is also partial. We can no longer say culture in the singular when referring to a particular person or a particular situation. The plural is now obligatory. A particular moment or a particular person or a particular group is never about just one culture. It is always about cultures.

Recall the earlier thought experiment: Several descriptions applied to the same person—generation, gender, religion, etc.—and each of those labels represented categories that people now refer to as cultural. What I did, or what the Cuban Jewish 1980s-college-era saleswoman from Mississippi did, was some mix of all those cultures and more.

If we had to connect all possible combinations of those five culture labels for a source and a target languaculture, assuming that each label had only two values, we would need to craft 25 × 25, or 210, culture translations, for a grand total of 1,024. That’s a lot of cultures. It seems a little unreasonable, not to mention laughable, especially considering that most boundaries labeled in the thought experiment wouldn’t be all that complicated—critical at a comparatively small number of rich points, but not too many.

On the other hand, we can’t just ignore the fact that all of us now are a mix of a little of this and a little of that, and the this’s and the that’s can be anchored in anything from a local family tradition to a global trend. And one’s attitude towards the this’s and the that’s can range from disgust to wild enthusiasm. In any particular situation, we craft a self, in Dorinne Kondo’s (1990) words. We put together who we are based on the contingencies and constraints of the moment. And what we use to put a self together involves that long list of cultural labels that all of us talk about now.

How did we get into this mess? Think of the old idea—culture as a closed, coherent system of meaning and action in which an individual always and only participates. No one today lives with a culture like that. What were we thinking?

Maybe it wasn’t so crazy in evolutionary terms. Think back to the so-called ancestral condition. Some 50,000 years ago, our ancestors turned into cultural animals in ways that went well beyond the archaic humans who preceded them. A symbolic world blossomed—language, paintings, sculpture, jewelry, ceremonial burial, and the like. This ancestral condition involved hunter-gatherers. They spent most of their time in small groups doing the same things over and over again. Contact with other groups was minimal. We know from game theory that one-time contact with strangers offers the best payoff if you assume the worst. And we know that with continual contact a cooperative strategy will pay off much better in the long run. The ancestral condition likely involved close-in cooperation with a small group that changed into worst-case assumptions upon contact with strangers.

These small, isolated, self-contained groups would grow a culture—in the old-fashioned Kroeber and Kluckhohn (1966) sense of the term—out of the practical activities that they engaged in. A variety of early cultures took shape as Homo sapiens adapted to a wide variety of environments. These first modern humans gained a tremendous selective advantage with their ability to use language to coordinate social action. And coordination of the same activities among a small, isolated group would produce the sort of culture that anthropologists used to dream of. Worst-case scenario assumptions with strangers would hold local cultural boundaries in place.

Not so long ago, finding the last of this kind of ancestral culture was a quixotic goal and a noteworthy event, not to mention an anthropological fantasy. Some readers may remember the discovery of the Gentle Tasaday in the early 1970s in the Philippines and the controversies it inspired. Were they the real ancestral culture, hidden away all these millennia, or not? Not, of course, as it turned out.

Anthropology preserved this ancestral image of culture, even though—by the time the field developed in the late 19th century—the image no longer applied. Even when I worked in a small South Indian village in the mid 1960s, it wasn’t difficult to maintain the delusion. Looking back at the village, I see the dozens of cracks in the image. But now, in most any community, the cracks in the traditional image have turned into canyons. Any community is about cultures now, plural, and everyone in that community has a different mix available, and everyone draws on a different subset of that mix in different ways. No person, or group, can be described, explained, or generalized completely with a single cultural label.

So now what do we do? If we’re going to save the culture concept, what changes do we have to make?

Culture is always plural. In today’s world we have to train ourselves to say, think and write it with an “s“ at the end. Never culture. Always cultureS. Anytime we use the concept with reference to a specific source and target languaculture, we have to understand that it will work only in part, some of the time, in any specific situation or for any particular person when we try and apply it.

So one strategy would be to accept that multiple cultures are always relevant and worry about how to explain their interactions.

Imagine, as a thought experiment, how cultures might or might not mix in a situation. A first issue might be hegemony. Hegemony is about how dominant a culture is once it is in play. As an instance of strong hegemony, I remember the time I lived on a sailboat on Galveston Bay. The live-aboard pier consisted of a small group of people who were as different from each other as a small group could be. The rule was—it was laid out explicitly for me when I moved in—that the only thing in life that mattered was the boat and the water. Talk about that or don’t talk at all. Early on I walked onto the pier carrying a briefcase. “What’s in there,“ asked another sailor. “My other life,“ I replied. “Good,“ he said, “that’s all the space it deserves.“ On the pier you were a sailor and nothing else.

A second issue is what we might call coupling, in the sense of what a culture drags in when it is in play because of what necessarily links to it for that person. For example, as I’ve gotten older, I’ve noticed that, more and more, generation overrides everything else. It doesn’t matter what I’m doing or what kind of person I’m talking with, if the other person and I are in the beyond-50 range, an “old fogey“ comment is likely to find its way into the discourse. The experience of aging among the baby boomer generation calls for comment all the time. Generational culture is one of those things that couple with most everything else.

A third issue is density, in the sense of how pervasive a culture is, how much situational relevance it has. For example, I’m an ethnographer. Often people assume I’m an ethnographer all the time, no matter what else I’m doing. That would be insane and destroy many things I enjoy with a degree of self-consciousness that contradicts the reason I enjoy them. I’m remembering an old cartoon with a couple in bed, where one says to the other, “What'll it be tonight, feeling or technique?“ But still, it’s true that my occupational culture distributes across a good proportion of what I do all day. It’s dense.

A fourth issue we can call attitude. Attitude just means how you feel about that particular culture. Is it something you’re stuck with but you try and ignore or conceal it? Or is it something that you’re proud of and want to display, maybe to the point where you overdo it? I once got into a conversation with a guy in a bar in a hotel where anthropology meetings were to be held. The two of us stood there, drinking beer, talking about sports and the weather and all those ordinary guy conversations. Finally he asked what I was doing, and I said I was there for a meeting. Then we both grinned and admitted we were professors of anthropology, and laughed long and loud. We’d both tried to hide that fact and just be ordinary guys. We laughed in part because it was just funny, and in part because it was a telling drama about the ethnographic aspiration to downplay the Ph.D. and just be one of the people.

A fifth issue is integration. Integration is about the degree to which people have figured out how to put their multiple cultures together into as seamless a bundle as possible. The saddest example I ever met was a Kurdish student at the University of Vienna. His parents were Iraqi and Kurdish, but he was oriented to the U.S. and Canada, although he was living and working and studying in Austria. In his own words, he was a chaotic and confused mess in terms of who he thought he was and how he thought he should speak and act.

A final issue I’ll mention here is volatility, the degree to which a culture in the mix is undergoing dramatic change. Years ago I worked on a study of international leadership. The representatives from South Africa, White and Black, said they really had no idea how to talk about South African culture in any coherent way right then, what with the volatility as the country shifted from apartheid to a multiracial state under then newly elected President Mandela. The questionnaire they were supposed to administer assumed a normal situation that one could measure. There was no normal situation, to put it mildly. What was interesting was how things were changing.

    Even this quick overview of some of the ways different cultures might mix and match for a person in a situation looks more like the problem rather than the solution. Hegemony, coupling, density, attitude, integration, volatility—it somehow sounds like the loyalty oath of a paramilitary organization rather than a solution to the problem of multiple cultures. Even if we could come up with a theory of cultural blends, what would be the use? It sounds like a strategy to preserve a concept that just gets in the way of figuring out rich points.

Maybe a better strategy would be to quit worrying about what culture might mean nowadays. The way I’m talking about it here, culture isn’t what we study anyway. Culture is a product of what we study. So what do we study?

Let’s go back to the ground again. We start with rich points, the marker of a boundary between an ethnographer’s LC1 and some people’s LC2. That’s the working assumption. Let’s just build from that for a moment.

Why would we be noticing a rich point in the first place?

Well, it could be a top-down kind of thing, or it could be a bottom-up approach. If the work is supported by a grant or contract, it certainly started top-down. Seldom do funders support a person to wander around and look for rich points. If rich points did come from wandering along through life thanks to independent wealth or unreimbursed curiosity, the work might start bottom-up. Insiders know that most ethnographers start top-down and then change because of unexpected bottom-up experiences. This gives us a bad reputation with funders but relevant and powerful results born of the ability to respond to what we learn after the funded study starts.

Let's first consider a top-down approach. A top-down approach begins by setting constraints around the world within which rich points will be sought. In this series of lectures, several constraints bounded the examples I used—a family law department in a courthouse, people physically dependent on heroin, truck drivers working as independent contractors, a group of guys starting up a U.S./Mexico international company, and a small village in South Asia, to repeat part of the list.

Notice how these constraints limit the world within which rich points come up. They make the research possible. They bound the world by specifying categories of people engaged in linked sequences of activities. It’s a pretty safe assumption that if people repeatedly engage in tasks together, understandings will emerge and become habitual, and turn into tacit background knowledge that newcomers will notice by its absence.

Our ethnographer, like any other newcomer, will lack this tacit background knowledge. As he/she enters into the actual world of those linked activities, or as he/she listens to people who perform those activities talk about them, rich points will surface that mark the boundary between the experienced and the novice. And those rich points will become the focus of an ethnographer’s work.

So in this top-down case, the culture in languaculture becomes whatever the ethnographer learns and makes explicit so that a person with the same languaculture as him or her can understand those people doing those activities like an insider does. Maybe the C in LC2 should just stand for context instead of culture, since that’s how an ethnographer, in the end, explains rich points.

But what about all those other cultures that we’ve talked about in this chapter? Don’t they play a role?

Very much so.

As top-down ethnography progresses, one looks for the relationship between the world bounded by those prior constraints and the rich points that come up. But one also looks for patterned variation in how those constraints are interpreted and managed. Notice the phrase patterned variation rather than the concept of variable. They’re different, very different. Generally we expect patterned variation to be a power law, that is, a few patterns will explain most of the variation, like the old 80/20 rule in business, 80% of the business comes from 20% of the products.

Those few patterns will emerge, in part, from the various culture blends we speculated on earlier. Which blends? Who knows until the work is underway. Notice the difference. Rather than worry about all possible cultures and all possible blends, we let what we learn from the person/activity sequence aim us at the relevant ones to explain the patterned variation that we find.

So in the end what do we get from this top-down approach? We predefine person-activity sequences that will be the source of rich points, and we pursue a translation that models tacit background knowledge together with the major variations in pattern, where those major variations, in turn, show us which additional cultures are most relevant. We don’t try to guess at all possible culture blends and how they work, an impossible task at any rate. We let the person/activity focus tell us which ones are most relevant.

Doesn’t that leave out a lot? Yes. Does it look anything like the traditional concept of culture? No. Should we just dump the traditional concept rather than trying to include it in our description of what we do? Intellectually I think maybe we should. Practically, though, probably not. More on that question in a moment.

What about a bottom-up approach?  I’ve done that as well. Two examples came up in the lectures, the first in New York and the second in Vienna. In New York I was struck with how the then new methadone maintenance program was interpreted differently in the clinic and in the street, and this difference led me to wonder where the program had come from and how it had developed. In Vienna the concept of Schmäh, which I had thought I already understood, turned richer and more complicated when I saw how widely it was used and how different those uses were.

Those examples started bottom-up, not top-down, because the initial question didn’t prespecify person/activity sequence boundaries within which rich points were sought. In fact, both these examples came out of situations that people like me dream of. In New York I had a regular paycheck and a diffuse assignment to just go out and find something interesting to do. In Vienna I was on sabbatical from the University of Maryland, a time designed to free faculty from the everyday madness of meetings and office hours to unclutter their minds and seek new ideas.

With bottom-up rich points, the approach changes, though, because the ethnography involves the many forms the rich point takes in different languacultures—in different domains and at different levels of scale. It asks the question of why that rich point exists at all, a blend of Malinowski and Foucault, what I think of as the Malinault or Foucowski approach. From Malinowksi I take the emphasis on learning a rich point by watching how it is used in a variety of different activity sequences by a variety of categories of people. From Foucault I take the question of where the rich point came from, why it exists at all, what history produced it and what political forces hold it in place?

This is a different breed of ethnographic cat. It foregrounds chasing the rich point across domains and levels rather than chasing it in order to translate a specific LC2 in use at some person/activity coordinates in the social world. s, though, other cultures clearly come into play with bottom-up rich points as well. As rich points are chased from place to place and level to level, they take different shapes depending on the different cultural mixes in those places and at those levels. And the way they are in play will vary along the lines of the different ways that cultures blend, as described earlier. This time, though, the variation is in how the rich point looks and in the way that one level controls how it looks at another.

Let me ask the same question as at the end of the previous section. Doesn’t this leave out a lot? Yes. Does it look anything like the traditional concept of culture? No. Should we just dump the concept? Intellectually I think maybe so. Practially, though, probably not.

So in the end, I’m stuck with culture. The concept has achieved widespread use among a bewildering variety of people, and most of those people assume I am an expert in whatever it is they take the concept to mean, and they assume that I will be delighted to hear the sound of it when we talk.

It’s odd to see ideas you’ve advocated for so many years get taken up and, in the process, turn into part of the problem rather than part of the solution. I’m remembering stories about how famous person X, at the end of his life, says “I am not an X-ian.“ Freud says “I am not a Freudian,“ Marx says “I am not a Marxist,“ or at least so the stories go. This is about a concept, not a person, but the idea is the same. Imagine an ethnographer saying, “I don’t study culture.“ Same kind of thing. Yet some days lately that’s exactly how I feel.

What should I do about culture?

First the popular use. Obviously, yelling at people whenever they say “culture“ isn’t a very productive strategy. So what I do is, I grin and bear it if the concept is used in a good way. What’s a good way? A way that isn’t so far from what it is that all of us have always fought for. If the word is used to recognize that a difference is in play that might mean something from another point of view, then it’s a good use of the term. If the word is used to counter the deficit theory—i.e., that those people are too stupid to know any better—then it’s a good use of the term.

On the other hand, if it’s used to label a difference without understanding anything about it, then it’s a bad use. If it’s used to justify a position of moral relativity, then it’s a bad use. If it’s used to explain away a difference without doing anything about it, then it’s a bad use of the term. If it’s used in a politically correct way with no consequences—“cultural competence“ is one of my current favorites—then it’s a bad way to use the term.

The list of good ways and bad ways could go on for awhile, but you get the idea. It’s not that a professional should never speak up. It’s that he or she should encourage a good direction and undermine a bad one, even if the culture term is being used in outdated ways described at the beginning of this article. It’s not about whether you can prove you know better; it’s about the moral value of the discourse in which you participate.

Now let’s get technical.

Among the professionals, the concept may well have outlived its original uses. At the most general level, the culture concept presents a classic problem. It was used to label what was assumed to be a closed, coherent system in equilibrium. Whatever we are dealing with today when we do ethnography, we are not dealing with systems like that. We are dealing with open dynamic systems co-evolving in their environment, complex systems on the edge of chaos, as the jargon sums it up, a lot of them all at once.

A well-known character from that world of complexity, unfortunately no longer with us, wrote a book called Open Boundaries (Sherman & Schultz, 1998). He was a character, Howard Sherman was, because he started out in philosophy, then went into business, and finally wound up advising companies. Here is a quote from his book:

When we try to close systems that in fact cannot be closed, we create a radical incompatibility with nature and business. The consequences of this incompatibility are severe: a retardation of coevolutionary development, a gross limitation of novel innovation and new possibilities, and an eventual and untimely death of the system. (p. 4)

Trying to make the old culture concept work in light of the kind of ethnography I do feels just like that, like taking a concept whose very core is about closed systems and trying to hammer it onto a fluid, dynamic world. It’s like trying to build a computer out of plywood.

But I did give it a try in this lecture. I came up with a notion that culture is another name for the translation between LC1 and LC2 that an ethnographer builds as a product of his/her work. But as we dived into the details, the traditional concept didn’t fit.

First, culture becomes visible only when an outsider encounters it, and what becomes visible depends on the LC1 of the outsider. There could be a big difference; there could be a little difference; and the translation would vary accordingly. Second, culture is relational. It is the property of no one and exists only as a translation enabler between LC1 and LC2, its content, again, being a function of which LC1 and LC2 define the boundary. Third, the boundary of the source LC2 that an ethnographer sets is partial and fuzzy, typically involving some categories of persons engaged in activity sequences. Fourth, bottom-up ethnographies are driven by the multisite, multilevel chase after a particular rich point. Bottom-up work is outside even this definition, since such ethnographies focus on how and why rich points shift and change as they cross social domains and levels of scale. Fifth, if we look at any particular kind of person or situation or group, more than one culture will always be in play. We have to think of the plural, cultures, for a full understanding of any given moment. And finally, if we do try and make sense out of all relevant cultures at some moment, we have to examine not only them but also the many different ways they can interact with each other.

I’m happy with this list as a series of steps on the way to a reformulation of the culture concept, given what it is we do and how the world works now.

But does it make any sense? If culture is taken as just another name for the translation that an ethnographer crafts, why do we need that other name? Especially when the culture concept carries so much intellectual and emotional freight within anthropology and in its many uses in everyday discourse?

Should we dump the culture concept? Intellectually I think so. Practically, probably not. Even with the proper revisions, the concept carries so much historical and personal noise that any new content is obscured. But politically it serves many purposes in everyday discourse that any professional would approve of.

A reader will have to make his or her own judgment call. I’ll continue to live with the contradictions. A colleague commented after the original lecture that she wanted to save culture and continue to use it professionally, but now she’d always think about it whenever she uttered what she called the C word. That’s good enough for me, at least right now.

But personally, in my professional work, I think I’ll stick with the concept of translation. Calling it culture is fine, for the professionals, as long as we change what it means away from the old image of the ancestral condition and realize what it has to mean and can’t mean now. And that exercise, as I hope I’ve shown, challenges our ability to wrap our minds around two wildly different concepts that just happen to have the same name. To modify a grisly line from the Vietnam era, it’s like destroying a concept in order to save it.

Just one more thing, as the famous TV detective Columbo used to say. Culture in the old-fashioned sense actually had two meanings, and I’ve dealt with only one of them here.

The meaning dealt with here started with the old idea of culture as a particular society’s shared system of meaning and action. Culture, in this sense, emphasized differences between one group and another. The study of culture was about the elucidation of those differences. That’s the sense of culture that has been stretched and changed and applied here. Rich points, the driving force of the whole enterprise, are about differences between one languaculture and another.

But in the traditional use of the term, that wasn’t the only meaning. Culture was also used to label what was uniquely human, what it was that differentiated us from the other primates and from the archaic forms of humanity that came before us. Modern humans have culture. Animals don’t.

Culture has turned into a problem in this meaning as well, but for now ignore the many fuzzy boundaries that have surfaced in such fields as primate communication and evolutionary social science, boundaries that show that many behaviors previously considered human specialties exist among other species as well. But in spite of these more fluid boundaries, it is clear that humans are a breed apart, for better or for worse.

Culture, in this sense, is about human universals, the things that unite us and make it possible to connect with people right away and, eventually, with the right intentions and a certain amount of work, make sense of what the others say and do.

Ethnography is making sense out of human differences in terms of human similarities. Without the universal connections, the human similarities, translation would not be possible. Those universal human connections are what this second meaning of culture is all about.

This aspect of culture I haven't dealt with. A couple of stories here and there touch on it. For the moment, I just want to mark the omission. Human universals are the ground against which the figure of rich point translation is possible. This meaning of this other part of the culture concept deserves its own treatment. I want to foreground its significance and note that it has not received that treatment here. It’s a topic for another day.

For the present the focus on culture as shared meanings and practices has produced headaches enough. It grew out of an expanding European world as other parts of the planet were brought under its sway. It signaled a difference between Europe and the occupants of those distant places. Anthropology claimed it as their core competence and created a science of humanity. Isolated small communities far away from Europe and the U.S. became the unit of study.

The problem now is, the historical conditions that bred the concept and the historical conditions under which we now want to use it are wildly different, contradictory, light-years apart in the kinds of worlds we attempt to understand. Maybe the concept belongs to that earlier stage of history. Maybe using it now is like trying to find the gills on a cat. Or maybe it can be rescued by adapting it to today’s world, as I’ve tried to do in this article. It’s not clear right at the moment. But if I’ve convinced the reader to always wonder what they, or anyone else, mean when the C word is used, I will have done the job I set out to do with this lecture.


1. Twelfth Night, Act 3, Scene 4.


Agar, M. (1995). Language shock: Understanding the culture of conversation. New York: William Morrow.

Agar, M. (1996). The professional stranger: An informal introduction to ethnography (2nd ed.). New York: Academic Press.

Kondo, D. K. (1990). Crafting selves: Power, gender, and discourses of identity in a Japanese workplace. Chicago: University of Chicago Press.

Kroeber, A. L., & Kluckhohn, C. (1966). Culture: A critical review of concepts and definitions. New York: Random House.

Kundera, M. (1984). The unbearable lightness of being (M. H. Heim, Trans.). London: Faber and Faber.

Sherman, H., & Schultz, R. (1998). Open boundaries: Creating business innovation through complexity. Reading, MA: Perseus.

 International Journal of Qualitative Methods 5 (2) June 2006