In the previous blog, Part 2, a first building block of an anthropological perspective was proposed,  awkwardly named “learning task communicative competence from those who actually do it.”  This part of the perspective makes it easy to see, in retrospect, why the shift from a traditional Indian village to a U.S. treatment center for heroin addicts  wasn’t as big of a deal as I thought it was going to be. I just kept doing the same thing in Lexington, even though by the anthropological rules of the times I wasn’t supposed to be there. I wondered what “patients” and “staff” were up to and how they got it done. I spent time with them in “the joint”—the residence, the dining hall, the chapel, the gym, therapy groups. I was still looking at things anthropologically, even though Lexington was out of bounds as a traditional  “field site.”

Here in Part 3, I’d like to suggest a second building block, this one with the much simpler name of “pattern.” Pattern will be a second part of the argument that this perspective—I’m still not sure whether to call it an “epistemology” or an “ontology” (or possibly a “pathology”)—works the same once you learn it no matter what kind of anthropology you’re doing, including applications where you’re not actually doing research at all.

Traditional behavioral/social science obsesses over “variables” rather than “patterns.” For some domain of situations or persons, a variable is an attribute that can be identified in each specific case. The rules for identifying it are set by an “operational definition” which means a guideline for how to assign a numerical value to a particular observation, that assignment being both “valid” and “reliable” with reference to the actual variation the researcher is interested in. It will be “objective,” meaning that any person anywhere else would assign the same value to the same observation. This language will be familiar to anyone who has taken Stat 101 or Intro to Research Design in a social-behavioral science program.

The thing is, if you’re trying to learn communicative competence for a particular task, “variable” gets in the way. “Variable” narrows your attention. It means stop the flow, then isolate, operationally define, and figure out how to assign a number to some fragment from one moment to the next, never mind what else is going on in those moments. Variable has its uses, an example in a moment, but not when you’re trying to learn task-based communicative competence. That’s about figuring out a pattern in the dynamics of a moment, not ignoring everything and chasing only the measurement of what the traditional behavioral/social sciences would call a “prior” and the rest of the moment be damned.

According to the dictionary, “pattern” can mean many things—repetition, like a pattern of fur on a cat, or a template, like a diagram that shows how impossible it will be to assemble your kid’s Christmas present before sunrise. But in this blog, pattern is meant to be more general and abstract, agnostic as to what is connected and what the connections are made of. The point about pattern is that it—rather than a variable—is what an anthropological perspective leads you to look for.

It’s easy to describe the difference formally. You notice some interesting “x.” With visions of variables dancing in your head, you want to define x so that you can build a set {x1, x2, x3 … xn}. Then you can input that set into the statistical software. If you’re learning a pattern instead, the abstraction looks different. Now when you notice an x, you look for something else, a “y”, and a relation “R,” such that xRy is true.

An anthropologist trying to learn task communicative competence is after patterns, not variables. We notice something interesting and we wonder what else it might be connected with. Gregory Bateson—I fall back on his work again like I did in the previous blog—made famous the phrase “the pattern that connects.” You see something interesting? What else is it connected with in the task and in the world in which it is done? The puzzle metaphor used at the end of the previous blog still works, only now it’s more difficult (and interesting) because new pieces keep appearing on the table that weren’t in the original box.

This is not a normal attitude in behavioral/social science or organizational management, though it is in intelligence analysis, investigative journalism, police work, and history, but that’s another story. Ruth Benedict’s classic book is called Patterns of Culture, not “variables of culture.” That phrase would be more like Geert Hofstede’s “cultural dimensions.” Anthropologists who go to hell after they die are forced to read it for eternity.

Here’s an example that shows the difference between variable and pattern and one possible connection between them. Recently I had a conversation with a colleague about “big data.” Millions of variables are now available among all the digital breadcrumbs that we leave behind as we wander the forest of life. (That might make a good country/western song). How, my colleague wondered, could all those variables be packaged better for a smart phone user?

One answer we came up with was this: Big data variables have no pattern. Sure, you can spear a database with regression lines ´til the chips come home, but who knows if the linear equations have anything to do with what people really want to know when they ask Siri for help while engaged in their tasks? But … the tasks that the people with smart phones are actually doing—they do pattern, from the phone-owner’s point of view. If you knew the task pattern, you might be able to write some software that could cherry-pick the relevant big data variables and have Siri ready to respond with task-relevant information when you ask her a question. Or him, depending on which voice you selected.

That’s what the Target store chain did in one of the early commercial uses of big data. A pregnant woman shopping in Target follows a pattern. She changes her purchases as her needs change through gestation and birth and early childhood of her baby. Her purchases at time T give the marketing department a pretty good idea of other things she’d be looking for at time T + 1. The clever Target marketers issued ads and coupons at the right moment to direct her attention to new products she was probably already thinking of buying, or maybe she hadn’t yet thought of them but might buy them once she knew they existed.

The Target story shows the difference between variable and pattern and illustrates how the two might work together, possibly to accomplish things more significant than just to increase Target’s bottom line.

Is it really true that any pattern is possible in anthropology? No, not really, not once you’ve graduated from liminality into full clan membership. “Anthropology has no theory” scorn the traditional scientists. Actually, it does, but theories don’t look like Euclid’s Elements. Instead, they look like pattern templates that tell you, before you head out into the real world, “This important slot better be filled in in your pattern when you come back.” Marx says you’re going to have relation of task to the mode of production. Critical theory says you’re going to tie task to interests of the one percent. Bourdieu asks you what sorts of capital are in play. Semiotics asks what might have happened but didn’t, and how does what happened co-occur with other things that did, and when you’re done deconstruct the whole damn thing. Gender theory requires links between task and gender identity of human participants. Me, I’m a complexity guy, I want feedback loops and nonlinearities in my patterns.

But in the end, outstanding anthropology, in my view, doesn’t just funnel what it finds into a pre-fab pattern template. On the contrary, it puts the ready-to-wear pattern on the shelf, learns the tasks first, and then re-shapes or re-invents or just dumps the original template in favor of a pattern that preserves what was learned on the ground. Run-of-the-mill anthropology, on the other hand, concludes with, “So here we have another example of pattern X, for I am an X-ian.” It’s anthropology’s way of “replicating,” not to be scorned, but not as powerful as an innovation in how we understand the way the world works.

Anthropologists might agree, or so went the argument in the previous blog, that learning task communicative competence is the primary job from which all else follows. They will often disagree, to put it mildly, on what counts as the most important pattern template that connects up parts of a task with other things outside of it. Nevertheless, my point in this blog is that anthropologists will agree that a search for pattern, be it in a village or a treatment center for addicts, will be the first order of business rather than a search for variables that can be isolated and measured.

Maybe after you get the patterns straight, maybe then you want to define and measure a couple of their pieces—now you can call them “variables” if you’re so inclined.  Maybe you want to do some traditional behavioral/social science to see if the pieces act like the pattern says they should across a wide range of tasks. If you do that, you will already have a running start. Any two pieces of a pattern that co-occur most of the time as you learn them, now converted into variables, will probably produce a graph, maybe nonlinear, that’ll knock at least one sock off a pattern-challenged peer reviewer, who will then ask, “How did you measure that?” And at that point you can show them the pattern and knock the other sock off, too, because you didn’t have to make up some pattern in the conclusion. Instead, you started with it already in the introduction.

In the previous blog, I mentioned an example of a pattern problem in the South Indian village. Since the Naik, the traditional headman, wasn’t acceptable as a leader, I wondered how Naik-assigned tasks got done. I looked at–sometimes literally–several conflict cases and learned that the Naik pretty much played no role whatsoever, even though that was one of his main jobs. What actually happened was that the “Daav,” officially “village council” but better translated as “group of respected old guys,” pretty much ran the show. The Daav, as it turned out, was much more fluid in membership on any particular occasion than its description implied. Sometimes it involved participants from outside the village.

There’s more to that story, but the point here is that an anthropological perspective aimed me at the question, what is the pattern that best models how actual tasks get accomplished? When I started work in the addict treatment center, I did the same thing.

For example, one pattern that jumped out right away was the difference between how addict-patients talked during “treatment” tasks with professionals and how they talked when doing things with each other in the residential unit, the dining hall, the gym and so on. In this case I got lucky. It turned out that there were two items of folklore known to most addicts in the joint. They were called “toasts,” and they modeled the two patterns pretty well.

I’ve written about the two toasts elsewhere, in the Journal of American Folklore long ago. I’ll just tell you the first line of each one to give you the flavor. One, called “Honky Tonk Bud,” starts with “Honky Tonk Bud, the hip cat stud, stood diggin’ a game of pool.” The other one, called “King Heroin,” starts with “Behold my friends for I am King Heroin, known to all mankind as the destruction of men.” Bud’s story, spoken in street language, shows an accomplished street hustler. King Heroin, in standard English, describes a social-psychological failure.

The toasts summarized two patterns of addict communicative competence that I was seeing across different tasks. A major policy problem was that the professional literature and the Lexington program only acknowledged King Heroin as who heroin addicts were. They were also Honky Tonk Bud, in the joint and the more so when they went back home and hit the streets.

“Pattern” is my nominee for the number two fundamental on the list of what an anthropological perspective is all about. As we set out to learn task communicative competence in a world we’re researching or practicing in, we go on a pattern quest, a search for the “pattern that connects,” the way that the different pieces that we learn in our Level Two learning connect up into a bigger picture of people and the tasks that they do. And we don’t just look inside the task. We look outside it as well, and back in time, and around it in the broader world.

And, like task communicative competence, the pattern quest is also part and parcel of applied research, or practice, or short term projects like I’ve done since I left the university in the mid-90s. Here’s a brief example.

The Thomas Edison museum in West Orange, New Jersey, wondered why townspeople and former workers did not participate in museum tasks, such as volunteering or helping with interpretation or just showing up at events. This was an interesting twist on task-based communicative competence—explain the absence of people that the museum staff thought should have been there because they supposedly already had it. So I visited with a small sample of townsfolk and retired workers.

It turned out that, in 1973, McGraw-Edison, the corporate descendant of the Thomas A. Edison company, closed up shop and moved south to a right-to-work state. This was a severe economic blow to the town of West Orange. In my first interview, with a town leader, in the 1990s, I heard what most other people I met with said, in one form or another. “Every time I drive by the museum I see a closed factory.” Most people I spoke with saw the “factory” closing as a betrayal, the beginning of a decline in their community. That the former economic center of their town was now a “museum” only added insult to injury. It did not fill them with wild enthusiasm to serve as a resource for “closed factory” activities, to put it mildly. That pattern was clear in no time at all.

Pattern, and the task based communicative competence described in the previous blog, strike me as foundation stones of an anthropological perspective. They are simple to state and profound in their consequences. They go together like the proverbial bread and butter, a useful cliché that reminds me that I’ve made a living with them in academic, applied research, and practitioner working worlds for five decades.

So what is the next fundamental I should propose in the next blog, Part 4? The problem with this blog series is, once I started, the law of “one damn thing leads to another” reared its ugly head. Possibly a sign of the rambling tendencies of old age – or possibly another sign of an anthropological perspective. I only have space in my two-week run for one more blog that describes one more fundamental. There is more than one left on my list. What to pick?

Well, speaking of one damn thing leading to another: As I thought back to the old days, another fundamental leapt out, the notion of “emergence.” This time, though, there’s a twist. We used the word all the time in the university but didn’t really think much about what it meant. I took it for granted in the village. But it turned into a major headache in the treatment center. What I’d learned as a fundamental part of an anthropological perspective turned into an argument I had with colleagues from the day I arrived in Lexington until the day I left the drug field in the early 2000s. The transition from village to joint was easy as far as using the perspective went; the view of this fundamental from scientific and medical and administrative colleagues? Big problem.

So, let’s talk about emergence to show a hard part of the change from village to joint. That’s where we’ll go in Part 4.