Computers and Sociocultural Anthropology

Savage Minds welcomes guest blogger Nick Seaver.

The fundamental requirement of anthropology is that it begin with a personal relation and end with a personal experience, but […] in between there is room for plenty of computers.
– Claude Lévi-Strauss, epigraph to The Use of Computers in Anthropology 1

Recent years have seen the growth of what we might call “alternate universe anthropology.” People with little or no training in anthropology are taking on big sociocultural questions, and they’re doing it with computers. We find PhDs in Electrical Engineering trying to algorithmically define musical genres, computer scientists modeling family ties in social networks, and autodidact software developers designing “content discovery” apps around their own theories of cultural influence and flow. If sociocultural anthropology didn’t already exist, people might reasonably assign the name to this stuff.

To academically-trained anthropologists, these computery projects can seem like they come from another planet. Anthropology, we know, is the stuff of notebooks, close friendships, and handkerchiefs tucked into your glasses. The social theories that pop out of companies like Facebook and the cultural analyses produced under the banner of big data are typically objects of scorn among anthropologists. We do not see them like we might see the “cultural” theories of members of an Amazonian tribe or the “social” theories of members of an ethnic enclave in southern Europe. We see them as simply wrong.

However, the fact that we have competition in producing accounts of culture and society should be no surprise: as Nicholas Thomas has written, “the objects of anthropological knowledge […] have never been exclusively anthropology’s own.”2 One could argue that the defining epistemic feature of sociocultural anthropology is its necessary coexistence with competing explanations, both from the people anthropologists study and the others who study them.

I conduct my own fieldwork with the developers of algorithmic music recommender systems in the US (think Pandora, iTunes Radio, etc.), and my goal is to investigate their theories about “culture,” how they think about those theories, and how they mediate between the idea that culture is intrinsically subjective while algorithms are intrinsically objective. Trying to relate stories from the field to other anthropologists, I find that we tend to see this cultural theorizing as the work of uneducated interlopers who are solely driven by economic or technical concerns. There are some serious issues with this situational suspension of anthropology’s trademark interpretive charity (Do we really think that engineers have succeeded in becoming homo economicus or that their theories of culture are simply errors to be dismissed?), but one of the most vexing is the idea that these knowledge projects, because they are essentially computational, are not really “anthropological.”

This is a case of selective memory: For nearly as long as computers have existed, there have been anthropologists making use of them.3 In my attempts to make sense of the knowledge practices I encounter in the field and to remind my own discipline that computers did not arrive from some distant land to cause us trouble, I’ve been studying the history of computing in sociocultural anthropology. Although much has changed in terms of hardware, software, and popular imaginaries about computers since we started messing around with them in the 1950s, there are some remarkably persistent debates that keep popping up — about formalism, quantification, and the division of research labor, among others. In many cases, computers were embraced by parts of the discipline that were later disavowed by the mainstream as we moved down a more symbolic, interpretive, literary path.

Returning to some of our past disciplinary engagements with computers might help us think about how we imagine the boundaries of our discipline, and it can help us make sense of this strange world of “alternate universe anthropology,” where ideas about culture we’ve cast out have been picked back up or reinvented. Over the next two weeks, I’ll be posting a series of sketches from this history. They are by no means exhaustive, and they are missing plenty of interesting work,4 but I think they are instructive. I hope that you, dear readers, will take to the comments to point out gaps and oversights and, hopefully, to share your own histories and experiences from the trading zone between computers and anthropologists.

A hint of what’s to come:

  • Structuralism: Thinking with Computers: Structuralists like Claude Lévi-Strauss and Edmund Leach were fascinated by digital computers both as metaphors for cultural processes and as tools which might fulfill structuralism’s methodological promises.
  • Ethnoscience: Being Scientific with Computers: As the scientific standing of sociocultural anthropology was debated in the post-war period, computers became tied up in broader debates about the merits of quantitative and formal methods.
  • Cultural Ecology: Modeling with Computers: For cultural ecologists and cyberneticians, analog computers offered models for feedback systems in the mind and the environment.
  • Personal Computing: Ordinariness and Materiality: The introduction of the personal computer allowed computing to happen in the field, which led to a number of new problems regarding dirt, humidity, and the typing up of field notes.
  • Computing: From Method to Object: At the close of the 1980s, as computers moved into anthropologists’ traditional field sites and those field sites expanded to include “high tech” settings, computers shifted from being primarily a tool to being an object of study in their own right. With computers as both objects of and tools for anthropological inquiry, we are thrust into the clutches of reflexivity, and hopefully it is not all that bad.

  1. Dell Hymes, ed., The Use of Computers in Anthropology (New York: Wenner-Gren Foundation, 1965). 
  2. Nicholas Thomas, “Becoming Undisciplined: Anthropology and Cultural Studies.,” in Anthropological Theory Today, ed. Henrietta Marks (Cambridge: Polity, 1999), 262. 
  3. Although I’ll be focused on electronic computers, historians of computing are fond of noting that “computers” used to be people, primarily women, who performed calculations for their employers. In line with that history, one of the first computers to be mentioned in American Anthropologist was actually a woman named Amy Barrington, hired in 1907 by Francis Galton to work with Karl Pearson on a eugenic survey of England. 
  4. Most notably, this history peters out well before the contemporary growth of various flavors of digital ethnography and anthropology, and it neglects work that was done with computers through the 1990s

4 thoughts on “Computers and Sociocultural Anthropology

  1. Looking forward to these posts! I’m a first year Anthro student at UCR, and I will be studying communication around the Typhoon Haiyan disaster in the Philippines. It’s definitely tricky working out an understanding of computers/culture/community.

  2. Very much looking forward to the series! I’m wondering whether one element will be computer models, such as agent-based models, that try to investigate dynamic patterns of cultural change within social groups without necessarily assuming that doing so fully captures or provides a reduction of cultural change.

  3. Nick — this is a great subject and I am glad to see younger anthropologists taking it up. You may be interested in looking the syllabi of the courses I have taught on the subject. Like you, I trace the origins of the anthropological engagement with the computer back to structuralism and trace its development through various phases, where the computer changes aspect from metaphor, to tool, to object of investigation. A key figure in this transformation of perspective is Lucy Suchman, In any case, here’s a link the the original syllabus —

  4. @ontoligent: Thanks! That syllabus is wonderful and thorough (although I had to request permission to access the file). It seems like in contemporary graduate training, symbolic and interpretive anthropology cast a long shadow that blocks out a lot of various post-war formalisms. That causes two big problems: 1. we understand that things like thick description are a good idea, but we don’t get a sense of the contexts in which such things emerged, and 2. we don’t realize that many contemporary computational knowledge practices have antecedents in our own disciplinary ambit. This is not to say that we need to start doing mathematical or cognitive anthropology again, but many current arguments against/about “big data” or “algorithms” are based on an overly rigid and unreflective idea about how the borders of anthropology/ethnography are historically constituted. (One of my favorite weird examples of this is the bit of time during which ethnoscience was called “the new ethnography” — at least it is a good way to troll people who continue to think of ethnography as some timeless method…)

Comments are closed.