Structuralism: Thinking with Computers

This post is part of a series on the history of computing in sociocultural anthropology.

In his foundational 1955 article “The Structural Study of Myth,” Claude Lévi-Strauss outlined the program for a structuralist, cross-cultural study of mythology. The basic premise is prototypical structural anthropology: to analyze myths, one must decompose them into their constituent units (or “mythemes”). Thus decomposed, hidden mythical patterns can be made evident. These patterns are the real “content” of myths, according to Lévi-Strauss — they persist across different tellings of the same myth, and they reflect the inner structures of the mind. More important for the structuralist project, they recur in different myths, cross-culturally, reflecting the psychic unity of mankind.1

Lévi-Strauss's notecards (1955:435) Lévi-Strauss’s notecards (1955:435)

Materially, such a structural analysis required note cards. With a mytheme on each card, they could be physically rearranged into a two-dimensional grid, with the rows and columns indicating their shared features.2 However, there was a problem. As Lévi-Strauss writes:

At this point it seems unfortunate that, with the limited means at the disposal of French anthropological research, no further advance can be made. [...] A variant of average length needs several hundred cards to be properly analyzed. To discover a suitable pattern of rows and columns for those cards, special devices are needed, consisting of vertical boards about two meters long and one and one-half meters high, where cards can be pigeon-holed and moved at will; in order to build up three-dimensional models enabling one to compare the variants, several such boards are necessary, and this in turn requires a spacious workshop, a kind of commodity particularly unavailable in Western Europe nowadays. (443)

This sudden concern for materiality is striking in contrast to Lévi-Strauss’s focus on the abstract structure of mind. According to him, the progress of structuralism is very literally halted by the size of his office! Compounding the problem of proliferating notecards was the proliferation of analytical dimensions. Two or three dimensions could be grasped intuitively and physically, but beyond that, “progress in comparative mythology depends largely on the cooperation of mathematicians who would undertake to express in symbols multi-dimensional relations which cannot be handled otherwise” (436). Computers offered a kind of conceptual space that could substitute for physical space: “as soon as the frame of reference becomes multi-dimensional [...] the board-system has to be replaced by perforated cards which in turn require I.B.M. equipment” (443).

Although Lévi-Strauss did not use computers as tools, he used them to imagine the future of structuralist analysis — the recovery of deep patterns from expanding corpora of mythological material, arrayed in dimensions that could only be grasped with the aid of computers. The putative ability of computers to handle boundlessly large data sets and multidimensional relationships between symbols reflects Lévi-Strauss’ structural ideal: the incorporation of all mythological material into a single analysis, sorted along all possible axes. By collapsing boards, cards, and multiple dimensions into a single figurative device, the imagined computer allows Lévi-Strauss’ ideal method to exist, in theory.

Edmund Leach on the Computer Edmund Leach on the computer

In his 1961 Rethinking Anthropology, Edmund Leach focused on a different connection between structuralism and computing: their reliance on binary opposition.

If an engineer tries to explain to you how a digital computer works he doesn’t spend his time classifying different kinds of nuts and bolts. He concerns himself with principles, not with things. He writes out his argument as a mathematical equation of the utmost simplicity, somewhat on the lines of: 0 + 1 = 1; 1 + 1 = 10. (6-7)

In Leach’s imagination, the particularities of actual computers do not matter—they provide a system in which unfathomable (or perhaps simply unfathomed) complexity is spun out from the purest of binaries. This style of explanation, Leach argues, is useful for understanding “what goes on in society, how societies work” (6) and the potency of binary codes: “although the information which can be embodied in such codes may be enormously complex, the basic principles on which the computing machines work is very simple” (7).3 Although imagining culture as 1s and 0s may be “frivolous” (7), Leach nonetheless endorses a kind of mathematical formalism:

The merit of putting a statement into an algebraic form is that one letter of the alphabet is as good or bad as any other. Put the same statement into concept language, with words like paternity and filiation stuck in the middle of it, and God help you! (17)

Algebra — the substitution of variables for constants, so as to describe forms and relationships — allows Leach to think that he can avoid the tangles of “concept language,” purifying cultural heterogeneity into fundamental relationships or topologies. Variables, the language of mathematics and computation, provide the figures Leach desperately wants, neither “good or bad,” but purely symbolic — a status notably unavailable to empirical cultural facts, which tend to be tangled up in all sorts of complicating webs.

To use the ultimate anthropological cliché, we might say that for Leach and Lévi-Strauss, computers proved “good to think.” They served not as tools for calculating, but as tools for thinking about mythical, social, or cultural orders. As computers grew in influence and popularity over the course of Leach’s career, he regularly returned to them as models of mind, society, and culture. In his book on Lévi-Strauss (1970), Leach himself suggested that computers functioned totemically for scientists like classes of animals did for “primitive thought”: as symbolic structures people could use to “make sense of the events of daily life by reference to codes composed of things outside themselves” (95).

The structuralist engagement with computing inaugurated a number of enduring themes for the use of computers in anthropology:

  1. Computers frequently function as both tools for studying and metaphors for thinking about objects of anthropological interest, be they minds, social rules, cultural knowledge patterns, or ecological systems.
  2. Computers, figured as indefatigable calculating machines, help to imagine promising futures for labor-intensive formal analyses: large-scale data collection and complex symbolic operations are made more plausible by the assumption that imminent technologies will make them easier.
  3. Computational methods (or in this case, the imagination of them) draw anthropologists into interdisciplinary methodological discussions about science, systematicity, and rigor: for Lévi-Strauss, this was primarily with mathematicians (see Andre Weil’s mathematical appendix to The Elementary Structures of Kinship); for Leach, with software engineers (Leach was fond of tracing his approach to anthropology back to his own training as an engineer).

Without engaging computation practically or materially, Leach and Lévi-Strauss mine it symbolically for homologies. If computers can spin great complexity out of simple binaries, why not culture? If “explanation” can be reduced to a statement of fundamental principles for the computer, then why not for society? The fact that their computer is an imagined one makes the homology all the more powerful: Leach can think that computers are only about 1s and 0s and Lévi-Strauss can imagine them freeing up the space in his office, working with pure and decontextualized symbols, because these computers do not exist. There is no hot room full of engineers, troubleshooting, coding, and constructing the image of purity and unflappable logic that appears in these texts. The computer, for Leach and Lévi-Strauss, is just an object to think with, providing a decontextualized, rigorously symbolic, and tirelessly iterating model of structuralism taken to its logical limits.

  1. Roland Barthes describes the basic actions of structuralism in “The Structuralist Activity” as “dissection,” by which an object is broken up into parts (mythemes, phonemes, themes, etc.) and “articulation,” by which those part are reorganized into relations with each other. Structuralists thus produce a simulacrum of their object, but, Barthes notes, this is an “interested simulacrum,” constructed expressly for the purpose of making certain features or relations thereof more evident than they had been in the original. Robin Horton, in his article “African Traditional Thought and Western Science,” argues similarly for theorizing in general: “All theory breaks up the unitary objects of commonsense into aspects, then places the resulting elements in a wider causal context. That is, it first abstracts and analyses, then re-integrates” (1967:62). For a modern example, see how Netflix analyzes movies
  2. This gridding practice was criticized by many anthropologists as a moment of subjectivity masquerading as objectivity, or as Jack Goody called it, The Domestication of the Savage Mind
  3. See later, Horst and Miller’s argument in Digital Anthropology that this pairing of the simple and the complex is a feature of the fundamentally “dialectic” nature of digitality. 

7 thoughts on “Structuralism: Thinking with Computers

  1. This sudden concern for materiality is striking in contrast to Lévi-Strauss’s focus on the abstract structure of mind.

    I see where you are coming from, but reading about his note card computing was actually what helped me begin to understand his bricolage concept. I’m not an expert in Lévi-Straussian theory, but I think he probably meant physical bricolage as more than a metaphor for a cognitive style. I understand him to have been saying that the two were types of a kind.

    Computers, figured as indefatigable calculating machines, help to imagine promising futures for labor-intensive formal analyses: large-scale data collection and complex symbolic operations are made more plausible by the assumption that imminent technologies will make them easier.

    Madame Héritier made some early applications of that sort to kinship data.

    Héritier, Françoise. 1975. « L’ordinateur et l’étude du fonctionnement matrimonial d’un système omaha. » In Les Domaines de la parenté: filiation, alliance, résidence, ed. by Marc Augé, 95–115. Dossiers africains. Paris: F. Maspero.

  2. Thanks, Matthew. I think you’re absolutely right — the “science of the concrete” was actually quite material, in ways that complicate some of our contemporary received notions about structuralism and the mind. I should have been clearer that I was referring more to that received wisdom than making an argument about structuralism per se. (And thanks for the reference!)

  3. @Nick

    A great beginning to the series. To deepen your exploration of the materiality of the “science of the concrete,” you might also want to consider other metaphors besides computers that L-S used to describe his research. In Tristes Tropiques, he describes walking across a French landscape and imagining the geological processes underlying its topography; then notes the resemblance of his thinking to that of Marx and Freud, who both, like the geologist, seek underlying structures to explain what they see on the surface. Also in Tristes Tropiques, he describes his project as attempting to develop a “Mendelevian table of the mind,” a set of basic elements from which all possible cultural patterns are constructed, as all of the material substances of the world are constructed from molecular combinations of the chemical elements. In the “Overture” to The Raw and the Cooked, in which he directs our attention to “the logic in tangible qualities,” he likens the formation of knowledge to the formation of stars, clusters and galaxies from cosmic dust, noting how first bright spots then larger structures emerge, while their edges and the spaces between them remain obscure. Plainly the computer was only one of several metaphors that stimulated his thinking.

    Also, at the end of the day, the greatest flaw in his work was his reliance on Hegelian logic, which reduces every problem to thesis, antithesis, and a mediating synthesis. In Tristes Tropiques he describes himself has having become so adept at this form of thinking, common in the French philosophy of his schooldays, that he could instantly construct a plausible argument on any subject at all, buses in Paris, for example. Perhaps needless to say, Hegelian logic is a far remove from the logic of the Turing machine that underlies all modern computing.

  4. Thanks, John. For the purposes of this series on computing, I obviously had to narrow my focus, but it’s interesting to note the resonances between various aspects of structuralism and contemporary ideas held by people working on big data knowledge projects and the like. I’ll have to think more on the (dis)connection between Hegelian logic and Turing’s ideas about computing: I suspect the point you raise is quite valid for Lévi-Strauss, who was more interested in computers as devices that might extend analysis (in partnership with mathematicians and the like) than as models for culture. Leach, on the other hand, is quite explicit about his use of binary, Turing-like logic as a model. This is maybe to be expected, given Leach’s interest in statistics (as in Pul Eliya) and mathematization more generally. I don’t know what Leach had to say about Hegel, though.

    One interesting thing that’s emerged from my research, some of which has involved interviewing people who were working in American cognitive and mathematical anthropology in the 70s, is how they think of structuralism as a (non) antecedent to that work. These were people who eventually actually used computers (and who I cover a bit in the new piece posted today), and they basically considered Lévi-Strauss a charlatan, given the subjective nature of his analyses and selection of data. The point is, I guess, that being inspired by computers does not necessarily entail a specific or singular course of action.

  5. I’d just like to reinforce the connection between computation and materiality that’s come up, a point hammered home to me by one of my teachers, David Hays (who was one of the participants in the Wenner-Gren conference uncovered in Nick’s next post in the series. All real computing – as opposed to the abstract thinking about Turing machines and automata – is a material process and so is subject to material constraints set by the number of processing units, amount of memory, and, of course, time.

Leave a Reply