Distant Reading in Lévi-Strauss and Moretti

This is a somewhat revised revision of a post I made at The Valve in November of 2009 under the title, Lévi-Strauss 2: Subject and Object. As the number indicates, it was the second in what would become four posts on my encounter with Lévi-Strauss's structuralism early in my career. The purpose of this revision is to recast that post as an argument that what Lévi-Strauss was doing, in particular in his study of myth in the four volumes of Mythologies, was a species of distant reading.

The term, distant reading, of course, is Franco Moretti's and, I presume, the rationale of invoking distance is to stand it in contrast to the (forms of) close reading so-beloved of the New Critics and many of their successors. What Moretti did in Graphs, Maps, Trees is indeed quite different from close reading of any species. In one chapter large numbers of texts become reduced to two items each, a publication date and a genre label. In another Moretti draws maps of where characters are located and how they move in geographical space. In a third he traces how certain stylistic features move from one text to another without giving any attention to nuances within any given text.

That is quite different from what Lévi-Strauss did with his myths, each of which he recounts in summary form, thus giving each one of them specific detailed representation within his text. He had quite a bit to say about each one of those myths, situating them within kinship structures and their associated behaviors, and commenting upon local geography, flora, and fauna. He was thus, in a sense, quite "close" to those myths. Yet . . .

Consider what Moretti says about distant reading. It is “a specific form of knowledge: fewer elements, hence a sharper sense of their overall interconnection. Shapes, relations, structures. Forms. Models.” Those terms, relations, structures, forms, models, certainly characterize Lévi-Strauss's concerns. His work was a sustained meditation on relations and models.

As for distance, it has its varieties too. Lévi-Strauss was not examining texts from his own culture; he was examining texts from other cultures, cultures to which he was an outsider, distanced if you will, though he did make the odd claim that his accounts of the myths were, as such, yet other retellings of them. He was not reflecting on those texts as a subject within the social world in which those texts circulated. As for Moretti, he did not read, in any ordinary sense, most of the texts that figured as data points in his Graphs chapter. And his readings of (fragments of) other texts were not interpretive within any recognized school; he was not taking them up as a subject and, as a subject, offering reflections to those living in view of those texts. In both cases, then, we have a distance from subjectivity, albeit a distance structured through different means.

With these things in mind, let us move closer to the method Lévi-Strauss employs in The Raw and the Cooked. But let us first look at the method through the eyes of others.

Lévi-Strauss and the Subject

Let us begin with a passage from Derrida's famous essay, “Structure, Sign, and Play in the Discourse of the Human Sciences,” revised from his presentation at the 1966 structuralism symposium held at Johns Hopkins. For example (p. 256 in Macksey and Donato, The Languages of Criticism and the Sciences of Man, 1970): “In effect, what appears most fascinating in this critical search for a new status of the discourse is the stated abandonment of all reference to a center, to a subject, to a privileged reference, to an origin, or to an absolute arché.” On the next page (257): “There is no unity or absolute source of the myth. . . . Everything begins with the structure, the configuration, the relationship. The discourse on this acentric structure, the myth, that is, cannot itself have an absolute subject or an absolute center.” From the subsequent discussion, in response to Jean Hypolite: “. . . this center can be either thought as it was classically, like a creator or being or a fixed and natural place; or also as a deficiency, let’s say; or something which makes possible “free play” . . . and which receives—and this is what we call history—a series of determinations, of signifiers, which have no signifieds [signifiés] finally, which cannot become signifiers except as they begin from this deficiency.”

That is to say, talk and thought of the center, the subject, and the sign are intimately bound together. Eugenio Donato, in his contribution to the symposium (“The Two Languages of Criticism”), notes (p. 94): “It is the possibility of maintaining the discontinuity between the order of the signifier and the order of the signified that permits Lévi-Strauss to avoid dealing with the problem of the individual subject and makes for the extreme rigor of his work”  [emphasis mine, WLB]. Note that phrase, “extreme rigor”.

That’s how things were conceived in the middle of the fray, when structuralism was still unfolding, when it had not settled into the past. Now let’s consider a passage from Maurice Bloch’s obituary for Lévi-Strauss that appeared in The Guardian. He is now reflecting on  those same issues, albeit informally and for a general audience (and, of course, at a temporal distance):

There is also another, even more fundamental, way in which his thought seeks to rejoin that of the mythology of the Amerindians as he understands it to be. Myths have no authors. Their creation occurs imperceptibly in the process of transmission or transformation over hundreds of years and across hundreds of miles. The individual subject, the self-obsessed innovator or artist so dear to much western philosophy, had, therefore, no place for Lévi-Strauss, and indeed repelled him. He saw the glorification of individual creativity as an illusion. As he wrote in Tristes Tropiques: "the I is hateful". This perspective is particularly evident in his study of Amerindian art. This art did not involve the great individualistic self-displays of western art that he abhorred. The Amerindian artist, by contrast, tried to reproduce what others had done and, if he was innovating, he was unaware of the fact. Throughout Lévi-Strauss's work there is a clear aesthetic preference for a creativity that is distributed throughout a population and that does not wear its emotions on its sleeve.

Bloch thus echos Donato's point, that Lévi-Strauss abandons the individual subject. Moretti would seem to do that as well. But when he abstracts texts to publication dates and genre labels, when he follows certain formal practices, from one continent to another (Europe to Latin America) with little concern for the particular texts in which they appear, is he perhaps doing, or gesturing toward, more, even greater distance if you will. Relations, forms, models, what are these things? An abstracted structure sans center?

The System of Myth

Let’s turn to Lévi-Strauss and consider two passages from The Raw and the Cooked, not those problematic passages where he can’t distinguish between himself (and his discourse) and those Amerindian mythographers (and their discourse), but passages where he is deep in his analytic work.

This particular passage is near the end of “Fugue of the Five Senses.” He begins (p. 162): “These transformations are so scrupulously observed that the adoption of a particular point of view implies, in the case of any given myth and any given population, a correlative change in all aspects of the myth belonging to the same community, in which the opposite point of view was expressed.” Notice the phrase “correlative change,” implying that there is an economy to myth such that if you change one or three things here, you get correlative changes there. The passage continues (162-163):

To see this, we have only to compare two Caraja myths, M70 and M85. The first deals with the prospective immortality of humans alone; immortality was denied them because they went from below to above and chose to settle on the surface of the earth where they found fruit and honey (natural products) in abundance, as well as the dead wood which allowed them to light a fire (and cook). On the other hand, M85 contrasts the human condition with that of the animals that slough off their skins. Here the problem is no longer how to prolong life beyond its normal duration, but, as the myth shows, to restore youth to old men. Correlatively there is a descent instead of an ascent (the bird flying earthward); heavenly light is granted instead of earthly fire (which, as the myth is careful to point out, men already possess); and the arts of civilization replace natural resources. As has already been seen, the precondition of prospective immortality in M70 is not to hear; that of retrospective youth in M85 is to hear.

Those final two sentences gives us the correlatives. These two different stories follow equally from the same set of underlying relations. The only way to see that, however, is to compare them with one another and with other more or less closely related stories.

Now let’s consider a passage late in the book, where Lévi-Strauss reaches way back to the beginning to consider the first myth he brought into examination, the so-called key myth (pp. 270-271):

I have proved the existence of a parallel between the animal helpers who intervene in M1 and M124. In so doing, I noted, too, that in each myth a fourth personage comes onto the scene at the end and is not just an animal but a relative: a grandmother in M1, who acts positively by giving the hero a magic stick; an uncle in M124, who acts negatively by killing the alligator with his harmful fluid, for this uncle is a skunk. We observe, then, that the following transformation has taken place:

a) (M1) helpful grandmother (human)→(M124) helpful uncle (animal=skunk).

And since it has also been shown that M1 and M5 are symmetrical, it is not at all surprising that, via the medium of M124, the following transformation can now be proved:

b) (M1) helpful grandmother (human)→(M5) hostile grandmother (human≡skunk).

This having been established, we come to realize that the myth about the origin of diseases, in its two successive episodes, illustrates the two possible way for a woman not to behave like a mother: physically, if she is a grandmother or a woman who has got beyond the age of child-bearing; morally, if she is a young mother whose excessive appetite prompts her to abandon her child. The one kills metonymically by breaking wind (the wind is part of the body); the other by diseases she exudes metaphorically, since she is unable to evacuate the food she has eaten. However different they may be, these two solutions are referable to one and the same principle; if you remove the maternal element from femininity, what is left is stench.

The correlatives are there, and we’ve got some more linguistic terminology (swapped from classical rhetoric), “metonymically” and “metaphorically.” And we’ve got those transformations (one of which has been proved) which I mentioned in the previous post, and two of those quasi-mathematical formulas. Such displays are an important aspect of Lévi-Strauss’ expository rhetoric and the serve two functions. One the one hand, they provide a short-hand representation of correlated bundles of features and allow you to imagine those relationships in some kind of (collective) semantic or conceptual space.

Distant Reading and Objectification

But they also represent an attempt to objectify the system, that is, to turn it into an object out there that one can examine, an object with a finite number of discernible properties. Yes, this objectification is an abandoning of the subject, one that is more far-reaching and certainly more problematic than favoring the collective over the individual. (When they talked of Lévi-Strauss and the sign, I suspect that Derrida and Donato were more concerned about this abandonment, this objectification, than that other.) But the problematic nature of this objectification is not something I wish to take-on here and now. Rather, let’s just bracket it and set it on the self.

Consider the formal models of, e.g. Chomskyian linguistics. They exist unto themselves as logico-mathematical objects. These things can have no material existence except as it is embodied in and maintained by a human population where each individual embodies a slightly different model. Could it be, as I suggested above, that Lévi-Straus's ability to focus on relations, structures, and models was facilitated by his interest in the collective, rather than the individual? For it is only in the interaction among members of the collective that the (acentric) models of individuals are adequated to one another.

It is, however, one thing to construct such a formal system for syntax. It is, that is, it has turned out to be, quite a different matter to construct a formal system for semantics, which is what one would need to write an explicitly formulated grammar of myth, rather than allude to such a grammar, as Lévi-Strauss did in Mythologies—a topic I took up in From Bollocks to Lévi-Strauss on Myth. Chomsky and many others objectified syntax by regarding it as a set of relations among a finite class of items. But, if semantics is a relation between the language system and the world, where can we get finitude? How do we close that system?

Moretti, Lévi-Strauss, and Distance

What then of Moretti's work and Lévi-Strauss's? Both stand at a distance from their texts, and both, I submit, seek to objectify their textual worlds by abstracting relations, forms, structurs, and models from collections of texts. And yet their work is quite different.

And I do not think we can understand that difference by the trope of distance alone. It is not that Moretti, for example, is more distant from the text than Lévi-Strauss, who is, in turn further away than the New Critics or the deconstructionists. If we say that they are distant, but in different directions, well then we've given up on distance alone. What are those directions?

No, they are distant in an (almost) absolute sense. What they do in that detachment, if you will, is somewhat different. Distance enables, but does not dictate.

Let us return to Eugenio Donato. In 1975 he reviewed all four volumes of Mythologiques in Diacritics (vol. 5, no. 3, p. 2): “Lévi-Strauss and the Protocols of Distance.” He tells us that “what follows, then, is not an attempt to describe or evaluate Lévi-Strauss’ contributions to specific areas as much as to question some of the implicit or stated assumptions which Lévi-Strauss himself relates to the ultimate significance of his work.” He goes on to assert that “despite Lévi-Strauss’ repeated protestations to the contrary, the anthropologist is not completely absent from his enterprise.”

No, he is not. Nor is Moretti absent from his somewhat different enterprise. It is a trivial truism that none of us is absent from our enterprises. It is a question of how we construct our enterprises, what the rules are, and how they are shared. I believe that distance is but a means to objectification, a necessary but not a sufficient means.

As for objectification, I do not mean it to imply objective truth, though I believe it a precondition for it. Objectification proceeds through shapes, relations, structures, forms, models. Whether such a model embodies an objectively true characterization of the world, that’s a different issue. The many syntactic models proposed by linguists are objectifications, but it is not at all clear that any of them are objectively true. The same holds for the various formal models that have been proposed for conceptual or semantic structure, in the cognitive sciences.

Let us also remember that objectification must be built on description. Moretti's work and Lévi-Strauss's depend on description. To assign a genre label to a text is to describe it, to find motifs shared by different myths implies the act of describing myths by listing their components. You can't examine a structure if you can't describe it.

But we take description for granted, as though it were obvious and easy. It is NOT easy, nor obvious. Not when describing flora and fauna, and not for cultural texts. But that is another matter, another day, many of them.