March 21, 2005

Cargo cult linguistics

In response to my defense of our field from the charge that certain formerly-fashionable French literary theories were "derived from linguistics", several knowledgeable readers wrote in with references to Saussure and Levi-Strauss, quotations about Foucault from Piaget's Structuralism, and so on. OK, we're busted. I have to confess that there is a connection, although as far as I know, none of the French theorists in question ever actually learned anything much about linguistics, or ever actually engaged in linguistic analysis of the sound, form and meaning of words or sentences.

The transmission of influence seems to have been something like this: the anthropologist Claude Levi-Strauss was influenced by the linguist Roman Jakobson during WW II in New York, and after the war, Levi-Strauss in turn brought some of the ideas and terminology of semiotics and structuralism back to Paris. This contributed to the intellectual compost in which thinkers like Barthes, Foucault and Derrida germinated.

There's an excellent explanation of these influences in The Johns Hopkins Guide to Literary Theory and Criticism, which is available on line to subscribing institutions (and for the most part is remarkably readable, in contrast to the material it describes). Here's a bit of the entry on semiotics:

French semiotics, which developed directly from Russian Formalism and Prague structuralism and arrived in Paris via New York thanks to Roman Jakobson's influence on Claude Lévi-Strauss during World War II, made a critical contribution to the study of literary texts during the mid-1960s. A special issue of Communications, edited by Roland Barthes in 1966 and devoted to the structural analysis of narrative, contains articles by the leading European semioticians who had a profound impact on the future and evolution of literary semiotics. In his introduction, which owes a great deal to Louis Hjelmslev's rethinking and development of Saussure's concepts of sign, system, and process, Barthes ascertains that narrative analysis must be based on deductive procedures and must construct hypothetical models patterned on structural linguistics. He proposes a multilevel model of analysis in which each level is in a hierarchical relationship to the others and narrative elements have both distributional (if relations are situated at the same level) and integrative relationships (if situated at different levels). In turn, levels are defined as operations or systems of symbols and rules. Barthes then delimits three linked levels of description--"functions," "actions," and "narration"--in which a function has meaning only within the field of action of an actant, and action is meaningful only when narrated.

And from the entry on Barthes:

... Barthes was eager to promote his French brand of Structuralism for only a few years before he rejected most of its methodological assumptions. [...] His numerous essays and books, written over 25 years from the 1950s to the 1970s (some published posthumously in the 1980s), have taught a whole generation "how to read" (to quote Ezra Pound) and have accompanied that generation through increasingly rapid changes in theory. Besides, even though he retained a set of favorite concepts, Barthes's own swiftness of mind rendered these concepts mobile and capable of important shifts in meaning...

I think it's fair to call this "cargo cult linguistics". Just as some post-war islanders in the South Pacific engaged in ritual imitations of the airstrip activities of foreign armies, in the belief these actions would bring them cargo, so some post-war philosophers in Paris engaged in ritual imitations of the analytic practices of linguists, in the belief that these actions would bring them insight. The islanders carved wooden radio sets and sat mumbling in imitation control towers; the philosophers invented semiotic terminology and sat disputing in Parisian cafes. And just as the failure of cargo to arrive as expected led to social crises and theological reformations in the South Seas, the failure of stable insight to emerge in Paris led to "rapid changes in theory" and to "mobile" concepts expressed in an increasingly opaque style.

More from the JHU semiotics entry:

It is possible to trace two major tendencies in France that evolve from the intellectual activity of the mid 1960s. The first, founded on the Saussurean-Hjelmslevian legacy, best represented by work done by Greimas, has become known as the "Paris school" of semiotics. [... ]

The second tendency is represented by the large number of works that draw their inspiration from a radical questioning of the structural principles defining semiosis. Julia Kristeva and especially Roland Barthes were instrumental in this respect. Indeed, the latter begins his study S/Z by challenging the very possibility of structural analysis to account for the specificity or individuality of any text. He then shifts the problematics from that of science and ideology to that of writing and rewriting, in short to a semiotics of addressers and addressees, of signs and interpreters. In so doing he also substitutes a semiotics of codes for a semiotics of signs and processes and, without structuring or hierarchizing them, determines five codes under which all the textual signifiers can be grouped: hermeneutic (enigma), semic, symbolic, proairetic (actions), and cultural (references to a science or body of knowledge).

And to take things to yet another level, here's a quote from the entry on Derrida:

... one might say that Derrida uses this theory of the sign--a purely diacritical mark, made up of absence, since in language there are "only differences," according to Saussure--to criticize Husserl and Heidegger, while bringing the phenomenological inquiry to bear on the foundations (or lack of foundation) of structuralist scientism, thus forcing two very different traditions to grind ceaselessly against each other.

This seems to be a form of religious syncretism, in which the symbols and rituals of historically different cults are similarly made to "grind ceaselessly against each other".

In this situation, to assign blame (or if you prefer, credit) to linguistics for French literary theory is rather like saying that the clothing preferences of today's urban youth are "derived from prison". There's a historical connection, but we're talking about social gestures that pick up bits of the surface of things and give them a completely new significance as emblems of group identity.


Posted by Mark Liberman at March 21, 2005 09:49 AM