Language and the Structure of Desire

The last time we saw each other, Michel told me, with much kindness and affection, something like, I cannot bear the word desire; even if you use it differently, I cannot keep myself from thinking or living that desire = lack, or that desire is repressed. Michel added, whereas myself, what I call pleasure is perhaps what you call desire; but in any case I need another word than desire

Gilles Deleuze, “Desire and Pleasure.”

Sex is tricky.

Lynn Margulis, Symbiotic Planet

In this article I describe a model for parsing kinship relations which is based on the strategy used in optimality theory grammar as developed by Alan Prince and Paul Smolensky. There is a brief explanation of Prince and Smolensky’s theory for readers not familiar with their work, including how that theory relates to “mainstream” generative grammar in its current “principles-and-parameters” approach, and there are references to an extensive optimality theory electronic archive at Rutgers University. My model is a simple application of optimality theory which does not require any special understanding of the grammar in order to consider how it works: it was “built” with pencil and paper and the only computation it requires is counting no higher than four and crossing out markers in a two column chart. The model is used to do a limited simulation of a theoretical hypothesis:

desire is structured by the conflicting demands of affinity and descent found in kinship systems, and people who live under a given system learn to parse its grammar in the process of becoming “subjects” capable of predicating relationships with other people or objects that fall under the dominion of the system.

(Jackendoff, Nolan)

The question the reader may be asking at this point is the one most difficult to explain: namely, why would anyone want to construct or use such a model? My immediate answer to the first part of that question is the only answer theorists are ever able to give: I built the model because I was fooling around with a piece of theory when some of the pieces started going together this way. Granted, it wasn’t by accident that I was using this particular theory. I had just seen Kubrick’s Eyes Wide Shut at a theater and then watched a video of Atom Egoyan’s The Sweet Hereafter, which I hadn’t seen in theatrical release. I thought I was going to write a movie review and was getting ready to read the literary sources, when I decided to see if I might be able to do something with optimality theory by way of trying to explain the disturbing melancholy that surrounded sexual desire in Kubrick’s film and distinguish it from the equally disturbing yet redemptive figuration of desire in Egoyan’s film, a continuation and variation of the treatment of desire in Exotica.

Fooling around with optimality theory as a way to write a movie review probably seems a bit demented, but it was really a predictable piece of knee-jerk academic behavior perfectly at home in a David Lodge novel. I had talked with Smolensky once about his theory and did a paper a few years ago at an MLA convention using some analogies and speculations drawn from the grammar. Optimality theory starts with the central assumption that the function of a grammar is to provide an acceptable interpretation of speech even though the “surface” of any speech segment is structured by constraints that are “highly conflicting and make sharply contrary claims about the well-formedness in most representations” (Prince and Smolensky, 1993, p. 2).

One way to deal with this problem, the way now favored by the standard “principles-and-parameters” approach taken by Chomsky, is to assume that any conflicts within a particular language can be resolved so that each constraint holds true either in the surface structure or at some level of representation in the chain of derivations that generates the surface structure. This kind of “negotiated conflict resolution” is achieved because actual grammars are allowed local jurisdiction over interpretation by virtue of having the ability to set...