The task of bridging the "two cultures" of science and humanities that Herbert Simon is pursuing in "Literary Criticism: A Cognitive Approach" is important and timely. I believe that the great changes not only in literary studies but in many of the other traditional disciplines that have occurred in the last decades in large part call for establishing "bridges," rethinking disciplinary borders, and discovering modes of collaboration among people whose "meanings" and "processes" (key terms in Simon's essay) are very different. Reading this essay made me want to know Simon better, to read the essay on information processing and gestalt phenomena he mentions (a text, alas, I should know), to visit with him personally in Pittsburgh. Such bridges, though, are not as easy as this essay suggests-even granting Simon's claims of cognitive science's "imperialist" possession of "the common experiences of being human." (After all, we in literary criticism and theory like to think of ourselves as the great imperialists.) As Donald Davidson said some years ago, "it is one thing for developments in one field to affect changes in a related field, and another thing for knowledge gained in one area to constitute knowledge of another." If this is true-and I will argue that surely it is some of the time-then constructions of bridges, even if they are built as a beachhead for benevolent imperialism, are difficult indeed.
There are three forms of this difficulty that Simon's essay presents, and their discussion might both lead to further study and also underline the problems of the kinds of interdisciplinary work that I join Simon in thinking is necessary. The first is what George Steiner (1978) calls a "contingent" difficulty, born of (interdisciplinary) ignorance. In the case of this essay, what I miss most in its description of "contemporary cognitive science" is the inclusion of semiotics. Semiotics, unlike story grammars, systematically pursues the analyses of meaning beyond the confines of sentences, and like cognitive science it assumes the canons of scientific truth that Simon claims for "expository prose": correspondence to empirical evidence ("veridicality"), generalizing self-consistency ("clarity, absence of ambiguity"), and simplicity. Familiarity with semiotics would allow Simon to see how his operational definition of
meaning could benefit from the work of Charles Sanders Peirce. More importantly, it would have created the necessity of defining "process" more precisely than the essay does. Time and again, what is called process seems to me simply to describe serial accidents. In one example, the "recognition" of a face produces access to more or less accidental memories. But the issue of how the gestalt of "face" is apprehended as "meaning" is never addressed. The work of A. J. Greimas and Paris semiotics might help formalize more rigorously than the vague associationism presented here the processes of recognition, just as it could formalize what Simon calls the "symbolic encoding" of images (Cf. Greimas, 1987).
A second difficulty of interdisciplinary work is, in Steiner's terms, "modal." In the "cognitive science interpretation" of literary criticism presented here, criticism itself is described in the mode of cognitive description. That mode follows the veridicality, generalizing clarity, and simplicity of science. The result of this is to reify or, as Greimas says, to "substantify" activities in order to measure or compare them. But the mode of much literary criticism is not so much measurement and comparison aiming at simplification and generalization but, as Gaston Bachelard (1984) says of the "new" science of the twentieth century, the "complexification of what appeared to be simple." This is to say, criticism studies rhetoric, and the mode of that study is not to reduce differences to the same, but to study the complex interplay of differences. In a rhetorical analysis of Simon's expository prose, one could profitably pursue the figures of speech that substantify processes-in phrases such as "the meaning we inject into a text," the "extraction" of meaning, the "structure" of memory. The rhetoric of the essay, like the definition of representation in Simon's footnote 3 that reduces the process of representation to a referential index, does not acknowledge the different mode of understanding-the discovery of a pattern of meaning at odds with the professed meaning, in this case the professed search for "processes," not things-that rhetoric pursues.
A last difficulty of crossing the "related" disciplines of cognitive science and literary criticism that I want to mention is close to what Steiner calls "ontological" difficulties. Such difficulties, he says, "confront us with blank questions about the nature of human speech [and] about the status of significance." That is, they question the very veridicality, clarity, and simplicity of meaning altogether. I've already touched on such "ontological" difficulties in discussing the contingent and modal problems for interdisciplinary work. But it is most clear, in Simon's argument, in the assumption of "cause" and "subjectivity" as essential components of cognition and understanding. Throughout the essay cognition is seen to be the effect of particular "causes": the "perfectly definite processes that can be executed by mechanism," the subjective "ownership" and accidental "plurality" of meaning, a solely referential conception of representation. The semiotic and rhetorical interpretations of criticism (as opposed to biographical and philological readings Simon seems to use as models of criticism) replace such mechanical "cause and effect" explanations with "means and ends" as the most fruitful mode of explanation, just as semiotics, psychoanalysis, and even relativity (where defining "causes" are replaced by "operational definitions") do. In all of this work, "texts" are not simply things into which meanings can be "injected" or clearly separable from their "contexts" or empirically caused-and "owned"-by their authors. Moreover, in such work the subject of knowledge and cognition is not always simple, empirical, and clear.
In this work of criticism, text, author, and such methodological assumptions as "cause" and "ownership" take on different meanings in different contexts, so that they are not always simple, verifiable, generalizable. Contemporary literary studies, among other things, is calling these Enlightenment concepts and methods into question, not simply to reject them-they all do important work and can be understood, in particular contexts, in terms of precise operational definitions. They are calling them into question, I think, in order to temporalize them to show that sometimes they work, just as sometimes cognitive science is, indeed, the master of literary studies while at other times-when the question of "games" come up, for instance, or of semiotic "encodings"-rhetoric might well "master" cognitive science. Werner Heisenberg (Heisenberg, 1958) describes this situation in terms of "alternative" explanations: complementarity, he notes, encourages physicists "to apply alternatively different classical concepts which would lead to contradictions if used simultaneously."
The alternative application of discipline-bound concepts might well be a method and mode of interdisciplinary studies. It is, I think, a version of what we do when we engage in fruitful conversation, when we encounter new ways of apprehending phenomena, and even when, as Simon does, we try to imagine what the eighteenth-century patricians who wrote our constitution might say today. Such alternatives allow us to suspend the differences between cognitive science and semiotics by focusing on "symbolic encodings," between literary and expository prose by pursuing rhetorical analyses, and between the grand successes and the critical failures of the Enlightenment mechanical world view in our century by studying how they worked and what they accomplished. Such alternatives, like the use of analogies from related disciplines, are neither (wholly) simple nor verifiable nor unambiguous. Rather, they are time-bound so that retrospectively, as Heisenberg says of quantum mechanics, processes and things-even if they are fraught with complexity and difficulty-can prove to be simply, if momentarily, true.