Writing and the Disembodiment of Language

I

AS IS WELL KNOWN, the study of writing in relation to speech played an important part in opening the door to poststructuralist theory, especially in the seminal works of Jacques Derrida. ¹ Taking off from his rereading of Saussurean structuralism, Derrida famously made the deconstructive case that reversed and then dissolved the ancient understanding of writing as the secondary, parasitic representation of the authentic signs of spoken language. But the question may well be asked: How will our understanding of writing and speech change if we take off from a nonrelativistic theory of representation? For it is from a thoroughly relativized theory of signs that poststructuralist theory and its derivatives all one way or the other begin. We will look at this in more detail below, but for now suffice it to say that this kind of theory always takes it that representation precedes the real in a fundamental way. No natural relationship exists between any sign and its referent. Given this fundamental disjunction as a foundation, the various theories lead to the practical work of explaining how signification does manage to occur. For example, meaning is a function of logocentrism in Derrida; desire in Lacanian psychoanalytic theory; discourse in theory after Foucault; ideology in the various new historicisms; and so on. In contrast to these understandings of representation, cognitive linguistics—I will use this broad term to include findings from the related but differentiated subfields of cognitive semantics, cognitive grammar, cognitive rhetoric—claims that at least on a very fundamental level semantics always precedes any actual manifestation of language, which is to say meaning precedes representation. This is because a foundational level of universal concepts arises from biologically determined percepts. According to this view, crucial, general kinds of meaning are a function not of discourse or power but of the evolution of the human organism in the ecosystem of the third planet from the sun. ²

In order to see how the cognitive linguistic arguments run, we will take an example from George Lakoff's Women, Fire and Dangerous Things. Lakoff brings in the work of cognitive anthropologist Eleanor Rosch in order to explain the way categorization functions in our perceptual interaction with the world. Rosch's findings, obtained through empirical studies using standard scientific methods, strongly suggest that human experience in its most general sense is "preconceptually structured" by our innate ability to categorize according to base-level prototypes (p. 270). Humans, even children as young as two years old, have a kind of gestalt cognitive process that tends to be oriented toward certain objects and features of the world, such that these objects and features—called cognitive reference points or prototypes—are universally "known," apart from culture and language. An example of this would be the response to primary colors as opposed to secondary colors. Rosch showed that even members of cultures lacking words for the primary colors found it easier to learn names for those colors than for secondary colors (pp. 40-41). We are, it appears, cognitively attuned to the primary colors. Further, we are cognitively constructed so as to sort colors automatically by gradations away from or closer to the primary colors. Culture will of course teach us certain beliefs about colors, but we do not "learn" to respond and operate on this basic level of perception.

In the next step of the argument, we find that our "basic-level concepts correspond to [this] preconceptual structure" (p. 270). That is, in considering ourselves as linguistic creatures, we find that thinking or conceptuality also operates according to prototypes. Again, this is held to be true only of specifiable basic-level concepts. Though concepts that arise in cultural life may possibly "impose further structuring on what we experience," nonetheless, "basic experiential structures are present regardless of any such imposition of [learned] concepts" (p. 271). Evolutionarily, language would have to emerge after these innate perceptual structures. Language as a representation would also depend heavily on prototype categorization, not...