Cognitive Approaches to Literature Session, Modern Language Association Convention
New Orleans, 2001

Organizer: Lisa Zunshine, University of Kentucky

"The Cognitive and Anthropological Origins of Narrative"

Richard van Oort, University of California, Irvine

In this paper, I argue that narrative and language are must powerfully and parsimoniously explained by a theory of their coeval origin. A cognitive theory of narrative cannot ultimately be separated from a theory of the origin of language. In order to make this point, I examine two recent discussions of human cognition, Mark Turner's cognitive theory of literary narrative and Terrence Deacon's anthropological theory of language origin.

In his book The Literary Mind, Turner argues that narrative preexists language. For Turner, language provides the vehicle through which we can analyze and theorize the cognitive operations of narrative, projection and parable. Turner is not troubled by the fact that all of his evidence for the cognitive structure of parable is, if not exclusively "literary" in the narrow institutional sense, certainly linguistic, that is, only available in language. Thus, despite the inseparability of language and parable in Turner's analyses, the evolutionary role of language in the production of parable is curiously passed over. Indeed, in his final chapter entitled "Language," Turner makes clear his assumption that language origin is explicable solely in terms of the more basic, nonlinguistic cognitive processes of narrative, projection and parable.

I think this assumption is fundamentally mistaken. The danger of relying too exclusively on the synchronic at the expense of the diachronic is that it overlooks the fact that language itself is, as Deacon points out, an evolutionary anomaly. No other species has evolved language. Surely this fact should give us pause when considering the cognitive origins of narrative. From the diachronic or evolutionary perspective, we certainly cannot accept the inevitability of language origin. If from the vantage point of the synchronic present language acquisition is a highly predictable outcome for each individual, this cannot have been the case for the original language users.

More importantly for my particular argument, however, I think that a fuller understanding of the specific evolutionary and anthropological problem presented by the origin of symbolic representation can help to enrich - and ultimately provide a firmer foundation for - the cognitive model of literary narrative. My intention therefore is certainly not to reject out of hand the pursuit of a cognitive origin for our aesthetic and literary processes. On the contrary, I applaud such an endeavor. My goal is rather to contribute to this same general project by suggesting a more concretely anthropological and evolutionary context for the discussion of the cognitive origins of such basic literary processes as narrative, projection and parable.

But I think a wider view is needed in understanding the cognitive anomaly --which is to say, the anthropological specificity--of literary >processes. I therefore turn to Deacon's book The Symbolic Species, which is an ambitious and, I think, highly successful attempt to address the question of language origin from the perspective of a wide variety of disciplines (including evolutionary anthropology, cognitive science, neuroscience, philosophy, psychology, linguistics, and primatology).

It is of course impossible for me to summarize all the implications of Deacon's wide-ranging discussion. Nor am I qualified to do so. The details of brain research I leave to the neuroscientists. For our specific purposes, however, it is enough to have grasped the basic theoretical model that informs Deacon's account of language origin.

The key to understanding Deacon's account lies in his model of cognition as fundamentally a process of interpreting increasingly more complex layers of reference. In an ingenious reinterpretation of Peirce's categories of icon, index, and symbol, Deacon describes each level of cognition as involving a different mode of reference. Each level requires mastery of the one below it. Thus, for example, indexical reference is not possible without first being able to recognize objects iconically.

Let me summarize very briefly Deacon's triadic hierarchy of levels of >reference. Iconic reference is equivalent to stimulus generalization. It is the default mode of representation, when we recognize one stimulus as similar to another. Indexical reference builds on iconic reference. In this case, the stimulus is generalized not merely at the level of recognition, but is also associated with an interpretive response. This is the classic level of stimulus response theorized by Skinner, and it can generate quite complex forms of representation. For example, vervet monkey calls are indexical signals that have evolved to refer to highly specific categories of objects (e.g., categories of predators, such as leopards, snakes, and eagles).

Though indexical calls, such as those used by vervet monkeys, appear superficially to be equivalent to language (i.e., an arbitrary "sign" "refers" to an object), Deacon argues that such a viewpoint fails to grasp the crucial difference between indexical and symbolic reference. Furthermore, it is only by understanding this crucial difference that we can appreciate the evolutionary anomaly of language and why other species have failed to acquire it.

How is symbolic reference different from indexical reference? Unlike the latter, symbolic reference is not primarily motivated by an association (whether learned or innate) between a sign and an object; rather, external reference is produced only after being threaded through a highly distributed system or network of signs which exist virtually, that is, independently of the contiguous space-time context of the particular object being referred to. This is not the case for indexical reference. In order to be interpreted indexically, an index must be associated either spatially or temporally with its referent. For example, in order for a stimulus, such as a red light, to refer to an object or event, such as the dispensing of food from a chute, the stimulus must be continually associated with the object or event. If the association is not maintained, if for example the red light were to go on over repeated trials without food being dispensed, or if in the case of vervet monkeys all predators were suddenly removed from the monkeys' habitat, then the connection between signal and object would rapidly be forgotten.

This is not the case for symbolic reference. We remember the meaning of words independently of the appearance of their worldly referents, which sometimes may in fact never be present (as in the case of God). This is because symbolic reference is generated by the relations between words rather than by the association between an individual word and its object. Only by understanding the virtual system first can we then agree on its corresponding reference. As Deacon puts it, language possesses a fundamental "duality of reference"; a word refers both to other words and to the world. Indeed, in some cases reference remains wholly fictional. Deacon cites the example of "impossible" or imagined referents, such as, unicorns or angels.

The latter are classic examples of Turner's idea of projection and the blended space. For Turner, projection is the key to parable. Following Deacon, I would argue, furthermore, that projection is not only the key to parable, but to language, and indeed to narrative as well. In this sense, the ability to construct what Turner calls "small spatial stories" is not narrative at all - at least not narrative in the peculiar anthropological sense that doubtless motivates Turner's analyses of literary and ordinary-language narratives. Typical small spatial stories that we perceive independently of language--e.g., moving along a path, eating an apple, drinking from a cup--are instances of cognition that require no higher complexity than indexical representation. To be sure, they are, as Turner assumes, prelinguistic. But they are not specifically human. On the contrary, the level of cognitive representation they require is no more than what we share with the rest of the animal world.

This last point explains a fundamental ambiguity in Turner's discussion of narrative. In the key early chapter entitled, significantly enough, "Human Meaning," Turner argues for the anthropological specificity of his basic concepts of narrative, projection and parable. Yet, as is clear from Deacon's discussion of indexical reference, there is nothing specifically human about recognizing small spatial stories. We duck when we see someone aim to throw an object at us, because we interpret the raised arm as an index of the entire sequence of throwing. But so does a monkey or a bird or a dog. This is not a form of projection in the specific "literary" sense that Turner finds typical of human cognition. It is a form of pattern completion or inference that requires no higher complexity than indexical reference.

The moment of specifically human cognition occurs when a particular image or story is mapped or projected from one context to a totally different context, as when we compare the passage of time to a flowing river or a courageous man to a lion. This form of projection or reference is neither iconic nor indexical. It is symbolic. But as Deacon persuasively shows, symbolic projection is by no means an inevitable consequence of indexical reference. The projection of small spatial stories, upon which Turner would base his cognitive model of "the literary mind," is in evolutionary terms a highly problematic moment, and it cannot be taken for granted. The fact that only humans and a handful of painstakingly trained chimpanzees and bonobos have acquired this cognitive ability points to the evolutionary anomaly of language.

How then are we to explain the moment of symbolic projection? The deferral of indexical reference requires a shift in focus from the immediacy of the association between a sign and its worldly object to the sign as a projection of a "signified" that exists, not in the first place in the world, but in the fictive space of the imagination. By shifting the interpretant's focus from the horizontal reality of small spatial stories to the vertical reality of the blended space, the symbolic sign defers our decision to respond indexically. Instead we can always return to the sign for "another look" in order to generate a new imagined narrative. The focus on the internal form of the sign itself - on the relation between words, rather than the relation between the word and the thing - is the hallmark of aesthetic response. If we are to believe Deacon and Turner, and I think we should, it is also the hallmark of the human. [R.v.O.]

Works cited:

Deacon, Terrence. The Symbolic Species: The Co-Evolution of Language and the Brain. (New York: Norton, 1997).

Turner, Mark. The Literary Mind. (Oxford: Oxford UP, 1996).

Back to Literature, Cognition & the Brain