index
The Preconceptual Structures of Fictional Narratives
Conferences
Francis F. Steen, Department of English, UC Santa Barbara
A talk given at the Ninth Annual Conference on Literature and Linguistics, North Texas University, Denton, February 6-8, 1997 
 
 

Introduction

"The central issues for cognitive science are the issues of the literary mind," Mark Turner says in his recent book, arguing that the grammatical structure of language is the product of literary modes of processing information: story, projection, and parable.

Most stories, however, never become literature; the vast majority of narrative imaging remains private, or exchanged once in a conversation: the story you form about going to the store, or calling a friend. Only a vanishingly small fraction of all stories thrive and spread as stories, as fictional narratives that have an immediate appeal.

In this talk, I'll address the question of the preconceptual structures that make fictional narratives possible, [and touch on the question of their proper domain.]

A. Computational tasks

When Turner says that "the central issues for cognitive science are the issues of the literary mind," he is making a computational claim: that imaginative narratives solve computational tasks. What are these tasks? Here are some suggestions:
  1. Simulation: access information from several cognitive domains (an information and retrieval system)
  2. Conceptual blending: creatively apply conceptual structures from several source domains onto target domains (an innovative frame-shifting system)
  3. Flowing scenarios: integrate fragmented pieces of information into a coherent narrative (an integrating system)
  4. Behavioral scripts: organize fictionial information into programs of action (an action planning system)
Simulation, conceptual blending, flowing scenarios, and behavioral scripts suggests broadly the kinds of computational tasks imaginative narratives solve: broadly speaking, a kind of metarepresentation, or framed representations.
 

B. Metarepresentations: the framing of representations

So the proposal is that the same cognitive architecture underlies pretend play, mind reading, and storytelling, and that it is characterized by a decoupling of the specific content from primary representation, achieved by pretending to frame the content in terms of an attitude of an agent.

Not all information, however, is equally worthy of being imaginatively narrated, and not all narrative imagining can be called literature.
 
 

C. The epidemiology of representations

I would like to appeal to Dan Sperber's notion of the epidemology of representation. He points out that most stories are never communicated at all, and "most communication remains a local and ephemeral affair" (TLS 27/12/96). But there are stories that are communicated again and again - religious beliefs, fairy tales, poetry. Contrary to what Richard Dawkins and others have claimed, this does not make them true replicators - a story is never simply copied by the listening; it is continually transformed in transmission. Some stories - or elements in stories - prove more resilent than others; they appear to incorporate strange attractors that tend to draw the variations back to a persistent theme. What are some of these attractors? What is it that makes stories interesting?

As the anthropologist Donald Brown has shown - the result is not particularly surprising - all human cultures pretend, understand each other's actions in terms of mental states, and tell stories. Storytelling has a pretty good claim on being an integral part of what it means to be a human being, and thus to belong - parhaps somewhat paradoxically - to the realm of biological meaning-making.
 

Nested realms

As a simple model, think of three realms, rougly nested; each contributes to the total meaning.

Our understanding of the natural is that chemical templates known as genes catalyze cellular reactions in an ordered fashion, leading to the gradual construction of the organism. The genetic templates vary naturally, replicate, and become more or less widespread and common based on natural selection - which in effect act according to functional engineering criteria. Cognitive design being a vital aspect of functional behavior, we have all reason to think this biological meaning-making has contributed to the creation of our cognitive abilities - as Mark Turner puts it in The Literary Mind, the human genotype contributes to "setting up a nervous system that will reach certain target values under experience" (25). Since narrative imagining "appears to be a fundamental target value for the developing human mind," we can ask the question, which problems did our ancestors face that got solved by narrative imagining? To answer this question requires a bit of reverse engineering.

In other words, if the capacity to produce metarepresentations is universal, this is good evidence that this constitutes a target value, and thus that the capacity has been subject to natural selection. Simon Baron-Cohen's work on mind-reading (the ability to explain behavior by attributing epistemic mental states to people) makes a very eloquent argument for considering this a cognitive adaptation - again, not in the sense that it is simply pre-programmed in the genes, but that the genetic templates in a normal environment will with a high degree of reliability reach this adaptive target.

As we have seen, mindreading is a metarepresentational system closely akin to storytelling. The question I would like to raise is whether storytelling can be argued to have a proper domain in itself - a set of problems it is uniquely adapted to solve. One candidate for this is the set of challenges associated with cooperation.
 

First stage: the emotional accounting system

Since life is not a zero-sum game, benefits to both parties can be had if certain conditions are met. The paradigmatic example is the Prisoner's Dilemma: the prisoner has been told he will get a reduced sentence if he informs on his accomplice, while his accomplice will be shot. He assume his accomplice has been offered the same deal. If neither of them talk, however, both will go free. How does trust develop? Under which circumstances should they be cooperative, and under which circumstances defect?

Arguments by Hamilton and computer simulations by Axelrod propose that a stable solution to the problem of cooperation must have two components:

These rules may sound vaguely familiar; in ethics, we know them in a slightly different formulation: The winning strategy in Axelrod's contest is generally referred to as the Tit for Tat rule; however, it is only the second component that somewhat fits the name. The first component is recognizably the explicitly formulated Golden Rule of several ethical traditions. It is also important to further specify the second rule: it is never an instruction simply to turn uncooperative. Rather, it is a strategy with a kindness component and a punishment component.

First of all, it says to mirror the same behavior back to the person; the Old Testament version captures the fact that this mirroring can be deliberately punitive, and may well have to be carried out at a cost. (The human propensity to punish cheaters poses a puzzle to economists working with a model of utility maximization: if individuals were simply looking to their own immediate benefit, nobody should be willing to pay the cost of punishing another for defecting.) Punishing, in game-theoretical terms, is an expense players need to take on in order for the cooperative game to be stable.

Secondly, the second rule is a rule about the next round only. As soon as the other begins to cooperate, the offer should be accepted. Invariant vengefulness is a losing and thus invadable strategy.

If we step back a moment and consider how these mathematical and game-theoretical models may apply to the evolution of cognitive adaptations in human beings, the suggestion is that Axelrod's Cooperation Game is an abstract formulation of a significant adaptive target in human social evolution. Is there evidence for such an adaptation?

The development of an uninvadable strategy of cooperation requires that defectors be penalized to an appropriate degree, so that the average payout for cooperating will be higher than the average payout for defecting. This requires an accounting system, and humans have evolved an accounting system that has a strong motivational emotional component: friendliness, resentment, gratitude, owe, deserve, loyalty, betrayal, and punishment are all mediated emotionally. To get a sense of the rich field of cognitive adaptations devoted to the problem of ensuring cooperation, consider the wealth of linguistic terms that indicate its various stages:

The Semantic Field of Cooperation
 

Indicating intent to cooperate kind, generous, gentle, friendly, helpful, honest, open, warm; trust, faith, confidence 
Indicating the other cooperated  grateful, thankful, appreciative; wanting to return a favor; obligation, indebtedness; to be in someone's good books 
Indicating the other defected  disappointed, bitter; feeling gypped, taken advantage of, exploited; suspicion, betrayal, treason 
Indicating intent to cooperate after defecting  guilty, feeling sorry; remorse, bad conscience 
Indicating a punitive response to the other's defection  resentful, angry, annoyed; to bear a grudge; to ostracize, to punish, to get back at someone 
Indicating a conciliatory response to the other's defection  forgiving, relenting; to try again, to give someone a second chance, to make up; reconciliation 
 

A central condition that must be met for the cooperative solution to emerge is a sustained series of encounters: a condition matched in the environment in which humans evolved.
 
 

Second stage: stories as a manipulable accounting system

Once the emotional accounting system is in place, the task of making a finer set of distinctions arises. Distinctions may be made regarding: Such a large number of potentially relevant variables lend themselves to different weighting schemes and 'spin control'. It is possible we have adaptations that dispose us to be generous when it comes to evaluating our own contribution and stingy when it comes to evaluating the opponent's. At this point, we are already making stories about what happened, running simulations in order to evaluate and remember the balance of dues and obligations.

In order to clarify what is at stake, and cooperate effectively, individuals would need to negotiate these stories-that is, they would need to begin to tell each other stories, and negotiate a common account. This process is open to cheating, and thus to the detection of cheating, or lying.

If several individual tell each other stories, the result would be a network of binary systems of dues and obligations. Stories would thus function to keep track of dues and oblications in a network of individuals, cooperating on an individual and personal basis.

Personal factual stories appear to be a proper domain of stories, and may be the only proper domain.

Third stage: coalitional accounting

If an individual could make a story that implicated the other in one common system of dues and obligations, a coalition would have been created.

One example of this is religious coalitions, which often depend on accounts of ancestors or powerful beings with whom the group is portrayed as having entered into coalitional agreements. Religious literature in turn often consists in attempts to persuade the divinity to become part of the speaker's coalition, to intercede on her behalf.

However, any story that implicates a group of people in a common obligation creates a coalition. This appears to be one of the significant features of an epidemiology of stories: they create a sense of community.

Conclusion

In closing, what must be borne in mind is that most of human history is prehistory. In terms of the natural selection of preconceptual structures, the hundreds of thousands of years of life as hunter-gatherers - in Biblical myth, the Edenic stage - dwarf the brief span of civilization. It is stories that have made us forget this fact-in the Bible, the pre-agricultural stage is brief indeed-but it is also stories that can help us remember. If the target values of a core repertoire of human interests and motivations can be traced back to the central adaptive challenges of our Pleistocene ancestors, stories that have appealed to human beings for centuries constitute the raw material for an archeology of the natural history of the human mind.
 
 
index
 Next: Conferences
Conferences
Cognitive Cultural Studies index
Bibliography
© 1997 Francis F. Steen, Communication Studies, University of California, Los Angeles