James A. Winn
In his challenging paper, Simon notes that the various critical acts involved in attributing meanings "can be more or less reasonable, hence. . . .proper subjects for rational discourse." Near the conclusion of his argument, he even avers that "there is no reason for the Schools [of criticism] to quarrel." But if the larger context of his exposition is "an experiment in communication between the two cultures" delimited by Snow-if Simon truly holds that "it is important for our society that this communication be improved substantially"-then there is more cognitive material to be thrown into the experiment than just the meaning of "meaning"; there is also the meaning of "rational discourse." The role of evidence in selecting among competing meanings is another crucial component of any improvement in the "communication" Simon hopes to enhance. Under such rubrics as "heuristics for decision-making," the topic of evidence has undergone a renascence in cognitive studies. While the original thrust of these researches may have been iconoclastic, emphasizing deviations of American voters or college sophomores from normative rationality, the community is now pursuing a much larger program, seeking to understand how people come to interpret contingent features of the world in the way they do. The relation between "meaning" and "context" that Simon discusses is near the lowest level of a hierarchy at each level of which the knowing agent controls the context in which meaning is to arise. At the simplest level, "control" is the choice of attending to one channel of communication or another. At higher levels we encounter themes from the psychology of rationality: the perception of contradictions or their denial, the organization of empirical perceptions as instantiations of schemata, and the like. Near the top of this ladder of abstraction is the formal awareness of one's own logic, whether Aristotelian, Peircean, computational, or whatever. And somewhere hereabouts is the cognitive mode we commend to Simon's readership as the most promising candidate for the bridging communication in question: the branch of cognitive science, arising as much out of natural philosophy as out of psychology, that engages in the serious study of how people weigh uncertain evidence. There are several famous subspecialties within this field. There is, for instance, the line of research associated with Ludwik Fleck and more latterly Thomas Kuhn, about the genesis of scientific "facts," which strives to replace the normative obsession of the conventional philosophy of science with logical "truth" by a much more psychological notion of rational degree of belief. There is the literature of heuristic decision-making, studies of judgment incorporating perceived probabilities or risks, and the related literature in economics dealing with utility and its paradoxes.
Rather as life usually follows art, the deepest penetration of this old approach into cognitive science is normally associated with studies of visual perception: a rich and wonderful literature about computational and experiential requirements of any system capable of parsing a scene before it. Older than all these branches of cognitive science is the rhetoric originally called "inverse probability" and now, especially by the statisticians, "Bayesian inference." As originally introduced, it is the reasoning about "hypotheses" by reference to their "consequences." Simon refers obliquely to one famous example of Bayesian inference in criticism: Mosteller and Wallace's (1964) study of the authorship of the disputed Federalist papers. Firmly ensconced in the statistical canon, this classic nevertheless failed to establish a school of Bayesian literary criticism. Our goal in this short note is more modest: to note additional applications of the rhetoric of evidence-especially the assessment of conflicting evidence-and suggest their potential role in bridging the cultures.
The "rhetoric" to which we refer is no postmodern invention. Rather, it is a frequent concern in the somewhat oldfashioned but still fruitful modes of literary criticism called biographical criticism and close reading. Even practitioners of more recently fashionable modes of criticism would agree that the "context" (sensu Simon) apposite to any text must include aspects of the poetic art itself. Poets often define their speakers, their subjects, and their audiences by exploiting rhetorical conventions that simulate the gestures of personal relations. Let us consider one brief example, drawn from Dryden's touching elegy for the younger poet John Oldham (1684):
O early ripe! to thy abundant store
What could advancing Age have added more?
It might (what Nature never gives the young)
Have taught the numbers of thy native Tongue.
But Satyr needs not these, and Wit will shine
Through the harsh cadence of a rugged line.
A noble Error, and but seldom made,
When Poets are by too much force betray'd.
Thy generous fruits, though gather'd ere their prime
Still shew'd a quickness; and maturing time
But mellows what we write to the dull sweets of Rime. In a less complex poem, the first couplet, a rhetorical question, would imply an answer in line with conventional flattery: "advancing Age" could have added nothing to the extraordinary and precocious talents of Oldham. But Dryden slyly surprises us by answering otherwise: age and practice might have taught Oldham to write in smoother meter. Such a criticism of the dead person in an elegy sometimes strikes modern students as inappropriate, but flaws are often among the things we miss when a person dies. A Dean in our university once began a eulogy for a recently dead faculty member by observing that the departed "was a lousy poker player," a statement many of us found especially moving, since the generosity and lack of rational calculation that made the dead man an inferior gambler were among his endearing qualities. It is hard to see how a cognitive notion of meaning will ever be able to cope with this sort of gentle wit or irony. Consider also the way that Dryden doubly qualifies his criticism. Before he has even completed his verb ("It might. . . .Have taught"), he parenthetically points out that "Nature never gives the young" skills that can only be earned by practice. Once he has completed his criticism, he goes on to point out that metrical fluency is not necessary in satire, calling Oldham's roughness "A noble error." In the most complex gesture of this passage, he contrasts the "quickness" of Oldham's "generous fruits" with his own smoother verse, which he depreciates as "the dull sweets of Rime." But the passage as a whole demonstrates Dryden's confident mastery of metrical effects. When speaking of Oldham's roughness, he writes with deliberate roughness: "Through the harsh cadence of a rugged line." By adding an extra foot to the line about the "dull sweets of Rime," he enacts in little the effect of "maturing time." And there are many other examples. A reader sensitive to the interplay of syntax, diction, and sound should recognize that Dryden does not really wish us to devalue his own verse by contrast with Oldham's. The rhetorical gesture is a fond one, but is not finally meant to be taken very seriously. To devise a program sensitive to this range of poetic effects, cognitive science would need to broaden radically the context of "meaning" suggested by Simon to include all the features that differentiate the perception of art from mere cognition. But then it would be literary criticism itself, no longer "cognitive science."
The "context" of such criticism includes aspects irreducible to meanings-in-the-world. The effects of the elegy on Oldham are (in part) poetic enactments of Dryden's personality as he contemplates the particular death of John Oldham and the general relations among talent, Nature, time, and the making of poetry. Such devices, central to the pleasure and wisdom we derive from great poetry, can be appreciated only when the critic subjects an agreed-upon corpus of facts (words, their meanings, Oldham's life and times) to an artful sequencing. This is not "meaning" but argument: the critic puts forward his interpretative claim in a manner that, he hopes, cannot be overturned except by the brute discovery of new historical inscriptions (new letters, new ephemera Dryden might have read). Such an art is no different in principle or in practice from what Bruno Latour (Latour, 1987) refers to as the "trial of strength" between two antagonists in a scientific setting: that interpretation triumphs which is capable of mustering the larger number of allies. But it differs considerably from the calm appeal to shared meanings to which Simon believes the critic is limited.
Simon is concerned with the chasm between the cultures; so are we, and with the implications of this continuing estrangement for the health of the academy, the community, and the polity. Toward his claim that "cognitive science can say a great deal about literary criticism," he has applied the cognitive notion of "meaning" to critical arguments. There is an equal amount of enlightenment to be had from the origins of cognitive science in natural science as well as psychological science-pursuing the application of cognitive notions of "argument" to critical arguments. The sense of "wonder" with which Simon closes applies as much to the mystery of logic as to the mystery of perception. In our view, both forms of wonderment would be useful in bridging the chasm of misunderstandings between the arts and the sciences.