4: Monitoring Fictional States of Mind
metarepresentation: under advisement and taking into consideration specific circumstances under which it was brought forth.3
My third example comes from Austen's Persuasion. I take as my starting point an observation of the literary critic Ellen R. Belton, who notes that when the novel's protagonist, Captain Wentworth, thinks that he is emphatically not interested in his former fiancee, Anne Elliot, he is, in fact, deceiving both himself and the reader. Belton argues that although it takes some effort for us to see through Wentworth's self-deception, once we do see through it, our attention shifts toward the ultimate source of that incorrect representation, the author herself:
Why, we immediately ask ourselves, does Captain Wentworth insist so strenuously on his not wishing to marry Anne?... It may be that [he] does not know his own mind. The value, therefore, of being allowed inside the mind of such a character is partly an illusion, a device of the author's to make us believe simultaneously that we are being told everything there is to tell and that something crucial is being left out. [Something other than Captain Wentworth's faulty introspection thus has] to provide the missing explanations, to formulate meanings that at this moment are beyond the ability of the characters to understand and, ostensibly, against the will of the author to communicate.
When the reader recognizes this strategy, he becomes not merely an inquirer into the motives and intentions of the characters, but an inquirer into the motives and intentions of the author herself. Like detectives, we find ourselves asking forbidden questions: How does her mind work? Why is she telling us this and not that? What is being withheld?4
I have added emphasis to the parts of Belton's argument that vividly demonstrate how the insight provided by cognitive evolutionary psychologists can productively converge with that of literary critics, who comment on the tacit shift in our interaction with the text, that is, on our heightened attention to the source of representation, once the representation itself has proven less than reliable. Let me recast Belton's analysis in explicit cognitive-evolutionary terms. When we consciously decide that Wentworth's professed indifference toward Anne is really not a "fact" anymore, a certain cognitive adjustment takes place within our system of information management. Wentworth's indifference ceases to be a representation that can migrate with very few restrictions throughout our cognitive architecture, impacting "any other data with which it is capable of interacting."5 It has become a metarepresentation, framed by a source tag along the lines of, "Wentworth believes that. . . ." As a result, we are now moved to take a closer look at the author—the primary source of our information about Wentworth's mistaken beliefs—and ask ourselves what she is trying to achieve here. Is she trying to emphasize certain aspects of Wentworth's character (say, his relative lack of self-awareness) that could partially excuse his present behavior (e.g., his attention to Louisa Musgrove) and thus qualify him as a not infallible but still a suitable partner for Anne?
In other words, the conceptual adjustment that we go through here is similar to the conceptual adjustment that takes place once we learn that the information provided by Eve is wrong (e.g., "Adam is a bad colleague," "It is raining gold outside"). There, we refocus our attention on Eve. Here, we develop a new interest in reevaluating the sources of information about Wentworth's feelings, such as the story's narrator or its author (more about the relationship between the narrator and the author later). Paradoxically, this process of refocusing our attention both constrains and opens up our venues of interpretation. By beginning to treat the representation, "Captain Wentworth is indifferent toward Anne," as a metarepresentation, we "constrain" the scope of inferences we can draw from this representation and thus limit it largely to the possible motivation of our source of information. By doing so, however, we learn to ask new questions about the intentions of the author (such as, how do we account for Austen's emphasis on Wentworth's lack of self-awareness?) and consequently develop new ways of thinking about Persuasion. Our capacity for "monitoring and reestablishing the boundaries within which each representation remains use-ful"6 thus underlies crucially our practices of literary interpretation.
The cognitive-evolutionary research into our ability to "consider the source" does more, however, than just substantiate Belton's insightful reading. Thinking in terms of our metarepresentational capacity allows us to see a pattern behind a series of seemingly unrelated conceptual processes informing our interaction with works of fiction. We begin to recognize that the same cognitive predisposition, that is, our ability to process information under advisement, makes possible both the metamorphosis of the once-proud or -prejudiced protagonists into romantic lovers and the metamorphosis of the formerly trusting readers into "detectives" querying the author's motives.
Furthermore, such an approach allows us to make certain predictions about the "chess game" that could take place between the reader and the writer. The author who gives his or her readers a good reason to doubt a
5: "Fiction" and "History"
representation considered hitherto true in the context of the narrative can reliably expect the reader to start scrutinizing the source of that representation. For example, once the readers of Vladimir Nabokov's Pale Fire realize that Charles Kinbote's account of the country named "Zembla" contains serious contradictions and lapses, they have to begin wondering who Charles Kinbote really is and what about his life and past makes him tell such strange stories. Such a "guarantee" of what the reader will be thinking of and looking for offers the writer the possibility of calibrating, if he is so inclined, the kind and amount of information about the source of representation that the writer sees fit to provide. Nabokov responds to the carefully anticipated shift in the reader's attention by providing cues about Kinbote's real personality, and, although these cues remain maddeningly inconclusive, the author knows that the reader will return to them again and again as Kinbote's chronicles of Zembla become both further divorced from reality and peculiarly reconnected to it. Of course, all of this is done intuitively—neither readers nor writers think in terms of "information stored under advisement"—but the intuition does tend to follow a suggestive pattern. Some writers are particularly fond of choreographing this particular mental game, which may explain, at least in part, the not-altogether-unpleasant feeling that we sometimes get after reading their novels, which is that the writer had somehow anticipated and even planted some of our smart interpretive gambits.
5
FICTION" AND "HISTORY
ere are some questions that must have occurred to you while read
ing my discussion of source-monitoring and that I have actually elided as long as possible to keep my argument under some semblance of control: Aren't works of fiction themselves metarepresentations with source tags pointing to their authors? And, if they are, what allows us to consider any information contained in them, even the apparently incontrovertible, "Lydia ran away with Wickham," a "fact" or an "architectural truth"? Should not we try to envision a much more intricate system of degrees of metarepresentational framing that would allow for such nuances? Thinking through these questions inevitably means raising a series of far-reaching inquiries into the ways cognition structures and is in turn structured by culture. In the remaining sections of Part II, I map out a few parts of this forbidding territory and point briefly to some of its landmarks (a more detailed discussion deserves a separate book).
Commenting on the special metarepresentational status of fictional texts, Cosmides and Tooby observe that it is likely that stories explicitly labeled as fiction (e.g., Little Red Riding Hood) are never stored "without a source tag. The specificity of the source tag may be degraded, from (say) 'Mother told me [story X]' to 'Someone told me [story X],' . . . but . . . insofar as source tags are an important part of a self-monitoring system, one would expect them to be retained in some form as well."1 In other words, works of fiction, at least those clearly defined as such, seem to be metarepresentations par excellence, pere
nnially stored with either variously implicit source tags, such as "folk" in the case of Little Red Riding Hood and as "Anglo-Saxon bard(s)" in the case of Beowulf, or explicit source tags, such as "Jane Austen" in the case of Pride and Prejudice.
We can speculate, then, that it is our awareness that there is a source behind the representation that legitimates a variety of personal and institutional endeavors to resituate, reinterpret, and reweigh every aspect of a literary text. One may feel a bit nervous and yet somehow justified in questioning what so recently seemed to be the widely accepted view—for example, the notion, based on Austen's own direct assertion, that Captain Wentworth does not wish to marry Anne—because as long as we keep track of the source of a representation that involves complex human interaction (or of the mind behind the "gossip," so to speak), no aspect of that representation is completely safe from a reevaluation.
Consider in this context the poststructuralist concept of the "Death of the Author," first advanced by Roland Barthes in 1968, elaborated by Michel Foucault in 1969, and assuming since then a prominent place in literary theory. The concept refers not to the actual demise of the writer (who could be dead or alive at the time of discussion—it does not matter) but to the rejection of the traditional view of the author as the main agency and the "ultimate 'explanation' of a work."2 As Barthes puts it, "[T]he birth of the reader must be at the cost of the death of the Author,"3 meaning that the reader, at liberty to choose whatever interpretation (or interpretations) of the text strikes him or her as most compelling, assumes the position of authority formerly reserved for the writer.
As a literary-theoretical credo (versus, that is, an obituary notice), "the Author is dead" strikes a cultural nerve, and for a good reason. It seems to
5: "Fiction" and "History"
demand a conceptual readjustment peculiarly challenging for our metarepresenting mind: the erasure of the figure of the author calls for some kind of suspension—or deferral—of the process of source-monitoring. It is possible, then, that as a conceptual experiment, "the Author is dead" is exciting because it allows us to consider various implications of such a suspension of source-monitoring even if on some levels this suspension remains unattainable.
For, ingenuous as the concept of the "Death of the Author" appears to be, note its essential cognitive conservatism. The source behind the fictional text is not really eliminated—it is merely substituted by another source. It is the reader who now emerges as an author or one of many authors of the narrative—a game of substitutes testifying to, among other things, the tenaciousness with which we cling to the idea that there must be some source (e.g., an author, a reader, multiple authors, multiple readers) behind a narrative that bears distinct marks of fiction.
Of course the idea that fictional narratives are always stored in a metarepresentational format is useful only as long as it is carefully qualified. Cosmides and Tooby begin such a qualification by pointing out that "the falsity of a fictional world does not extend to all of the elements in it, and useful elements (e.g., Odysseus's exploits suggest that one can prevail against a stronger opponent by cultivating false beliefs) should be identified and routed to various adaptations and knowledge systems." They further assert that the "fact that fiction can move people means that it can serve as input to whatever systems generate human emotions and motivation,"4 which is to say that at least on some level those systems "don't care" if the whole emotionally moving bundle of representations is stored with a source tag identifying it as an "invention" of somebody known as Jane Austen. (I will return to this issue of "not caring" later in my discussion of detective novels.)
But even if on some level we are ready to weep and laugh at a story that we know to be somebody's willful invention, on a different level we can be very sensitive to any attempt on the part of the writer to pass his or her fantasy as a "true" and not a "meta" representation. As Cosmides and Tooby observe, even though "'false' accounts may add to one's store of knowledge about possible social strategies, physical actions, and types of people, in a way that is better than true, accurate, but boring accounts of daily life, [this] does not mean that falsehoods are, other things being equal, preferred. True narratives about relevant people and situations—'urgent news'—will displace stories, until their information is assimilated."5
Moreover, the attempts to pass fictions for such "urgent news" would be decried. Consider, for example, the indignation of the early readers of Daniel Defoe's Robinson Crusoe (1719), who had been promised by the title page of the first edition "a just history of fact" but had then been given what they perceived as a "feign'd" story. Responding to the heated charges of "lying like the truth," Defoe felt compelled to tell his "ill-disposed" critics that his "story is not "feign'd" and "though Allegorical, [it] is also Historical, [because it contains] Matter of real History."6 And, indeed, Defoe's novels contain plenty of information that could be stored either with no scope-limiting source-tagging at all or with a relatively weak tagging (note the importance of introducing the concept of gradation to our discussion of source-tagging; I shall come back to it later). For example, Robinson Crusoe contains information fully compatible with our basic ontological assumptions about causation, naive physics, mental states, and so on, as well as the information compatible with culture-specific semantic knowledge, for example, that eighteenth-century Englishmen engaged in overseas trading, that they used slave labor, that they followed their primogeniture laws. This is not to mention that Robinson Crusoe is also a good source of potentially useful inferences of the kind indicated by Cosmides and Tooby in their "Odysseus" example above, such as that one can survive even in the direst circumstances by exercising resourcefulness and self-reliance, or that wanderlust may come at a serious cost. Strictly speaking, the presence of all this ontologically, semantically, and emotionally true information allowed Defoe to claim that his novels were "true histories" because they contained "matters of fact," and yet his critics felt justified in accusing him of lying and violently discarded his "matters of fact" claim. Their outrage7 seems to indicate a strong conviction that certain representations should be publicly acknowledged as fictions as a whole even if many of their constituent parts satisfy a broad range of truth-value requirements.8
Other epochs and cultural settings offer examples of a similar drive to differentiate the "true stories" from "feign'd," even if the criteria and indeed the very vocabulary of "truth" are different in each case. The famous fourth-century B.C. entry in Chinese Zuozhuan (the commentary on Spring and Autumn Annals, which covered the reigns of the twelve Dukes of the state of Lu from 722 to 491 B.C.) tells of the three historians who chose to be killed, one after another, rather than agree to falsify the account in the same Zuozhuan that recorded the murder of Duke Zhuang of Qi by his chief minister. Although Zuozhuan undoubtedly contains some "falsification of records, precisely to suit those in position of power," its testimony about the heroic historians clearly intends to impress upon the readers that
5: "Fiction" and "History"
the commentary would not sponsor the promulgation of political myths, even if the word myth itself did not exist in ancient China.9
Similarly, a sixth-century B.C. Greek historian, Hecataeus, ridiculed other people's tales as "absurd" and presented "his own accounts .. . as true" (alethes), and in the following generation Thucydides distanced "himself from those whose accounts are 'more suited to entertain the listener than to the truth.'" As being beyond scrutiny (anexelegktos), such stories, wrote Thucydides, "won their way to the mythical" {muthod.es), a term, that, as G. E. R. Lloyd observes, "clearly acquires pejorative undertones" when used "in a collocation associated with unverifiability."10
Finally, as a more familiar example, think of our own bookstores' commitment to carefully demarcating shelves containing fiction from those containing nonfiction, even though the former offer plenty of information that deserves to be assimilated by our cognitive systems as architecturally true, and the latter contain a broad variety of cult
ural fictions (just consider treatises on dating and dieting!).
A cognitive perspective on "fiction" and "history" allows us to qualify the argument sometimes made by my colleagues in literary studies that the notion of "truth" is a relatively recent Western invention and that other epochs and cultural settings do not share our preoccupation with that elusive entity. To support this argument, they typically point out that other peoples' notions of "history" and "fiction" are very different from ours; for example, something that we certainly classify today as a myth could be considered, say, 2,000 years ago, a historical truth about the origins of a nation. What my examples from eighteenth-century England, fourth-century B.C. China, and sixth-century B.C. Greece demonstrate is that although on some level, some of our cognitive systems do not distinguish between actual situations and deliberate fictions—for example, the tears that we shed while reading a touching novel are real enough—on another level, people have always cared deeply about the difference between "true" and "feign'd" stories and were even willing to die for their right to call the myth a myth.
WhyWeReadFiction Page 9