Furthermore, Christopher Frith has suggested that since "selfawareness cannot occur without metarepresentation," that is, the "cognitive mechanism that enables us to be aware of our goals, our intentions, and the intentions of other people," specific "features of schizophrenia might arise from specific abnormalities in metarepresentation."4 The failure to. monitor the source of a representation thus can lead to patients' perceiving "their own thoughts, subvocal speech, or even vocal speech as emanating, not from their own intentions, but from some source that is not under their control," whereas the "inability to monitor willed intentions can lead to delusions of alien control, certain auditory hallucinations, [and] thought insertion."5
For example, the metarepresentation, "I intend to catch the bus," could be perceived by a schizophrenic patient as "Catch the bus," and "My boss wants of me 'you must be on time'" as "you must be on time,"6 thus making the patient experience delusions of control or think that he/she hears disembodied voices talking to or about him/her. The latter, called "a third person hallucination," can result from perceiving a metarepresentation, such as, "Eve believes 'Chris drinks too much,'" as a "free floating notion 'Chris drinks too much,'"7 and so forth.
Note that although people with autism also lack metarepresentational capacity (to the same degree to which they lack Theory of Mind), the above delusions associated with failure of source-monitoring are typical for patients with schizophrenia but not for those with autism. Frith and his colleagues explain it by the "markedly different ages of onset" for autism and schizophrenia. The former manifests itself in the first years of life, whereas the latter usually develops in the early twenties, when the
patient's theory of mind is already in place:
The majority of autistic children fail to develop [Theory of Mind]. They are unaware that other people have different beliefs and intentions from themselves. Even if they manage, with much effort and after a long time, to learn this surprising fact, they will be only able to infer the mental states of others with difficulty and in the simpler cases. As a consequence they cannot develop delusions about the intentions of others. Furthermore, they will know, over a lifetime of experience, that their inferences are likely to be wrong and will therefore be ready to accept the assurance of others as to the true state of affairs.
In contrast, schizophrenic patients know well from past experiences that it is useful and easy to infer the mental states of others. They will go on doing this even when the mechanism no longer works properly. For the first 20 years or so of life the schizophrenic has handled 'theory of mind' problems with ease. Inferring mental states has become routine in many situations and achieved the status of direct perception. If such a system goes wrong, then the patient will continue to "feel" and "know" the truth of such experiences and will not easily accept correction.8
In Sections 8 and 9 below, I focus on fictional protagonists failing to keep track of themselves as sources of their representations of other people's minds and thus "feeling" the truth of their (wrong) mind-attributions. I show that such failures could be used by the authors wishing to tease their readers by making them unsure of what is really going on in the story and which representations originating in the characters' minds they could trust. However, before I get to the narratives that cultivate this kind of conceptual vertigo in their readers, let us consider a more manageable example of a character clearly marked off by the author as mentally unstable.
Fedor Dostoyevski's novels feature many self-deceiving sufferers. Prominent among them, however, is Katerina Ivanovna Marmeladova 0Crime and Punishment), a gentlewoman by birth and education, now a desperately poor widow dying of consumption among her starving children. Katerina Ivanovna repeatedly invents stories that enhance her past and future and immediately starts believing in these fantasies herself, to the raucous delight of cruel onlookers. For example, at the funeral of her alcoholic second husband, she comes up with the idea that she will soon receive a pension for him (which can never happen), and she decides to use that pension to open a boarding school for refined young ladies. Some
2: Metarepresentational Ability and Schizophrenia
of her listeners are simply amused by such ravings, but others, such as her landlady, find her plans as to how to run the school, which county to locate it in, and whom to hire so convincing (for Katerina Ivanovna herself believes in them) that they begin seriously advising her on how to ensure the hygiene and good morals of her pupils (405).
Katerina Ivanovna does not like the thought of accepting advice from her landlady (whom she considers infinitely beneath herself), and she lets it show. The disagreement between the two women escalates into an ugly fight. At this moment, a temporary lodger enters the room, a respectable well-to-do lawyer Petr Petrovich Luzhin. Earlier, Katerina Ivanovna had told everyone that Luzhin was a friend of her first husband, a protege of her father, and the very man who would use his significant connections to secure her the pension (all of which is, of course, her invention). Now Katerina Ivanovna turns to this near-stranger for support:
"Petr Petrovich!" cried she, "at least you protect me! Impress upon that stupid beast that she cannot treat this way a gentlewoman in distress, that there is court and justice .. . I will to the Governor-General. . . She will answer for it.. . In the memory of my father's past hospitality, protect us orphans!"
"Excuse me, Madam .. . I beg your pardon, excuse me, Madam," Petr Petrovich was trying to get past her. "I've never had an honor of meeting your dear father, as you well know yourself. .. beg your pardon, Madam!" (Someone in the room roared with laughter.) "And I have not the least intention to participate in your endless squabbles with [your landlady . . |."
Katerina Ivanovna stood still, unable to move, as if struck by lightning. She could not comprehend how Petr Petrovich could disavow the hospitality of her dear father. Having once invented that hospitality, she now completely believed it herself. . . . (407-8; translation mine)
I have no intention of "diagnosing" the poor Katerina Ivanovna with selective amnesia or schizophrenia, but I do want to point out that her delusions clearly stem from the failure to monitor properly the source of her representations. Katerina Ivanovna's "I wish I could get a pension for my husband" changes to "I get a pension for my husband," and her "I wish this respectable and influential man (i.e., Petr Petrovich) were a friend of my first husband and a protege of my father" registers in her mind as "This respectable and influential man was a friend of my first husband and a protege of my father." Note that because these representations are allowed to circulate freely, that is, without "tags" pointing to herself as their source, in Katerina Ivanovna's mind they produce inferences that can corrupt the already existing stores of knowledge. After all, Katerina Ivanovna's late father had been a socially prominent figure, and Petr Petrovich could have been, in principle, welcomed in his house, if the two men had ever had a chance to meet. What happens here is that Katerina Ivanovna's original memory of her father's house is now corrupted by the conviction that Petr Petrovich used to be a frequent guest there. (Compare it to the hypothetical situation above, in which the information that it is raining gold, when assimilated without a source-specifying tag, such as, "It was Eve who told me," begins to impact our other knowledge stores and results in harmful behavior, such as canceling classes, quitting the job, maxing out on credit cards, etc.).
EVERYDAY FAILURES OF SOURCE-MONITORING
f course, it is not just the hapless Katerina Ivanovna who invents stories about the state of affairs in the world and begins to act upon them as if they were real. We all do it. In many cases, such self-deception is quite beneficial—as one of the more level-headed (or just differently insane) characters from Crime and Punishment observes, "Best lives he who dupes himself the best" (502). But generally, especially if we consider the closely related issue of personal memories, it makes sense to think of our partial failures to keep track of some of the sources of our representations as part of the normal functioning of the metarepresenting b
rain. When I say "normal," I mean to contrast it both with the sustained, pathological pattern of such failures typical for schizophrenic patients and with the deliberately planned and carefully highlighted instances of such failures in the works of fiction.
I was reminded some time ago about everyday failures of our sourcemonitoring—failures that we do not even register consciously unless pressed by circumstances—while reading the account of Martha Stewart's trial in The New Yorker (Stewart had been accused of insider trading and subsequent lying to federal agents). The author, Jeffrey Toobin, refers to a "curious" testimony by one of Stewart's close friends, Mariana Pasternak, who, at one point, could not identify the source of one of her memories:
3: Everyday Failures of Source-Monitoring
Pasternak's appearance ended on a curious note. In her direct testimony, she said that, in another conversation in Mexico, Stewart had commented about [the tip of her broker who had advised her to sell her stocks in the biotech company ImClone]: 'Isn't it nice to have brokers who tell you those things?' But under [the defense lawyer's] cross-examination, she said, 'I do not know if that statement was made by Martha or just was a thought in my mind'—a concession so dramatic that it brought a gasp from the spectators. But then, when the prosecution questioned her again, Pasternak said her 'best belief' was that Stewart said it. (70)
I suspect that the main reason Pasternak's concession "brought a gasp from the spectators" is the charged atmosphere of the courtroom and the specifics of this particular case, in which so much hinged on reconstructing who said exactly what and exactly when. Had any of the "gasping" spectators been asked to trace the exact sources of this or that representation of his, it is likely that he would feel just as uncertain about certain aspects of it as Pasternak did.1
One may ask, then, why we should posit our metarepresentational ability as a special cognitive endowment when it seems that we are routinely unsure about the sources of our representations. The answer to this question applies equally well to the question of why we should posit our Theory of Mind as a very special cognitive adaptation when in fact we routinely misread, misinterpret, and misrepresent other people's states of mind. To adapt one of Ellen Spolsky's insights, both the metarepresentational ability and the Theory of Mind are not "perfect" in some abstract, context-independent sense. Instead, they are "good enough"2 for our everyday functioning: however imperfect and fallible, they still get us through yet another day of social interactions.
Thus, in the example above, the trial witness may have difficulties pinpointing the exact source of her personal memory, but even her apparent failure is thoroughly structured by her metarepresentational ability. That is, she knows that the representation, "Isn't it nice to have brokers who tell you those things?" does not simply describe the state of affairs but also expresses somebody's opinion. Even if she strongly agrees with the truth of this sentiment, on'some level it has still been processed in her mind with a tag limiting its source to two people, either herself or Martha Stewart. The potential for a misattribution or uncertainty (e.g., "Was it really me or Martha?") falls within the same functional range as (to return to the example from Part I) our mistaken interpretation of tears of joy on our friend's face as tears of grief. In the latter case, our range of readings is drastically and productively limited to the domain of emotions; in the former case, Pasternak's range of attribution is drastically and productively limited to two people (as opposed to, say, 150 other people of her acquaintance).
Though not "perfect" (in some rather abstract way), this is surely a "good enough" cognitive scenario, of the kind that we live with daily. Evolution, as Tooby and Cosmides frequently point out, did not have a crystal ball:3 the adaptations that contributed, with statistical reliability, to the survival of the human species for hundreds of thousands of years and thus became part of our permanent cognitive makeup profoundly structure our interaction with the world, but even when they function properly, at no point do they guarantee a smooth sailing through concrete complicated situations or the instinctive knowing of the exact origins of every aspect of our personal memories.
4
MONITORING FICTIONAL STATES OF MIND
owever little we may know at this point about our metarepresenta
tional ability, applying what we do know (or at least hypothesize strongly) to analysis of fiction results in the same embarrassment of riches as does the application of the Theory-of-Mind research. We start realizing that our capacity for storing representations under various degrees of advisement profoundly structures our interaction with literary texts, although, just as with the Theory of Mind, specific historical and cultural circumstances shape the specific forms that such interaction takes. Broadly speaking, whereas our Theory of Mind makes it possible for us to invest literary characters with a potential for a broad array of thoughts, desires, intentions, and feelings and then to look for textual cues that allow us to figure out their states of mind and thus predict their behavior, our metarepresentational ability allows us to discriminate among the streams of information coming at us via all this mind-reading. It allows us to assign differently weighed truth-values to representations originating from different sources (that is, characters, including the narrator) under specific circumstances. The ability to keep track of who thought, wanted, and felt what, and when they thought it, is crucial considering that the majority of our fictional narratives, from Homer's The Iliad, Shikibu's The Tale of Genji, and St. Augustine's Confessions, to Tolstoy's War and Peace and
4: Monitoring Fictional States of Mind
Achebe's Things Fall Apart, center on the characters' reweighing the truth-value of various cultural and personal beliefs.
Consider Austen's Pride and Prejudice. Elizabeth Bennet (and, through her, the reader) can get over her prejudice toward Mr. Darcy because one of the important representations on which she has based her deep dislike of him—Mr. Wickham's account of how Mr. Darcy had mistreated him in the past—is stored in her (and our) mind as a metarepresentation. The agent-specifying source tag, "Mr. Wickham says that. ..," ensures that the information about Mr. Darcy's cruelty and superciliousness is partially restricted from becoming such an integral part of Elizabeth's worldview that no information to the contrary would be able to make any dent in it.
Similarly, Mr. Darcy is able to reconsider his views of himself, Elizabeth, and Elizabeth's sister's feelings toward his friend Mr. Bingley only because he can see these views as metarepresentations: emanating from himself, at a certain time, and for certain reasons (unlike, say, Dostoyevski's delusional Katerina Ivanovna, who is not aware of herself as the source of some of her representations). For example, Darcy used to believe that Elizabeth's sister, Jane, did not love Mr. Bingley and wanted to marry him only for his money and that, furthermore, in marrying any of the Bennet sisters, a man of his own or Mr. Bingley's position would lower himself in the world. These were the sentiments that informed the letter that he sent to Elizabeth shortly after his unsuccessful marriage proposal to her. Later, however, Mr. Darcy is able to assure Elizabeth that the letter was written "in a dreadful bitterness of spirit" that he does not feel anymore; or, to adapt Elizabeth's own apt description of the situation, "the feelings of the person who wrote [that unpleasant] letter .. . are now . . . widely different from what they were then" (248).1 In other words, Darcy has revised his previous views because they have been "stored" in his mind with an agent-specifying source tag, such as, "It was me who felt it," and a time tag, such as, "several months ago, when I was angry at Elizabeth Bennet and mistaken in my earlier representations of Jane Bennet's feelings."
(Our inquiry into the workings of our metarepresentational capacity may also shed a new light on the unpleasant and yet undisputable power of ad hominem arguments. The subconscious appeal of such arguments is a mirror reflection of our tendency to scrutinize the source of representation once the content of representation becomes suspect. Throw a strong a priori doubt on Mr. Wickham's character and see if Elizabeth Ben
net will take his stories about Mr. Darcy's iniquity quite so uncritically, even if she is already predisposed to dislike Mr. Darcy.)
A different example from the same novel: Austen's famous opening sentence, "It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in a want of a wife," derives at least some of its ironic punch from the play between its status both as representation and as metarepresentation. This sentence activates in its readers two rather different information-processing strategies, for it is framed simultaneously as an "architecturally true" statement and a statement to be processed under advisement. On the one hand, the tag phrase, "It is a truth universally acknowledged," literally pressures us to let the idea that "a single man in possession of a good fortune must be in a want of a wife" circulate completely freely among our other knowledge stores, thus influencing our future behaviors in a broad variety of ways (and, we assume, influencing with equal intensity the behavior of the novel's characters). On the other hand, phrases such as, "It is a truth universally acknowledged," or "as everybody says," or "as everybody knows," are generally a peculiar lot, for they also tend to alert us to the possible metarepresentational nature of the information that they introduce. Somewhat paradoxically, they can be easily interpreted as implying an interested source of representation even as they deny that there is one. They seem to hint that somebody wants to manipulate us into doing something that would benefit him or her by having us take a certain precept as a "universal" truth. What if you are a single man in possession of a good fortune, and yet you have no desire for marrying whatsoever? Who is it that wants to coax you into believing that you certainly are "in want of a wife"?
Austen's very next sentence provides an answer to this question, presenting a community of people for whom the idea that a well-off man needs a wife is not a metarepresentation but incontrovertible Truth (a semantic memory, if you will): "However little known the feelings or views of such a man may be on his first entering a neighborhood, this truth is so well fixed in the minds of the surrounding families, that he is considered as the rightful property of some one or other of their daughters" (1; emphasis added).2 The immediately following exchange between Mrs. Bennet and her husband narrows down our suspicions even further: it is the mothers of genteel but poor girls who will benefit if the rich young men of their acquaintance share their own absolute conviction on the subject. Some of the ensuing comedy of the novel is foreshadowed by this outlined-in-the-first-sentence clash between the sensibility that has selfservingly assimilated the idea that a rich man needs a wife desperately and immediately (for many of Mrs. Bennet's antics do result from apparently believing it unconditionally!) and the sensibility that holds this idea as a
WhyWeReadFiction Page 8