And if calculate we must—as, for example, when knowing that one of our pleasant fellow passengers is, in effect, a predator but not knowing which one—we had better have all of our attention focused on the problem at hand. Trying to figure out who among our present company is a murderer involves not only attempting to read the minds of everybody around us but also constantly imagining our behavior from their point of view, for we don't want the criminal to guess that we suspect him/her. Imagine walking leisurely round the really hungry lion, picking up its tail, and casually patting it on the head, all the while pretending that the lion is not even there. Not an ideal situation for analyzing the feelings of one's beloved.
But, one could say, reading about the homicidal maniac is not the same as being actually stalked by him. By the same token, trying to guess together with Austen's Anne Eliot whether Captain Wentworth still loves her is not the same as actually going through such emotional upheavals yourself. It could be very difficult to do both at the same time in real life—to think, that is, of how to outsmart a rapidly approaching murderer while you are figuring out what your beloved really meant yesterday when he said that the weather was particularly friendly for outdoor rambles—but what prevents us from combining these two "activities" in our imagination? Why can't we control our emotions and manipulate
3: Metarepresentationality and the Detective Story
them into "multitasking" by reminding ourselves that we personally are not threatened by the murderer-at-large and we personally are not worried about Captain Wentworth's feelings?
This question dovetails a much larger issue: "how is it that we respond emotionally to literature at all?"16 For a detailed analysis of this issue, I refer the reader to Hogan's two most recent studies.17 For the purpose of the present argument, I want to focus on his observation that our emotional response to fiction is a "matter of trigger perception, concrete imagination, and emotional memory. The issue of fictionality just does not enter."18 Note, incidentally, how well this works with my earlier argument that once we have bracketed off the fictional story as a whole as a metarepresentation with a source tag pointing to its author, we proceed to consider its constituent parts as more or less architecturally true. "To know that something is fictional," Hogan continues, "is to make a judgment that it does not exist. But existence judgments are cortical. They have relatively little to do with our emotional responses to anything. The intensity of emotional response is affected by a number of variables . . . [which] include, for example, proximity and speed, vividness, expectedness and so on."'9
To illustrate Hogan's point about the variables affecting our emotional response to fiction, think of yourself reading the second to last chapter of a murder mystery. You know that a murderer, whose identity is still hidden, is getting closer and closer (the issue of proximity) to the protagonist you have come to associate with. You know how the protagonist feels sitting there trapped in her own creaky house (the issue of vividness) with no phone lines working, and the slightly intoxicated neighbor, who has accidentally wandered in earlier, as her only and clearly inadequate, protection. Then—boom!—the suddenly sobered-up neighbor turns out to be the murderer (the issue of expectedness), and there seems to be no escape for the heroine now.
And when it comes to all these emotion-triggering variables, we have to remember that after tens of thousands of years of collective cultural experience of storytelling, authors have at their disposal a bag of rather effective tricks aimed at emotionally hooking us on whatever mind-reading scenario they are activating. A compelling love story knows how to push your emotional buttons by making you guess and second-guess the characters' states of mind because it is built on the bones of millions of forgotten love stories that didn't. Detective narratives have not been around for so long, but still, given that fewer than one-half of one percent of such narratives published since the nineteenth century have survived in cultural memory,20 we may assume that authors have learned a thing or two about how to keep you on the edge of your seat with guesswork concerning the mental processes of their characters.
It could be, then, that the narrative that attempts to be simultaneously a high-intensity love story (i.e., a love story that keeps us working hard on figuring out the lovers' state of mind) and a high-intensity detective story (i.e., a story that keeps us working hard at figuring the suspected criminal's state of mind) proves "too much" for us, at least in the currently familiar literary format. Just imagine a narrative that skillfully forces you to anxiously keep track of the thoughts of twelve different people (for any of them, or perhaps all of them, as in Christie's novel, could be involved in that deviously arranged murder) and that also forces you to hang with bated breath on every sidelong look of the heroine who apparently does not want to show her rival that she cares that her beloved has read the letter that the rival has written to him five years ago about that conversation that the heroine and the hero had as children in that garden at their eccentric aunt's estate because that letter implies that the rival is much more emotionally suited to the hero than the romantic heroine herself is, and so on. Clearly, something's got to give. Hence, we have successful detective stories with some romantic elements, but the metarepresentational framing needed to process those romantic elements is carefully calibrated so as not to compete with the metarepresentational framing required to process the detecting elements of the story. Conversely, we have compelling romances with elements of detection, but the metarepresentational framing of the detecting elements is skillfully subdued so as to add some extra level of mind-reading to the story without making it compete with the main type of mind-reading expected from its readers.
Of course, in its present embryonic state, a "cognitive-literary" perspective may not be able to explain why certain combinations of different kinds of mind-reading in the story are more felicitous than others. Still, it points us to the areas of cognitive research to watch. If, at least on some level, the narrative focusing on a romance and the narrative focusing on the detection of the murder may appeal to differently specialized adaptations within our Theory of Mind module (e.g., the one evolved to facilitate mating and the one evolved to facilitate predator avoidance), then the narrative that combines the two by demanding an equally high emotional attendance both to the romance and to the detection of murder overloads some of our attention-focusing and information-processing systems.
Literary history can be thus viewed as a continuous experimentation with recombining metarepresentational units that used to feel overwhelming for our representation-hungry brain-mind but that have come
3: Metarepresentationality and the Detective Story
to feel pleasurable in new, hitherto-unexpected, ways. Hybrid genres emerge all the time as a testimony to this experimental endeavor.21 Who knows?—in five hundred years, we may have a genre of murder mystery/romance/family chronicle that will hit our Theory-of-Mind "spots" in all the right ways and feel as "natural" as a "pure" detective story feels today. In fact, I would say that because the combination of the equally emotionally engaging detective plot and romance plot remains so challenging today, we have a "guarantee" of sorts that writers will continue experimenting in the direction of integrating the two. The culturally embedded cognitive "limits" (i.e., the limits that became apparent only because of certain paths taken by literary history) thus present us with creative openings rather than with a promise of stagnation and endless replication of the established forms.
Meanwhile, let us take a closer look at the detective mysteries that have indeed incorporated romance into their main plot of detection. First of all, it seems that many writers have learned to skirt the issue altogether by either having their detectives go through regular and not particularly involving love affairs or by keeping them married. Both casual affairs and marriage require a minimal amount of metarepresentational framing involved in figuring out the romantic partner's state of mind.
Thus the thirty-something female detective has a reasonably clear idea of what a
college student who ogles her at a party is thinking (as in James's Unsuitable Job for a Woman). Similarly, the newly married, hunky, but unfortunately swamped-with-work Marlowe knows exactly what his rich and idle wife really wants (as in Raymond Chandler and Robert B. Parker's Poodle Springs). Casual affairs and married states are good for the detective story because they let us focus our mind-reading energies on figuring out the crime and suspecting everybody, while still making us appreciate that all-human aspect of the investigator's personality. It really is a neat narrative trick. No modern-day Dashiel Hammett would be able to quiz Sara Paretsky or Sue Grafton on the subject of their heroines' unnatural celibacy, for, look: V. I. Warshawsky and Kinsey Millhone hop in bed with a different man in nearly every novel or else reminisce about their recent affairs.
On the far opposite end of the spectrum are the detective story writers who overinvest in the romance, an instructive sight. Sarah Dreher's novels feature a shy travel agent, Stoner McTavish, whose emotional energies are focused on winning the heart of the enchanting Gwen and who prevents murders as a way of deepening her relationship with Gwen. The compelling romance part of the story leaves very little room for guessing the states of mind of potential criminals: Gwen is a delightful mystery, but we can easily figure out what the baddies are thinking.
Ironically, although I am not sure if there is any causal connection here, the "real," challenging, metarepresentationally framed mind-reading that we expect from the detective story is substituted in one of Dreher's novels by plain old telepathy. I stopped considering Something Shady a detective narrative after the heroine, locked in the room by the criminals who were getting ready to kill her, apparently sent a psychic signal to her aunt in a different city, and the aunt started calling the baddies' enclave, thus nearly distracting them for a while from their evil designs. (Like Peter Rabinowitz, I feel "entitled to assume that the supernatural cannot intrude" in the detective narrative.22)
Dreher's overinvestment in romantic mind-reading at the expense of "detective" mind-reading provides an illuminating contrast to Allingham's Sweet Danger, The Fashion in Shrouds, and Traitor's Purse, and Sayers's Gaudy Night. Those four novels were just as ambitious in their attempt to break the celibate mold of the murder mystery, but they succeeded where the "Stoner McTavish" series fails,23 and here is why: the "relationship" plots in Allingham and Sayers are engaging enough, but they are skillfully underemotionalized compared to the gripping detective plots of each novel. The mutual attraction of Amanda Fitton and Albert Campion is cute, but, for some reason, we are just as happy to keep their romance unresolved until the next published installment of Campion's adventures, whenever it comes, as are Amanda and Albert themselves. Similarly, in the case of Gaudy Night, we understand early on that Harriet Vane either will marry Lord Peter Wimsey after a requisite amount of soul-searching or will not, but we don't particularly care anyway. By contrast, we do care about the identity of the increasingly violent college malefactor, and we dutifully begin to suspect every innocent middle-aged professor whom Sayers grooms for that wicked role.
In other words, both Allingham and Sayers "advertise" their stories as detective narratives with a strong element of romance by increasing, in particular, the amount of time they spend talking about their protagonists' love interests. Because, however, they do not build in a strong metarepresentational framing for this aspect of the story, that is, they do not make us guess, second-guess, misread, and then head-slappingly correct our misreading of, the characters' romantic feelings, the romance remains a tame "junior partner"24 to their main business of detection.
Hammett's The Maltese Falcon is an important example of yet a different strategy. Brigid O'Shaughnessy, the woman with whom the private
3: Metarepresentationality and the Detective Story
FIGURE 4. "What else is there that I can buy you with?" Sam Spade and Brigid O'Shaughnessy before Sam finds out that she killed Archer.
FIGURE 5. "When one in your organization gets killed, it is a bad business to let the killer get away with it—bad all around, bad for every detective everywhere." Sam and Brigid after he realizes that she killed Archer.
investigator Sam Spade falls in love, is one of the suspects in the case of the murder of his partner, Miles Archer. The criminal and the romantic aspects of the novel are so intertwined that if Brigid is concealing the truth about her role in the killing of Archer, it means she is lying to Sam about her feelings for him, for had she really loved him, she would not have kept him in the dark about the story of the crime (figures 4 and 5). The romantic mind-reading thus nearly completely overlaps with the mind-reading oriented toward the detection of the crime.25
Here is what is particularly interesting about this frugal "two-for-one" scenario. On the one hand, I have argued above that because, at least on some level, the romance plot and the detection plot "feed" their respective information into different adaptations within our Theory of Mind module (i.e., the mind-reading adaptation geared toward mate selection and the mind-reading adaptation geared toward predator avoidance), writers may generally have a difficult time when they want to combine the two plots so as to give them an equal emotional weight within the story. Hammett, however, seems to have circumvented this difficulty by merging the two plots into one. To understand some of the emotional effects of such a merger, think again of my earlier examples of the cultural images of "maneater" and "ladykiller" that emphasize the danger of falling in love with a predator. The detective story in which the investigator's love interest is also one of the suspects exploits the suggestive cognitive ambiguity of such a situation. Such a story derives titillating emotional mileage from making the readers mix the inferences from the mate-selection aspect of mind-reading with inferences from the predator-avoidance aspect of mind-reading.
Misreading the mind of the predator by approaching him/her with the view of romantic relationship may result in a personal disaster, as so happens in Hammett's Maltese Falcon, Paretsky's Bitter Medicine, and Hitchcock's Vertigo. On the other hand, the love interest may turn out to have been unjustly suspected of predatory tendencies (as is Vivian Sternwood in the Hollywood version of Chandler's The Big Sleep or Linda Loring in the original The Long Goodbye). In a very mild variation on the two-forone scenario, Karen ¥>)cws)tis Alley KatBlues, the policeman-boyfriend of the female investigator, Kat Colorado, gets entangled with a woman implicated in the crime that Kat is trying to solve. By solving the murder case, Kat thus also gets to figure out the feelings of her boyfriend who has been acting strangely lately. Alley Kat Blues is significantly more invested in romance than many of the detective novels discussed above (though it does not approach the level of Stoner McTavish or Something Shady), and
4: Always Historicize!
it pulls it off precisely by creating a situation in which the mind-reading oriented toward solving the crime overlaps with the mind-reading oriented toward figuring out the feelings of the romantic partner. In other words, unless used too often and thus rendered predictable, the conveniently economic focusing of the two different kinds of mind-reading on one person can work for writers who are intent on opening up that "very tight little box" of the classic detective story.
Note that by construing a spectrum, on the one end of which there are detective stories with the minimum of romance and, on the other, the stories in which romance overwhelms detection, with a variety of other narratives gravitating toward either end of the spectrum, I do not intend to pronounce on the aesthetic value of any of these books. Nor do I make any sort of prediction about their chances for survival in what Franco Moretti calls the "slaughterhouse" of the long-term literary market. Instead, I suggest that by articulating the difference between the love story and the detective story in terms of the dominant type of mind-reading required from the reader in each case,26 we can reground and systematize our intuitions about the differences between the two genres. In general, we can now start thinking of our concept of the literary genre as reflecting, at least on some l
evel, our intuitive awareness that even though all fictional narratives rely on and tease our Theory of Mind, some narratives engage to a higher degree one cluster of cognitive adaptations associated with our ToM than another cluster of such adaptations.
j~4~
A COGNITIVE EVOLUTIONARY PERSPECTIVE:
ALWAYS HISTORICIZE!
P o what have I really been saying by insisting on grounding our enjoyed ment of detective stories in the workings of our metarepresentational ability and our Theory of Mind? A quarter-century ago, in his influential Adventure, Mystery, and Romance, John Cawelti cautioned literary critics about the dangers of assuming that the process of writing and reading fiction is "dependent, contingent, or a mere reflection of other more basic social and psychological processes."1 Have I recruited research from the currently fashionable field of cognitive science to smuggle in the same old fallacy of explaining away a complex cultural artifact as a mere reflection of a basic psychological process? Have I tumbled headlong into the pit that Cawelti warned us about?
If I have, my response is to dig myself further in. I shall start by quoting more of Cawelti's argument:
In the present state of our knowledge, it seems more reasonable to treat social and psychological factors not as single determinant causes of literary expression but as elements in a complex process that limits in various ways the complete autonomy of art. In making cultural interpretations of literary patterns, we should consider them not as simple reflections of social ideologies or psychological needs but as instances of a relatively autonomous mode of behavior that is involved in a complex dialectic with other aspects of human life.2
WhyWeReadFiction Page 20