Literature, the Emotions, and Learning Noël Carroll

Philosophy and Literature, Volume 44, Number 1, April 2020, pp. 1-18 (Article)

Published by Johns Hopkins University Press DOI: https://doi.org/10.1353/phl.2020.0000

For additional information about this article https://muse.jhu.edu/article/754505

[ This content has been declared free to read by the pubisher during the COVID-19 pandemic. ] Noël Carroll

LITERATURE, THE EMOTIONS, AND LEARNING

Abstract. This essay is an exploration into the ways in which literature can function as a resource for emotional learning. After rejecting the identification model of emotional education, an alternative model, employing the notion of criterial prefocusing, is introduced. I then go on to explain how this approach to emotional education can not only account for the reinforcement of existing emotional tendencies but can also change social sentiment.

I he subject of this essay is the way in which literature, by engaging Tour emotions, contributes to our emotional intelligence. In reading works of literature, we are almost constantly called upon—or mandated— to mobilize our emotions in the process of understanding the text. In this way, the literary text ineludibly guides us through a rehearsal of the pertinent portions of our affective repertoire. For example, we do not fully understand Iago unless we despise him, nor do we understand Dorothea Brooke adequately without being sad- dened by the way in which she is entrapped in her marriage to that stick, Casaubon—that is, unless we lament “what a waste.” Moreover, in order to be absorbed in a plot, we need to align ourselves emotionally with the protagonist so that we feel suspense when his or her goals are imperiled and joy when they are achieved. Literature has the capacity to put our emotions through their paces and to refine them by confronting us with diverse characters and

Philosophy and Literature, 2020, 44: 1–18. © 2020 Johns Hopkins University Press. 2 Philosophy and Literature situations, thereby training us emotionally. That training will be the thesis of this paper. However, before pursuing the topic any further, a few caveats are in order. First, although my focus is on emotional learning, I do not mean to suggest that this is the only sort of learning available through read- ing literature. One may acquire propositional learning from literature both of particular propositions—such as that “Napoleon was driven from Moscow” from War and Peace—and of general propositions—such as that “love is chemical” from A Midsummer Night’s Dream.1 Furthermore, we may not only derive propositional beliefs—beliefs that such and such—from literature. We can also acquire know-how. From Lady Murasaki Shikibu’s Tales of Genji, the aspiring Japanese aristocrat could have learned how to behave in the Heian court—that is, could have learned its rules and etiquette, including even how to conduct a seduction properly. And, of course, we all have Bram Stoker to thank for teaching us how to dispatch vampires. So the kind of learning from literature that I discuss in this essay is not the only kind there is. Indeed, it is not even the only kind of emotional learning on offer in literary works. In addition to sharpening our emo- tional intelligence by giving it a workout in the process of understanding (and evaluating) characters and scenes, literature also affords us the opportunity to enter the minds of characters as they both experience and reflect upon their emotions, desires, goals, motives, options, and so forth. In this way, a literary work can provide us with models that we can, so to speak, observe and thereby imitate when we consider our inner affective lives.2 In short, there are many modes of literary learning other than the one that I am primarily exploring in this essay. Another disclaimer that I need to signal loudly is that literature is not the only cultural practice that promotes emotional learning, including the sort of emotional learning upon which I am focusing. For not only are our emotional skills honed by literature but also obviously by our caregivers’ direct instruction about what to feel in respect to what, by our teachers and our peers, and also by social institutions including reli- gion, sports, and politics. This should not be surprising, for emotional education, as I will show, is indispensable for cultural survival. Thus, it makes sense that a culture will have redundant channels supplying such an education in order to assure that the job gets done. That is why we have a network of back-up resources guaranteeing that emotional instruction will be delivered one way or another. Noël Carroll 3

What the preceding caveats amount to—in philosophical jargon—is the admission that the sort of emotional learning being examined in this essay is neither necessary nor sufficient for the learning that literary works afford. It is not the only kind of learning available from literature, nor is it a kind of learning unique to literature. Indeed, it may not even figure in some literary works at all. And although the discovery of neces- sary and sufficient conditions for the object of study of a philosophical analysis is typically the holy grail of such pursuits, it strikes me in this case as an impediment to understanding the complex ways in which literature functions as a multitasking, cultural practice.

II A culture—by which I mean (admittedly vaguely) an ensemble of customs, practices, rituals, traditions, norms, memories, myths, ideals, narratives, examples (and probably much else) around which a people coheres, recognizes, and identifies itself—must reproduce itself in order to survive. That is, not only must it physically reproduce its population from generation to generation, along with the means of supporting itself materially. It must also reproduce the next generation, in large measure, in its own image if the people in question are to continue as a cohesive group. For, if the group is to persist, it needs, in the main, to be coordinated in terms of the expectations and judgments we make of each other and of ourselves. There must be a high degree of conver- gence in the group concerning what people think, even if the culture in question sustains some perennially contested issues. Indeed, in some cases, the transmission of those perennially contested issues may be a significant element in the group’s identity. But indisputably, cultural perseverance over time requires shared beliefs that such and such, including shared ideals and histories, and also shared skills, especially social skills—that is, knowledge of how to behave in certain recurring public and interpersonal situations. However, in addition to learning what to believe and how to act, the member of a culture must be educated, as Roger Scruton puts it, in what to feel.3 That is, in order to navigate her culture successfully, one must be able to comprehend what it is fitting or appropriate to feel in a wide range of circumstances, such as: that grief is typically how one feels upon the occasion of loss; that anger is how one feels in the presence of injustice; and so on. Moreover, our emotional intelligence also extends to the ability to discern what other members of the group are feeling. The 4 Philosophy and Literature emotions are, among other things, a means of communication whose significance we must learn to decipher in order to join the conversation of our culture. One needs to sense another’s anxiety in order to reach out to her as well as the other’s contempt so as to protect oneself. Thus, in sum, in order to be a fully functioning member of my culture, I need to share with my fellow citizens an enormous number of beliefs that such and such is so, as well as a large body of behavioral schemas regarding how to act, in addition to an emotional repertoire governing what to feel in the appropriate circumstances. In this regard, literature is an immense cultural resource for transmitting each of these sorts of cultural information, including insight about how believing that such and such, and knowing how to act and what to feel, can fit together and interact. But for now, I just want to focus upon the way in which literary texts can contribute to learning what to feel.

III Interestingly for my purposes, one of the earliest, if not the earliest, discussions in the Western tradition of the education of the emotions occurs with reference to literature. What I have in mind are books 2 and 3 of Plato’s Republic, in which Plato was concerned with the education of the prospective guardians, the young people who eventually would become the rulers of the ideal state or republic that Plato envisioned. Plato was particularly concerned with the emotional effect the literature that the fledging guardians were exposed to would have upon them. And in the course of pondering their curriculum, he developed the first theory of the affective impact of literature, a theory that would shape thinking about our emotional engagement with literature for centuries thereafter. That theory is rooted in the notion of identification, the idea that in responding to stories we identify with the characters in the text—that is, we take on their emotions. In ancient Greece, students learned to read by reciting from classic texts like Homer’s Iliad and Odyssey. In doing this they would speak out loud the lines of various characters. Plato feared that when they repeat- edly read the lines of certain characters—voicing emotions of which he, Plato, disapproved—the students would become infected with those emotions and, thus accustomed to them, would be disposed to indulge them in their own lives as the occasion arose. For example, in reading book 11 of the Odyssey, a young person might read Achilles’s lament: Noël Carroll 5

Glorious Odysseus: don’t try to reconcile me to my dying. I’d rather serve as another man’s laborer, as a poor peasant without land and be alive on Earth, than to lord over the lifeless dead.4

Plato worried that if students read this line aloud they would identify with the emotion that Achilles is venting—self-pity, specifically self-pity with respect to death. Plato was afraid that engendering this feeling in the guardians had to be avoided because it would compromise their capacity as warriors. That is, having tasted of Achilles’s aversion to death, they might hesitate to plunge into battle. Likewise, Plato did not want the guardian children to mouth the lines of characters expressing fear, since he believed that cultivating fear would undermine their commit- ment when in battle. Nor did Plato want the young guardians to speak the words of gods disputing angrily among themselves, anticipating that this display might dispose the guardians, once grown, to find such strife legitimate, thereby opening the doorway to internecine conflict within the ruling class, or even to civil war. Plato felt by reading lines of poetry, readers would take on or iden- tify with the emotions expressed in those lines. And where those lines aroused emotions that Plato thought would potentially subvert the abil- ity of the guardians to defend and rule the Republic, he argued that the poetry should be regulated and even censored for its potential to undercut the guardians’ martial capabilities. Clearly the Platonic disposition lives on in contemporary calls for the regulation and censorship of violent media, although, ironically, where today the concern is that such media might encourage aggressive feel- ings, Plato was frightened that the media he distrusted would inhibit the warlike tendencies of the Republic’s soldiers. Nevertheless, both Plato and contemporary censors presuppose that a primary, if not the primary, form of emotional engagement with the arts is identification, understood as a matter of the audience taking on the selfsame emotions as the characters in the story, as a result of those characters having the emotions in question. Achilles feels rage toward Hector; we feel rage. K. feels disoriented by his circumstances in Kafka’s The Trial; we feel disoriented. Although Plato thought of identification as an artifact of reading aloud, the notion of identification has continued to hold sway into the era of silent reading. A number of problems arise with the theory of emotional identifica- tion, especially if it is thought of as a general account of our emotional engagement with literature. 6 Philosophy and Literature

The first problem is empirical. Very often our emotions do not par- allel the feelings of the characters about whom we are reading. At the end of Oedipus Rex, we feel pity for Oedipus. He is feeling remorse for having killed his father and having had intercourse with his mother. Yet we have no cause for remorse. We’ve done nothing like what he has (I hope). Likewise the character in a suspense novel may be hap- pily on her way home after a hard day’s work, whereas we are feeling high anxiety because we know that she is being stalked. Here, because we know more than the character does, our emotions differ. Moreover, sometimes the character may know more than we do—so the detec- tive, for example, Sherlock Holmes, stays cool, while we are in a state of excited curiosity. These cases of emotional asymmetry between the characters and the audience are not rare. They occur often, if not most often, in our commerce with literature. But the theory of identification cannot readily accommodate them. Of course, sometimes the character and the reader may be in roughly the same emotional state. Reading Jurassic Park, I may be terrified of the raptors as I read about the harried characters’ terror. However, this might not be best explained in terms of identification because my state of terror often is not a result of identifying with the emotional state of the character but, rather, of my being moved by attending—coinci- dentally—to the same thing the character is attending to. In the story, the character is struck by the lethal potential of the raptor’s maw and is frightened. Reading about the raptor’s sharp teeth, I too am fright- ened. But I am not frightened because the character is frightened; I am frightened because of my own apprehension of danger in the fictional situation as I construe it imaginatively. Thus, even in cases where the character and the reader appear to share emotional states, identifica- tion may not be the right explanation, thereby further compromising its empirical reach. Call this the problem of coincident emotions. In addition to these empirical problems with the notion of identifica- tion, we come across conceptual problems. The most outstanding is the fact that very frequently—perhaps most frequently—the objects of the reader’s emotional state differ from the objects of the emotional states of the characters. Emotions have intentionality. They are directed at particular objects, such as fictional characters. Our pity, for example, is directed at Anna Karenina, who feels sad because her husband has prohibited her from visiting her son. Let us say that the object of her sadness is her exile from her son. But that is not the object of our emo- tional state, which is probably best described as pity for Anna. This Noël Carroll 7 fictional woman who has been separated from her son is the object of our emotional state. We have not been separated from our own child. That separation is not the target of our emotion—of our pity. Anna—a woman suffering because she has been separated from her son—is the particular object of our affective state. Thus, if the notion of identifi- cation presupposes that the reader and the character share identical or exact emotional states, the theory cannot accommodate the large number of cases where the objects of the emotions of the readers and the characters diverge, even though cases like that of Anna’s separation from her son are often uncritically taken as evidence for identification. It is true that our pity for Anna and her plight and Anna’s sadness over her separation from her son are both dysphoric emotional states. But they are not identical dysphoric emotional states. There are also logical problems with the theory of identification. It can lead to contradictions. For example, fictional characters are often in the affective state of love. In Wuthering Heights, Heathcliff loves Catherine. Let us suppose that we, as the theory claims, identify with Heathcliff. Are we too then in love with Catherine? But if that were so, why wouldn’t we be jealous of Heathcliff, since Catherine appears to prefer him over all others? Yet does that sort of jealousy ever occur? Here, of course, pace identification, love does not describe our emotional state, presupposing that we have an affirmative attitude or positive emotional attachment to Heathcliff. Rather we feel sympathy toward Heathcliff, at least as we read the first half ofWuthering Heights, because his pursuit of Catherine is being unfairly blocked. Our alliance to Heathcliff is driven by our desire that his desires be fulfilled. When they are not, we feel distress. Distress, not love—which is what identification theory would predict. Let me mention one last problem with identification as a general theory of the reader’s emotional engagement with literature, perhaps the most obvious problem. Identification cannot provide a thoroughly com- prehensive theory of our emotional engagement to literature because a great deal of our emotional engagement is shaped by noncharacter, often nonhuman, features of the fictional world as made salient in the descriptions of the landscape, architecture, climate, clothing, flora and fauna, and so forth of the environment, not to mention by the vocabu- lary and prosody of the writing. 8 Philosophy and Literature

IV Given the problems that beset identification theory—including the problems of emotional asymmetry, coincident emotions, divergent affec- tive objects, noncharacterological emotive stimuli, and the potential for contradictions—I maintain that identification cannot be the general way in which we engage literature emotionally and, for that reason, cannot serve as an account of how literature educates or, in Plato’s view, miseducates our feelings. Instead, I propose an alternative theory of our emotional engagement with literature in order to support an account of one of the ways in which literature may promote emotional learning. I call this approach criterial prefocusing. However, before presenting that approach, let me say a few things about the conception of the emotions that I am presupposing. The conception of the emotions that I am presupposing is not origi- nal to me. On this view, broadly speaking, the emotions are processes involving mental states that cause internal bodily states of the sorts that typically prime behavioral dispositions to act in certain ways.5 Specifically, emotions comprise evaluative mental states that involve appraisals of our circumstances in terms of whether those circumstances will enhance or impede our vital human interests. Fear assesses a present situation as dangerous; that alerts us bodily by quickening our pulse physiologically and/or sending a chill down our spine phenomenologically. Once this bodily fright alarm is triggered, we are ready to flee or freeze. The emotions are psychophysical mechanisms that we use to assimilate the environment. The emotions pick out those features of the environ- ment that will help or hinder us. In particular, they are fast mechanisms for sizing up the environment—fast, that is, in comparison with slower ways of assessing our circumstances such as deliberation or reasoning. Evolution has endowed us with these affective means of rapid appraisal, since it is adaptive to respond quickly to environmental stimuli, for even if the emotions sometimes mislead, it is better to be safe than sorry for purposes of survival. For our prehistoric ancestors, it was more advanta- geous to freeze in response to an ambiguous noise in the bushes on the chance that it was a predator than it was to ignore it and be attacked by a hungry bear hiding in the foliage. Crucially, the emotions protect our vital interests. Fear alerts us to danger; anger to injustice; jealousy to loss of affection; sadness to loss; and so on. The current psychological and philosophical literature contains some debate about whether our emotive appraisals are cognitive or completely Noël Carroll 9 affective. This is sometimes framed in terms of whether these appraisals can be initiated in the frontal cortex or are mostly an affair of the amyg- dala. For my own part, I believe that emotional processes can sometimes be initiated cognitively—as in the case of academic envy, which may emerge only after my sustained reflection upon my departmental chair’s differential behavior toward me in comparison to my better-looking col- league—even if, in the normal run of affairs, our emotive appraisals are more often immediate, noncognitive, affective responses. Nevertheless, despite my overall catholicity with regard to this debate, I lean toward the view that our emotional responses to literature are more likely the product of cognition, not only because they are read but also because they must be engaged imaginatively and understood. They are hard to model upon our perceptual responses, since they are not reducible to perceptual responses. Some neo-Jamesians, like Jenefer Robinson, take the startle response as the paradigmatic emotional response and then attempt to apply it to literature.6 This might make sense to a certain extent with film, but its application to literature seems to me strained and unlikely. How often, if ever, is one literally startled when reading literature—that is, startled as one might be by a sudden clap of thunder of the sort that makes you shudder or start in your chair? It has never happened to me and I have read a lot of horror stories. However, in any event, for my theory of criterial prefocusing, the ques- tion of whether emotive appraisals are primarily cognitive or affective is less important than the fact that either way they are appraisals—appraisals of features of the environment in terms of what advances or imperils our vital interests. A helping hand is appraised as a contribution to our well-being and elicits a feeling of gratitude in consequence. An oncom- ing car appears to be moving too fast to stop before hitting me, so I assess it to be dangerous, thereby eliciting fear. The emotions are mechanisms of selective attention—they organize or gestalt the environment by picking out that which will help or hin- der us. The selectivity of the emotions is grounded in the fact that they appraise the environment and/or elements thereof either positively or negatively. Moreover, insofar as they evaluate the environment, they are governed by criteria. They size up the environment against certain standards of evaluation—for example, whether the situation is danger- ous in terms of fear, or whether it is unjust in terms of anger. Thus, the emotions have certain criteria of appropriateness, which the medieval philosophers called formal objects (DTR, pp. 28–101). In other words, a stimulus has to be perceived as possessing certain definite features 10 Philosophy and Literature in order to call forth a certain emotional response. For example, the situation has to be perceived as dangerous (even if it is not) in order to call forth the emotional response of fear. An emotional response can be called irrational if it is an inappropriate response to its intentional object, as would be my response of terror to SpongeBob SquarePants, whom I know to be harmless.7 The important point here is that insofar as the emotions are bodily appraisals—whether cognitive or perceptual or a mixture of both ele- ments—they are governed by criteria of appropriateness. Standards exist for what counts as a suitable particular object of a given emotional state. To be a genuine particular object of the emotional state of fear, the object must meet the criterion of being dangerous; the object of anger must be, according to Aristotle, a perceived wrong done to me or mine; to be the object of envy, one must be perceived to possess something that I lack but desire (like a Lamborghini Aventador); and so on. With this sketch of the nature of the emotions in hand, let us see what it can tell us about our emotional commerce with literature. But first I briefly characterize the ways in which the emotions function in everyday life in order to draw a contrast with the ways they function in response to literature. In everyday life, the emotions function as searchlights. They scan the environment in order to selectively pick out features of the situation that will enhance or impede our vital interests and, when they appraise items as such, they sound bodily alarms that rivet our attention to the pertinent features of our situation. The emotions are selective—they zero in on elements in the surround that are relevant to our vital inter- ests. The emotions focus our attention—they direct us to what we are to attend to and how (evaluatively) we are to attend to it. And once they fix upon the particular object that animates our state, the emo- tions survey the scene for other elements that may be relevant to the presiding existential theme. That is, awakened by the smell of smoke, the emotion of fear urges on us in search of its source. The emotions are natural endowments, albeit culturally calibrated, that function as our first line of response to the environment in advance of and often in place of deliberation. However, things stand very differently with regard to literature. In a story, the environment and its inhabitants are described in such a way as to make certain emotional themes—such as danger or injustice—stand out. The author literally controls our attention insofar as we can only attend to what she chooses to describe or imply, and, furthermore, Noël Carroll 11 we can only attend to that in terms of the ways in which she describes them. Whereas in, so to say, nature, our emotions do the filtering, in literature, the scenes and sequences have been predigested or prefil- tered for us by the author. This filtering guides our attention irresistibly to the elements of the scene that call for an emotional response that the author has already primed by selecting and making prominent an object appropriate to the emotional response she or he intends to elicit. Suppose you are a novelist and you are writing on the currently popular theme of a zombie apocalypse. You will describe agents and events using, among other things, adjectival and adverbial modifiers, with an eye toward eliciting certain preordained, affective responses by bringing to the reader’s attention those elements of the situation or of the character that are criterially appropriate to the desired emotional state. In our imagined novel of the zombie apocalypse, you will describe in gory adjectival excess the suppurating bodies of the zombies, their decay and fragmentation, in order to engender the reader’s disgust by focusing our attention unavoidably upon these features. The aim of a novel about a zombie apocalypse presumably would be to elicit horror, an emotion in which a central component is disgust in conjunction with fear. Vividly describing the viscous deterioration of the zombies is a surefire way to achieve that. As noted above, I call this process criterial prefocusing. It is a matter of prefocusing because the author has described the relevant scenes and characters, prominently emphasizing certain aspects of the situation that are pertinent to the emotions the author intends to provoke. This prefocusing, moreover, is criterial because the objects, events, and char- acters—and their properties that have been selected for attention—are those that are appropriate to the intended emotion. That is, they satisfy the criteria of appraisal for that emotion type. In the example of the zombie apocalypse, the themes of impurity and incompleteness meet the criteria for appraising a human body to be disgusting. So in order to elicit the feeling of disgust in readers, the author will describe scenes of the relentlessly encroaching zombie bodies as impure and incomplete, in a way that readers cannot avoid, save by closing the book. However, if we don’t close the book but instead continue to read, we are, in effect, rehearsing our culture’s patent for disgust.8 Similarly, suspense is an emotion that involves presenting a situation in which an unwanted outcome seems likely to obtain—for example, that Goldfinger in the eponymous James Bond novel will be able to contaminate the gold in Fort Knox with atomic radiation. Thus, in order 12 Philosophy and Literature to elicit suspense, Ian Fleming must craft the scene in such a way that the unwanted event appears inevitable by saliently, criterially prefocus- ing such features as Bond’s impediments and the relentless countdown of the dirty bomb. In order to see how the conception of our emotional engagement to literature can illuminate the way in which literature can afford affec- tive learning, I look at Aristotle’s theory of tragedy, which is typically regarded as his rejoinder to Plato.

V For Aristotle, the aim of tragedy is to arouse certain emotions, notably pity and fear, the very emotions Plato found suspect. In order to do this, Aristotle maintained that the particular object of these emotions—the tragic character—has to be constructed or designed in a very specific way. Aristotle reasons that the character couldn’t be morally perfect, for the audience’s response to the destruction of a saint, say Mother Teresa, would be indignation, not pity. Nor, Aristotle contended, could the tragic character be an utter villain, like Hitler. For the audience would by pleased by that outcome; they would feel the bastard got his just deserts. Instead, Aristotle recommended that the construction of the character should be such that he/she is of average virtue, not evil, but not perfect either. The tragic figure makes a mistake, but Aristotle emphasized this should not be a morally blameworthy mistake. Recall that Oedipus makes a mistake; he is in Thebes because he fled from Corinth in order to escape the prophecy that he would murder his father and commit incest with his mother. The reason Aristotle thought that the tragic character, though of somewhat elevated status, should be of mixed virtue, but not evil, is so he/she will be like us, and this will elicit fear in the audience; for if calamity can befall someone like us, it can befall us. That is, we will be terrified by the prospect that bad things can happen to good people, or, at least, tolerably good people. At the same time, insofar as unearned destruction crushes the tragic character, we will pity him/her, since undeserved harm is a criterion for the emotion of pity. We don’t pity and shouldn’t pity agents who merit punishment. In short, Aristotle argued that in order to engender the tragic response—the conjunction of fear and pity—the characters Noël Carroll 13 must be designed in such a way that their outstanding attributes are prefocused so that they satisfy the criteria of fear and pity respectively. Now, at this point, some will recall that Aristotle says the tragic emo- tion involves not only pity and fear but also the cartharsis thereof. That is, tragedy subjects pity and fear to the process named by cartharsis. But what is cartharsis? Aristotle was not very forthcoming on the matter and scholars have been arguing about it ever since. The Greek word can mean many different things, including cleansing, purifying, purg- ing, or clarifying, among others. For various reasons, which I defend elsewhere,9 I find “clarifying” to be the most promising interpretation. So, in that light, I propose that Aristotle held that the aim of tragedy is the clarification of pity and fear. But what does that mean? Recall that emotions have criteria that determine the appropriate objects of the affective states in question. I hypothesize that what Aristotle maintained is that the genre of tragedy aims at the clarification via prefocusing of the attributes of the tragic characters and their plight in terms of the proper or appropriate objects of the constituent emotions of pity and fear that combine to provoke tragic affect. Tragic literature teaches readers, viewers, and listeners what calls for tragic emoting on our part. As I have mentioned, the appropriate objects of tragic affect are situations in which bad things happen to good people. Tragedy tutors our emotions in this regard. It trains us in matching these affects with situations of this general character. It instructs us in what to feel.10

VI Generalizing from Aristotle’s theory of tragedy and what it suggests about literature’s contribution to our sentimental education, we may hypothesize that by means of criterial prefocusing, the preeminent form of our emotional engagement with literary texts, literature cultivates our emotional intelligence. No matter how realistic, literature presents us characters and situations that are highly selective, making the emotion- ally relevant variables salient in a way not typically so perspicuous in everyday life. This sensitizes us through rehearsal to be alert to poten- tially emotional concerns that might otherwise fly under our radar. It also may give shape to affective configurations otherwise too vague for us to give a name to. Undoubtedly our emotional education via literature begins on our caregiver’s knee, as he/she reads our earliest stories to us, accompany- ing them with interpretations, enactments, examples, analogies, and 14 Philosophy and Literature digressions aimed at leading us to feel what the author prescribes, which is usually normatively correct for the given culture or subculture into which we are being initiated. The parent’s very tone of voice will cue us as to who in Hogwarts should be loved and who feared. Teachers then take over as we are introduced to the canon of our culture, not only to be entertained but to augment the range and subtlety of our feelings, generally for good, but admittedly sometimes for ill.11 Once out of school, some of us continue to read—again, not only for entertainment but to continue the education of our feelings. Stories are leading sources of the education of our emotions. Many cultures, including our own, possess revenge sagas—to mention a poten- tially unsavory example. We use these kinds of stories—which Ronald de Sousa calls paradigm scenarios—to shape our own feelings and to explain the reactions of others (RE, p. 182). These scenarios may be the result of a spontaneous folk origin—sustained by gossip—or they may the product of the arts, ranging from country-and-western ballads to realist novels. But whether of humble or high descent, they serve to guide readers, viewers, and listeners by means of criterial prefocusing in the ways of feeling that bind various cultures and subcultures.12

VII So far I have been emphasizing the way in which literature serves to reinforce the emotional profile of a culture. And certainly a major function of literature is to train us in how people like us are supposed to feel. But clearly that training is too narrow a view of the role of litera- ture in the education of the emotions, for it not only sustains existing patterns of feeling but expands them. Indeed, it can even change our emotional responses toward certain particular objects from revulsion to sympathy. But how can literature achieve this? In order to address this question, I would like to look at perhaps the most successful literary transformation of the emotions of a culture, namely Harriet Beecher Stowe’s Uncle Tom’s Cabin. Although many, I think unfairly, would chal- lenge its literary credentials due to its undeniably aggressive persuasive techniques, I think that the very clarity of its almost transparent devices makes it an exemplary starting point for engaging the question of how literature can change cultural sentiments. Roughly speaking, what Stowe achieved was a reversal in the senti- ments of many readers toward African American slaves. Whereas previ- ously many regarded these slaves as subhuman objects of disgust, Stowe Noël Carroll 15 was able to engender feelings of sympathy toward them. She did this, I hypothesize, by moving the slaves from what Jesse Prinz calls one emo- tive calibration file to another, or, in other words, by recalibrating them. According to Prinz, our emotions are keyed to calibration files— approximate records, including perceptual records or memories of the sorts of saliencies that caused our various emotional alarms to go off in the past.13 Presumably, these calibration files would also contain criteri- ally focused mental representation, de Sousa’s (criterially prefocused) paradigm scenarios, and perhaps even concepts. Given certain stimuli, say a rattlesnake, a specific calibration file is activated, resulting in a bodily alert that causes us to freeze in our tracks. But then, say, we scru- tinize the stimulus and realize that it is not a live snake but a toy snake. In that case, we would subsume it—or recalibrate it—under a different calibration file—perhaps that of “rubber toy”—which would neutralize our initial visceral response. Using this vocabulary, Stowe moved black slaves from being the subjects of one calibration file—that of repulsive savages—to being the subjects of another calibration file, where they could be construed as worthy of sympathy. Stowe mobilized two major strategies to this end. The first, which is maybe very period specific, was to emphasize the fervent Christianity of the slaves, especially Uncle Tom’s faith, which is of the magnitude of a saint. The second strategy is probably more interesting, since it supplied Stowe with the emotional leverage to move not only non-Christian read- ers around the world in her own day but secular humanists today. This strategy underscores that slaves belonged to loving families and that slavery is, as Stowe observed, a system that “whirls families and scatters their members as the wind whirls and scatters the leaves of autumn.”14 The book opens with the scene of two slave families being torn asun- der: Uncle Tom’s and Eliza’s. Eliza, her son Harry, and her husband George are able to escape the clutches of the slave merchants and find safety in Canada. But Tom is brought to market and seemingly every slave he meets on his way there and afterwards has a story to tell about families being torn apart as they were being auctioned off. Even more than the physical miseries of slavery, Stowe stresses slavery as an affront to the sanctity of the family. Although for the most part Stowe makes her case by showing rather than telling, at some moments she makes the explicit point of family dismemberment—namely that the dissolution of these black families is no different than the destruction of white families would be, and that whites should regard it with comparable horror. Little Eva, worrying 16 Philosophy and Literature about what will happen to his slaves should her father die, says to him, “Papa, these poor creatures love their children as much as you do me” (UTC, p. 438). Throughout Uncle Tom’s Cabin, the slaves are recalibrated or recatego- rized, by means of criterial prefocusing from the category of unfeeling, subhuman brutes to subjects worthy of sympathy because of their family values. This gestalt switch does not come from nowhere. It involves a calibration-file adjustment due to the invocation of another conceptual category or calibration file—specifically that of the sanctity of the fam- ily. Notice—and this is key—that for this calibration-file switch, another calibration file must be accessed in order for the recalibration to obtain. Or, to speak less jargonistically, the object of our emotions needs to be linked with another, already existing set of emotion elicitors. In this regard, de Sousa employs the useful metaphor of cantilever- ing.15 In order to add a balcony to a structure, we need the support of an existing wall. To change an emotion toward x to another emotion toward x requires building upon emotions and emotion elicitors (cali- bration files, including criteria of appraisal, paradigm scenarios, and concepts) that are already in place in the culture. We can be induced to have new feelings toward the relevant objects by redeploying sentiments antecedently in our emotional repertoire. In effect, we are modifying our preexisting emotional economy by associating a previously indiffer- ent or aversive stimulus with a different emotion file or set of saliencies. Applied to the case at hand, using de Sousa’s metaphor of cantilever- ing, the reader’s emotional alignment with and approval of the slave families is supported by the audience’s preexisting commitment to the value of the family unit as such, to its sanctity (which entails its prima facie inviolability). On the basis of this example, then, one way in which a literary text can contribute to the education of our sentiments, such as engendering sympathy for the oppressed, involves recalibrating the relevant characters as subject to a negative calibration file to being subjects of a positive file that is dependent upon—cantilevered onto—typically preexisting affective criteria, concepts, and/or paradigm scenarios. In this way, new emotional learning builds on earlier emotional learning. That is, literature can encourage affective change by appealing to preexisting emotive criteria of a generally wider and more abstract order. In this way, Stowe moves the slave family, via criterial prefocusing, from the culture’s subsumption of it under the category of the animalistic and disgusting to that of the human family united in bonds of love and Noël Carroll 17 affection—a family that, when threatened, merits our sympathy and, when sundered, demands our tears.

VIII In brief, I have argued that when our emotions are engaged in the normal course of reading, we are—for good or for ill—typically gaining and/or refining our grasp of that which our culture (or subculture) deems it appropriate to feel in the relevant situation. The author is able to tutor us affectively by the process that I have called criterial prefocus- ing. Reading becomes an exercise in the rehearsal of what our culture regards as fitting to feel. This account, however, does not entail that literature merely reinforces the emotional status quo. Emotional change can be secured by redeploy- ing our existing emotional resources in such a way that the particular objects of our emotions are—via criterial prefocusing—realigned or recalibrated, or recategorized or associated with different, albeit already available, feelings. In this way, the emotional learning afforded by literature, like learning in general, builds upon what is already there.

The Graduate Center, City University of New York

The author would like to thank Patrizia Lombardo, Ronald de Sousa, Paloma Attencia Linares, Joan Acocella, and the audiences of the NCCR in Geneva and UNAM in Mexico City for their comments. They are not responsible for the imperfections above. I am. 1. Here I am going to be talking about “believing that x” rather than “knowing that x” in order to, for the purposes of this paper, sidestep the question of whether we can derive knowledge from fiction. I think we can. But my topic here is learning and it seems to me unobjectionable that we can learn things that we do not know. For the “chemical” interpretation of Midsummer Night’s Dream, see Colin McGinn, Shakespeare’s Philosophy (New York: HarperCollins Publishers, 2006), pp. 17–34. 2. See Jenefer Robinson, L’éducation sentimentale, in Art and Its Messages, ed. Stephen Davies (University Park: Pennsylvania State University Press, 1997), pp. 34–48; and Jenefer Robinson, Deeper Than Reason (Oxford: Oxford University Press, 2005), pp. 154–94, hereafter abbreviated DTR. 3. Roger Scruton, Culture Counts (New York: Encounters Books, 2007), pp. 36–37; Roger Scruton, “Emotion and Culture,” in The Aesthetic Understanding (1983; repr., South Bend: St. Augustine’s Press, 1998), pp. 173–79. 18 Philosophy and Literature

4. Homer, The Odyssey, trans. Richmond Lattimore (New York: Harper and Row, 1965), pp. 180–81, lines 488–503; emphasis added. 5. William Lyons, Emotion (Cambridge: Cambridge University Press, 1980). 6. See Anthony Kenny, Action, Emotion, and Will (New York: Humanities Press, 1964). 7. Ronald de Sousa, The Rationality of Emotion (Cambridge: MIT Press, 1987); hereafter abbreviated RE. 8. Undoubtedly, with more complex cases, we may be acquiring and/or refining the affects in question. 9. Noël Carroll, Classics in the Philosophy of Art (forthcoming). 10. On Aristotle’s view of the emotions, we not only need to learn the appropriate objects of the emotions. We also need to learn to emote at the right level of intensity, for the right reasons, at the right time, and so forth. Literature, needless to say, can also afford the refinement of our emotions along these dimensions. 11. The examples discussed above have been primarily been of salutary uses of crite- rial prefocusing. But criterial prefocusing can also be used to disseminate the ethically questionable emotions that exist in every known culture. In both cases, the mechanism of criterial prefocusing is parallel. That is why I speak of learning in this paper, since we may learn not only morally good feelings but also evil ones. 12. One objection to my claims about criterial prefocusing may be that it is too simple to handle anything but the simplest sorts of literature, as some of my examples, like Jurassic Park, might suggest. The suspicion may arise that the notion of criterial prefo- cusing, however suitable to mass-market fictions, is not really useful for accounting for more sophisticated literary works, where the emotions in question are hybrid, complex, multifaceted, ambiguous, and/or ambivalent. Yet I am not convinced that this worry is warranted. For where the emotion aroused by the literary fiction is multifaceted, the concept of criterial prefocusing still seems to me to be the more perspicuous way to get a handle on the fiction in question. This would involve an exercise in reverse engineering, so to speak. That is, observe the features of the situation that the author has made salient and work backwards to the emotion or emotions for which these features would be appropriate. Where the emo- tions are hybrid, multifaceted, complex, ambiguous, and/or ambivalent, the features of the situation that have been made salient will point in different directions. A bittersweet melodrama, for example, will have arrestingly presented elements that are appropriate to joy and others that correspond to sadness. And other works consist of even greater complexity than this. Indeed, other hybrids may occur for which we do not at present have an emotive label, but which we may refer to by means of the work of fiction itself or its author (as we refer to a certain kind of anxiety as Kafkaesque). I see no reason in principle to think that my strategy of reverse engineering cannot be extended to works of greater affective density than Jurassic Park, although showing that would have to be demonstrated in another paper—or maybe a series of papers. 13. Jesse Prinz, The Emotional Construction of Morals (Oxford: Oxford University Press, 2007). 14. Harriet Beecher Stowe, Uncle Tom’s Cabin, ed. Jean Fagan Yellin (Oxford: Oxford University Press, 1998), p. 438; hereafter abbreviated UTC. 15. Ronald de Sousa, Emotional Truth (Oxford: Oxford University Press, 2011), p. 34.