The Nature of Emotion

Again, the preceding account of affect presupposes an account of emotion. There are in fact several competing accounts of emotion in affective science. Simplifying the diversity in the field, we may organize the options into two large alternatives along two axes. Both axes concern the mental organization or “architecture” of emotion. The first is a matter of the large structures of the human mind; the second is a matter of the key processes that engage those structures. As to the former, affective scientists may be roughly divided into those who treat emotion in terms of “dimensions” and those who treat it in terms of “systems.” Dimension theorists seek to account for the diversity of emotional responses by reference to a limited set of variables.³⁴ Two variables are particularly common in such accounts—valence and arousal. Most often, these variables are presented as bipolar gradients of subjective (or “phenomenological”) experience; however, they may also be articulated in behavioral terms. Phenomenologically, valence is a matter of pleasant versus unpleasant experience; arousal is a more simply a matter of high energy versus low energy. Behaviorally, valence is a matter of an inclination to maintain a state or situation versus an inclination to change it; arousal is an inclination toward motor activity versus a disinclination. Different ordinary language categories of emotion may be mapped onto a conceptual grid using these dimensions.³⁵ For example, depression involves high unpleasantness and low arousal, whereas anxiety involves high unpleasantness and high arousal.

Dimensional accounts may lose something in elegance but gain in explanatory capacity as they add further dimensions.³⁶ For example, intuitively, it seems that anxiety is not simply a matter of unpleasantness and high arousal, which would appear to characterize a number of emotions, including, for instance, certain sorts or phases of grief. Rather, anxiety would seem to involve at least two other factors (both cognitive, rather than narrowly affective, though with clear affective consequences). First, there is a temporal dimension. One is anxious about the future, not about the past. Second, anxiety appears to be related to one’s sense that the unpleasant future (e.g., a simulated event) can possibly be avoided.

What is most significant in the present context is that the preceding points suggest a connection with literary experience. Specifically, in a dimensional account, anxiety—either egoistic (i.e., for oneself) or empathic—would appear to follow the same general principles as literary suspense. To think about this, we might begin from Ed Tan’s insightful account of interest.³⁷ Given a preferred outcome, our interest is sustained through assessments of the likelihood that this outcome will be attained. Given the right intensity of preference, and an adequately aversive quality to the alternatives (e.g., when the hero must either subdue the villain or die trying), this covers not only interest, but also suspense. If anxiety is indeed related to suspense, this already begins to give us a potentially fuller understanding of anxiety in its relation to the (subjective) likelihood of preferred outcomes and the emotions inspired by salient alternatives. But that is not all there is to the issue. The study of literary and cinematic suspense alerts us to the great importance not only of overarching goal pursuit, but of detailed, moment-to-moment interactions, the particulars of the narrative trajectory. These are presumably no less important in anxiety than in suspense. Finally, literature and film may give us not only targeted suspense (e.g., regarding whether the hero will defeat the villain or be killed). They may also give us more generalized and undirected feelings of foreboding (due for example to the use of music in a film). Such background elicitors of emotion are presumably no less important in anxiety than in suspense, since anxiety too may be undirected or “generalized.” As these points suggest, a literary analysis of suspense may potentially contribute to a psychological analysis of anxiety as much as the reverse. In other words, this may be an example of how the study of literature may contribute materially to our understanding of general psychological processes.

A dimensional approach may have consequences for concrete interpretation as well. For example, both Jonson’s and Lǐ’s poems treat grief. But there are striking differences between them, differences made salient by an awareness of emotion dimensions. Both poems express a high degree of unpleasantness. But Jonson’s poem is more evidently backward-looking temporally and more consistently low in arousal or activity. Lǐ’s poem, in contrast, begins with a frantic search that appears pointless—indeed, a search that continues even though there is no longer any suspense, thus hope. Yet, there are moments of ponderous lethargy in Lǐ’s poem, moments more depleted in arousal than anything in Jonson’s poem. For example, the speaker wonders who will stand and pick a flower at her home, suggesting that she herself lacks the energy for even this minimal exertion. Conversely, awareness of the dimensionality of emotion is likely to draw our attention to the limited suggestions of activity that do occur in Jonson’s poem, as when he assimilates the burial of the child to his own act of paying a debt. This points toward a suggestive ambiguity. Is the father’s debt owed to God, who has exacted the child’s demise as payment perhaps for the father’s sin? Or is the debt owed to the child (“I thee pay” meaning “I pay to thee [as creditor],” rather than “I pay with thee [as, say, collateral]”)—a debt that might have been incurred due to some flaw in fathering? In any case, the lines (beginning with “My sin”) indicate the centrality of the emotion of guilt with its deep sense of responsibility for action (or inaction when action was required). In contrast, Lǐ’s poem seems to point more toward utter irrelevance of any action and the speaker’s helplessness concerning a loss that was entirely beyond her influence. The difference is presumably not merely a matter of these two poems, but of kinds of grief as well. Along with the preceding example (regarding anxiety and suspense), these observations suggest that the incorporation of dimensional analysis into literary study promises to be beneficial for both the understanding of emotion and the understanding of literature.

This is not to say, however, that such mutual benefits are confined to dimensional accounts. Indeed, despite the apparent attitudes of their advocates, dimensional and systemic accounts are not irreconcilable. To the contrary, the dimensions could simply apply across systems—disgust, fear, anger, and so on—as part of what makes all of those systems emotional. Systemic accounts commonly presuppose the variables invoked by dimensional theories, but link them to innate (or sometimes acquired) complexes of distinctive feelings, action tendencies, and so on. For the most part, our usual, pre-theoretical approaches to emotion and literature are systemic, presupposing the existence of fear, anger, disgust, and so forth, as fundamental categories.

The situation is similar with the second axis of difference in theories of emotion—that is, the nature of causal processes invoked to explain emotion episodes. The most common account of emotion in affective science is one or another version of appraisal theory. An appraisal theory is an account that understands emotions as interpretations of circumstances or events relative to goals or, more broadly, as judgments about the relevance of circumstances or events to one’s “flourishing,” as Martha Nussbaum would put it.³⁸ As Keith Oatley explains, “Emotions are evaluations of events as they affect our concerns.”³⁹ The main alternative to appraisal theory may be called “perceptual-associative.” It posits a more mechanical series of perceptions and associations as triggering emotional responses due to specific sensitivities in particular emotion systems. Appraisal theories suggest the importance of general trends or statistical likelihood. In contrast, perceptual-associative theories emphasize concrete particulars. For example, in one perceptual-associative account,⁴⁰ there are three sources of emotion system activation: innate sensitivities, critical period experiences, and emotional memories. Innate sensitivities include not only responsiveness to physical pleasure and pain, but also (among others) our receptivity to the emotion expressions of other people and our susceptibility to emotion contagion from witnessing those expressions. Critical period experiences are early events that have enduring organizational and dispositional consequences for our emotional systems, such as an inclination toward secure or insecure attachment bonding. Emotional memories are memories that, upon activation, revive the initial emotion itself (as when the recollection of an automobile accident revives the sense of panic that accompanied the past experience).

Appraisal theories appear to have the advantage in many complex social cases. For example, Jones reads a letter denying his application for tenure and is devastated. It seems plausible to see this as a function of his judgment that his flourishing has suffered a keen blow. Perceptual-associative theories, however, appear to have the advantage in cases where we respond to stimuli in ways that go against our assessments—as when someone fears flying more than driving, despite knowing that the former is much safer. Moreover, the perceptual-associative theories would seem to have the advantage in cases of emotional experiences that apparently do not involve the advancement or inhibition of one’s flourishing. For example, Juslin notes that “music would not appear to have any capacity to further or block life goals,”⁴¹ but it may be highly emotional. The two accounts may be partially reconciled, however, by noting that appraisal can be incorporated into a perceptual-associative account. Appraisal processes involve concrete simulations (e.g., Jones’s simulation of telling his spouse about the tenure denial and seeing her distress), the activation of emotional memories (e.g., of past failures or social exclusions), and other factors that would trigger emotion according to a perceptual-associative account. The difference is that, in the perceptual-associative theory, it is not the logic of the appraisal that produces the emotion; it is, rather, the concrete imaginations and recollections accompanying the appraisal that do so.

The opening of Jonson’s poem suggests an appraisal account.⁴² The speaker grieves because he had “hope” for his son. Technically, he simulated a future for the boy and he now realizes that the imagined future cannot be realized, with painful consequences for his own flourishing. But later lines of the poem suggest limitations of the appraisal account. As we have seen, part of the speaker’s emotional state is a matter of pity for the boy. That pity is inconsistent with his assessment of the child’s Heavenly state. This is presumably because he witnessed the boy’s sickness and death, a direct perception forming emotional memories, but he has had no such experience of the child’s postmortem bliss. In my personal experience, a key line in the poem is, “O, could I lose all father now!” This line bears directly on the issue we are considering. To me, this sentence is a devastating cry of despair, rendered all the more pathetic by its grammatical peculiarity. The phrase is ambiguous. But what I take it to mean—and what guides my emotional reaction—is that he could foreswear ever having been a father in the sense of ever having heard himself addressed as “father.” Behind this sentiment, then, is the now-searing, emotional memory of the boy calling to him with the title (“father”) that would ordinarily be sweet with the tenderness of attachment bonds—perhaps even addressing him from his deathbed. If this is accurate, the power of this line is partially clarified through a search for concreteness fostered by the perceptual-associative account of emotion.

Emotion Episodes

Of course, emotions are not simply mental structures or types of causal process. They are events, the activation of systems by causes. A key aspect of affective science—related to its connections with Anglo-American or “analytic” philosophy—is the analysis of emotion episodes into components. We tend to think of emotion as primarily a feeling. This is called the “phenomenological tone” of the emotion. But it is only one element of an emotion episode. Indeed, it is a rather strange element. Most components of an emotion episode may be articulated from an observer’s perspective. This is not to say that an actual observer would be able to acquire knowledge of all such components. It is simply to say that they can be phrased in a third-person idiom. Only phenomenological tone is essentially subjective.

Before turning to phenomenological tone, however, we need to consider some of the other components of an emotion episode. The first component is eliciting conditions. These are the circumstances that give rise to the emotion episode. Such circumstances most obviously include the events and properties of the external world. But they also include affective dispositions and moods on the part of the person feeling the emotion. Consider Jonson’s line, “O, could I lose all father now!” The suggestion is that he cannot bear even to hear the word “father.” Clearly, the cause of the emotion is not the word alone. It is only in the context of his mood of mourning that hearing the word becomes an eliciting condition for intensified grief. This is, of course, because it calls to mind his dead son—perhaps an emotional memory of the boy’s voice saying “father” (or, what is worse, a simulation based on a report of the son calling in his absence). In keeping with this account of the eliciting conditions, the speaker makes reference to the child’s voice, admonishing the dead son to answer an unspoken question about the inhabitant of the grave. In doing so, he implicitly makes the boy avoid saying “father.” Specifically, the speaker tells the boy to refer to “Ben Jonson his best piece of poetry,” not “My father his best piece of poetry.” Once noticed, the omission becomes striking.

A second component of an emotion episode is expressive or communicative outcomes. These are results of the emotion that do not serve directly to alter or maintain the situation as such. Rather, they serve to convey the emotion to those present (which may of course lead in its turn to changes in the situation).⁴³ For example, a cry of fear at the sight of a predator would serve to warn others of the danger. Note that this does not mean the communication is planned. Rather, such communication results from the spontaneous emotion expression and at least contributes to its evolutionary function. Communicative outcomes not only provide information to others. They tend to inspire emotion contagion or empathy and they are crucial for the acquisition of appropriate emotional responses in childhood (e.g., when a child learns to fear and thus avoid dangers due to the caregiver’s expressions of fear). (Emotion contagion is the egocentric experience of someone else’s emotion, as when Jones’s fear of a predator makes me fear for my own safety; empathy is the allocentric experience of someone else’s emotion, as when Jones’s fear of being denied tenure makes me fear for Jones.⁴⁴)

Communicative outcomes figure most obviously in film and theater. Carl Plantinga’s important essay on “The Scene of Empathy and the Human Face on Film”⁴⁵ presents a compelling analysis of viewer emotion as to a great extent mirroring that of the faces presented on screen, particularly in close-ups. In emotional mirroring, the observer experiences some version of the emotion expressed by the behaviors that he or she witnesses in the target (e.g., an observer may feel happy on seeing someone else smile). This is due to overlap between the neurologically based systems that respond to one’s observation of other people’s behaviors and the systems involved in one’s own enactment of related behaviors (e.g., other people’s smiling and one’s own smiling).⁴⁶ Plantinga’s point could be extended to other expressive or communicative outcomes, such as posture and gait.⁴⁷ Moreover, it could be complicated by noting the difference between a parallel and a complementary interpersonal stance (as discussed above).

Expressivity enters into literary works as well, however. This occurs in part through simulation and through subvocalization, the silent “voicing” of a speaker’s utterance. Tone of voice—including, variation in volume and pitch contour—is a very important expressive or communicative outcome. Consider again the line, “O, could I lose all father now!” Part of the impact of the line relies on subvocalization. Imagine it with the contour of, say, “O, could I use a frothy brew now!”

A third component is action readiness, which may combine with actional outcomes. Action readiness is the entire bodily and mental orientation toward behavior that will either alter the situation or maintain it, depending on whether the emotion is aversive or pleasurable. It involves the “priming” of motor routines (i.e., the partial activation of such routines, below the threshold of enactment), the preparatory tensing or relaxation of the musculature, and other, relevant physiological orientations. Actional outcomes are, of course, the actions themselves, such as running from a danger.

As already noted, Lǐ’s poem begins with an actional outcome, “search search” (xún xún; 尋尋‎). The suggestion may be that the speaker is looking for her husband. She is seated at the window, looking out through the clear, cold air, scanning the distance. But the reader infers that her husband is dead. Even if he is not dead, there is nothing she can do. She desperately wishes to change the situation, but she is helpless. Thus the only option is not to change the external situation, but to change herself. This actually leads to another component of an emotion episode, one parallel to the actional outcome—emotion modulation. One may seek to shift, diminish, sustain, or intensify one’s emotional response to a situation by changing oneself. For example, when struck by a funny idea in the middle of a funeral, one may try to turn one’s thoughts in another direction, shifting one’s emotion. Unable to engage in action that would change her situation, the speaker of “Shēngshēng Màn” seeks to change herself. First, she tries to rest; sleep would, of course, make her impervious to her circumstances, but she cannot sleep. Then she downs cup after cup of wine, but rather than enjoying the drink and raising her spirits, she finds it insipid (dàn; 淡‎). Like everything else that evening, the suggestion is that drinking only serves to remind her that she is alone (dúzì; 獨自‎). The actional and self-modulatory outcomes are thus inconsequential. That impossibility of amelioration is central to the despair of the work.

This leads to yet another component in an emotion episode—emotions tend to alter cognitive processing. In some cases, this processing is more or less purely informational. Positive emotional states tend to foster top-down processing, stressing generalities over particulars. Negative emotional states tend to do the opposite, stressing particulars over generalities.⁴⁸ We arguably see the latter in Jonson’s poem, with its emotional emphasis on the particularity of the son’s death, at the expense of generalities about divine beneficence and the bliss of Heaven. This, in turn, suggests a cognitive reason that grief is so resistant to consolation.

Other processing changes are more obviously emotional themselves, such as the tendency toward emotion-congruent processing, the interpretation of and response to ambiguous situations in a way that is consistent with one’s predominant emotion at the time.⁴⁹ We find a prominent case of this in what Chinese critics sometimes refer to as “emotion-scene fusion.”⁵⁰ This is not a misguided application of human properties to nature. It is an apt reflection of the way in which emotions guide our response to our environment. There is a touching case of this in Lǐ’s poem. Instead of seeing beauty in the fall blossoms, she finds them haggard, languishing, and harmed (qiáo cuì [or qiáocuì] sǔn; 憔悴損‎); implicitly, they are like the speaker herself. In other words, due to emotion-congruent processing, she understands and responds to the condition of the flowers in terms of her own physical condition, posture, and recent experience. It is a sort of reverse mirroring.

The mention of mirroring returns us to phenomenological tone, since mirroring gives us a hint of another person’s emotional experience, but not the experience itself (as shown by the flowers, which of course are not having any emotional experience). Again, phenomenological tone is the one component of an emotion episode that is essentially private; it is a subjective experience that cannot be observed by anyone else or even phrased in a third-person idiom. But that does not mean it is inconsequential. The pleasure or pain of emotion is part of its phenomenological tone. As such, that tone is precisely the part of an emotion that one would wish to share with others, appealing for their commiseration or celebration. But it is also the one part of emotion that is ultimately unshareable. The point is particularly clear in ordinary life, where our individual experiences often diverge sharply for the most personally consequential events. For example, even a friend or relative is unlikely to share one’s joy in success without jealousy or some other taint of egoism, and the grief that touches one’s own life is rarely common to friends, or even spouses. However, in some respects, the detailed elaboration of emotional particularity in a literary work may give us intimations of more fully sharing emotions with characters or authors or may enable a sense of elaborate emotion sharing with other readers.

One common view of phenomenological tone, associated with William James, identifies it as the subjective experience of bodily changes (“physiological outcomes”) that occur during emotion episodes. In other words, the “feeling” of an emotion (as Antonio Damasio calls it)—what many people are likely to think of as the core of emotion itself—is not ethereal or spiritual, but thoroughly embodied. In recent work, Joseph LeDoux explains James’s view in the following terms: “Different emotions are experienced differently because they involve different body signatures that produce different patterns of feedback and thus different feelings.”⁵¹ As LeDoux goes on to explain, this account was developed, extended, and rendered neurologically rigorous by Antonio Damasio in his theory of “somatic markers.” In Damasio’s account, bodily (“somatic”) feedback not only produces feelings; it also marks cognition, with important consequences for decision making, which is therefore not merely abstract and rational but deeply bound up with emotion and embodiment.⁵² Among other factors, this is presumably due to the fact that decisions bearing on action are a matter not merely of information processing (thus logic and evidence) but of motivation (thus emotion, including phenomenological tone) as well. Indeed, the general idea is closely related to the account of emotional response in fictional simulation given above. Decisions are in part a matter of simulating outcomes and experiencing particular positive or negative emotions in response to those simulations. The emotional responses have significant motivational consequences, thus significant consequences for decision-making.

Authors and Readers

Needless to say, real emotions are to be found only in real people. There are two sorts of real persons involved in literature—authors and readers. The preceding section referred briefly to emotion sharing. The idea applies to both authors and readers. Readers who particularly enjoy a work or find it especially moving are likely to want to share that emotional experience with other people, prominently attachment figures, such as friends or family members. As to authors, some accounts of literary production make emotion expression the primary purpose of literature. Some readers today are likely to associate this view with the Euro-American Romantic movement. But versions of this view may be found in different times and places. For example, in Japanese tradition, Lady Murasaki’s 11th-century Tale of Genji includes the following account of authorship: “the storyteller’s own experience …—not only what he has passed through himself, but even events which he has only witnessed or been told of—has moved him to an emotion so passionate that he can no longer keep it shut up in his heart.”⁵³

Writers such as Bernard Rimé have brought emotion sharing into affective science, stressing its centrality to human social life.⁵⁴ The most obvious function for emotion sharing is modulatory. Someone else may calm one’s anxiety or enhance one’s joy. This is related to a broader, longer-term function—the possibility of, so to speak, calibrating one’s own emotional responses by reference to the quality and intensity of other people’s responses. Another possible function for emotion sharing is in ascertaining who might be compatible as a friend: someone whose emotional responses might be congruent with one’s own. In the case of literature, these functions apply most obviously across readers, though, as Lady Murasaki suggests, sharing from author to reader is important as well. In keeping with this, the impulse toward emotion sharing seems clear in both poems. That impulse is unusually poignant in the case of Lǐ’s poem because it is evident that the speaker wishes to share her anxiety with her husband and to be reassured by him. The reader may be keenly aware that he or she cannot reassure the author, that her poetic act of emotion sharing in fact provided no benefit of emotion modulation for her. She remained alone (dúzì; 獨自‎) with the clutter of languishing blossoms and the heart-piercing (zhèng shāng xīn; 正傷心‎) cries of the passing geese, the only “acquaintances”⁵⁵ (shí; 識‎) mentioned in the scene.

The situation with Jonson is somewhat different. In a lovely line, Jonson has his dead son characterize himself as “Ben Jonson his best piece of poetry.” Unlike Lǐ’s poem, Jonson’s poem presents us with explicit self-consciousness about the poetic act of emotion sharing. Jonson knows that this is a poem that will be sent out to the public to be read and appreciated, and its author will be pitied. In connection with this, he presents a clear image of the “implied author,” Ben Jonson the poet and playwright. Indeed, the poem illustrates an ambiguity in the idea of an implied author. On the one hand, the phrase (“implied author”) refers to a set of textual norms; on the other hand, it refers to an image of the real author.⁵⁶ The two are related, and both may be linked to the real author,⁵⁷ but they are not the same. Even so, in both cases, special emotional concerns arise with respect to the implied author in an emotional context, prominently the issue of emotional sincerity. Readers may differ in the degree to which they are willing to allow Jonson’s grief to be sincerely compatible with the hint of vanity in praising the boy as his “best … poetry.” (This analysis assumes that expressions of grief, remorse, and loss of hope are fully sincere.)

The reference to differences in the reception of the poem returns to real readers. It also brings up the relation between real readers and implied readers. Like authors, real readers have real emotions (not just simulated emotions, like characters). Those emotions differ individually, while also manifesting some recurring patterns. The empirical study of reader response, including emotional response, gives us one way of uncovering such patterns. Such study has developed importantly in recent years through work by Marisa Bortolussi, Peter Dixon, Don Kuiken, David Miall, and others.⁵⁸

Perhaps the most important patterns across readers are not specific to literature, but extend across people’s general routines of thought and feeling. Perhaps the key aspect of readers’ affective response is empathy, which is strongly affected by out-grouping.⁵⁹ Empathy has been a topic of considerable affective science research in recent years; the bearing of this research on literary study has been discussed influentially and insightfully by Suzanne Keen.⁶⁰

Implied readers are textually guided simulations that provide norms for response to a literary work. Sometimes, critics object to the idea of an implied reader as it may seem to define the appropriate responses of a reader too narrowly. However, there are many cases in which we might wish to say that a given reader’s emotional response is misguided—as in the case of my student giggling at Lǐ’s poem. Conversely, we may wish to say that a poem fails aesthetically if its effect on real readers is very likely to diverge from its implied readerly norm. If almost everyone giggled on reading Lǐ’s poem, it would be important to know whether it was a poem of grief or a parody of a lament. That difference is at least in part a matter of the implied reader’s (normative) emotional response.

Narrators and Plots

Ben Jonson’s foregrounding of himself as a poet is not only a matter of forging an image of the implied author and establishing norms for reading this particular poem. It is also a matter of bringing out the narrator or speaker of the poem as a (simulated) voice distinct from the author or implied author. This is suggested by the references to Ben Jonson as “he.” Of course, one can simply take the “his” in “Ben Jonson his best piece of poetry” to be in the voice of the dead child. But the concluding couplet (“For whose sake henceforth all his vows be such, / As what he loves may never like too much”) is more general. It is spoken in the voice of the governing consciousness of the poem. That is the voice of the narrator.

Narrators raise special emotional concerns in literature. Most obviously, they invite degrees and kinds of trust or distrust. Often trust and distrust are interpreted in terms of proximity to or distance from the implied or real author. However, things are more complex than this. Indeed, Jonson’s poem suggests that, in some cases, trust may be enhanced by at least some degree of differentiation between the narrator and the author or implied author. There is, for example, something a bit strange about Jonson’s narrator’s claim that the dead child was Jonson’s “best piece of poetry.” It is important that it is given in a narrator’s voice. If the poem had said, “my best piece of poetry,” the egoism of the claim would have likely inspired distrust in many readers. After all, the metaphor works to eulogize the child only if Ben Jonson is a truly stellar poet. Our trust in the narrator’s judgment here is in part a function of the narrator diverging from the author and implied author, who are presumably too grief-stricken to consider such matters as professional success. Beyond such issues of trust, and consequent emotions, narrators may be more or less personified⁶¹ and as such may inspire a range of ordinary emotions in greater or lesser degree.

In addition to narration, narratologists distinguish plot or emplotment as part of how a story is told, thus an element of the “discourse” (as opposed to what is told—the story proper). Emplotment involves such matters as narrational manipulation of what information is given when (e.g., whether the outcome of the events is given at the start or deferred until the end of the telling) and how that information is articulated or construed. Plot too involves its own distinctive set of emotions. We have already noted that interest, in Ed Tan’s account, as well as suspense are a function, not only of the story per se, but of the timing and construal of information from the story, which is to say the plot. Since Meir Sternberg’s work on the topic, it has been customary to distinguish curiosity and surprise in addition to suspense.⁶² Suspense is a matter of outcomes; curiosity is a matter of causal precedents, once we know the outcome. Surprise is a function of expectation. Though certainly valuable, it is not clear that this is the most systematic way of dividing plot emotions. Combining Sternberg with Tan, we might suggest a dimensional account in which there are variables bearing on the degree of preference for any given event (whether an outcome or a precedent), degree of likelihood for that event, and prospective versus retrospective orientation. For example, a strongly preferred event with low likelihood that has occurred and been narrated is likely to result in a sense of relief, making this another plot-based emotion alongside suspense and curiosity. Surprise might be understood as bearing on a different dimension—consistency with anticipatory simulations. It may occur independently of such emotions as relief or as an inflection of them.

Lǐ’s poem presents us with a remarkably complex case of emplotment, mixing temporal orientations, perhaps by mixing a memory with a current situation. Specifically, the poem suggests a current evening when the speaker is alone following the death of her husband. But it seems to intermingle a memory of waiting in vain for his return in the past. We see this in the searching at the outset, which would seem pointless if the husband is known to be dead. This gives the poem a dual temporal orientation, with all the emotional complexity that this suggests. The ambiguity is enabled by the relative absence of tense markers in Chinese.⁶³ It is further developed by the images of approaching winter and approaching night that both suggest death is coming rather than that it has already come. This temporal and emotional duality is made more or less explicit in the closing line of the poem. The speaker asks how a single “zì” (字‎) could convey her feelings of “chóu” (愁‎).⁶⁴ “Zì” is often translated as “word” and “chóu” as “sorrow.” But it is also possible to translate the former as “character,” referring to the written form particularly, and the latter as “worry.”⁶⁵ In the context of the poem’s complex emplotment, the line is emotionally brilliant. “Chóu” does really summarize her feelings, because it conveys both the retrospective sorrow and the prospective worry. Indeed, the worry continues in the present through the haunting memories of searching and seeking and through the speaker’s distress over (simulations of) her own lonely future. In addition, the character for “chóu” (愁‎) contributes to this complex feeling, for it combines the characters for “autumn” (“qiū”; 秋‎) and “heart” (“xīn”; 心‎), the former placed above the latter as if the autumnal change, with its implications of winter and death, weigh down on the speaker’s heart. Indeed, the character for “xīn” is flattened in “chóu.”

Stories

Story is, again, what happens, what the plot partially represents and partially reorganizes. The story is what is done by or happens to whom and where it all occurs. In other words, story comprises the events, characters, and scenes. Sometimes the second and third (characters and scenes) are referred to as the “storyworld” and distinguished from the sequence of events that delimits the story, in a narrower sense. There are many ways in which both character and scene may bear on affect and literature. For example, there has been much work on character and “Theory of Mind,” the means by which we come to understand (or misunderstand) other people’s intentions, aspirations, and other mental states, including emotions.⁶⁶ In general, characters may have—or provoke in readers or in other characters—all the usual human emotions. After all, it is again part of our simulative capacities that we respond to simulations as we respond to comparable realities. The same general point holds for scenes or places. In each case, there may be particular sorts of relation that are especially consequential emotionally, such as the organization of characters into in-group and out-group members or of places into home and away from home.

The relation of emotion to stories (in the narrow sense) is perhaps more interesting. A number of writers have argued that there are cross-culturally recurring story sequences. Outside cognitive and affective science, these are sometimes posited as the result of “archetypes.” In Jungian psychology, archetypes are instances of “a tendency to form representations of a motif—representations that can vary a great deal in detail without losing their basic pattern,”⁶⁷ a tendency “inherited” through a “collective unconscious.”⁶⁸ In contrast to this approach, working from within affective science, I have argued that a specific set of cross-cultural story patterns may be understood as deriving from emotion systems in a rule-governed way.

First, I have argued that categories of story structures have the same general properties as other social categories. Specifically, such categories are not usually defined by strict necessary and sufficient conditions, but by roughly average instances that serve as “good cases” of a category. The point applies, then, not only to such ordinary, physical objects as tables and birds, but also to stories. The “good cases” are called “prototypes” (following Eleanor Rosch⁶⁹). It is important to distinguish (experientially defined) prototypes from archetypes, despite the similar sound of the two words.

Beyond this semantic point, I have drawn on stories from a range of traditions to argue that there are recurring genre patterns across traditions of literature and orature.⁷⁰ These genre patterns define genre prototypes and the prototypes may be explained by reference to ordinary features of emotion systems. First, the genre prototypes include romantic, heroic, sacrificial, familial, seduction, revenge, and other recurring story sequences. For example, the romantic prototype involves two people falling in love, encountering social resistance (often from parents), experiencing separation (often exile or threat of death), but then—in the full, comic version—overcoming obstacles to be reunited in an enduring and idealized union. The heroic story often develops in two sequences. One treats the usurpation (often by a close relative), exile, threatened death, and ultimate return of the rightful social leader; the other treats the temporary defeat of the home society by some social enemy.

In each case, the story pattern was not produced by some sort of collective sensibility. Rather, it resulted from algorithmically specifiable psychological processes relying only on standard mental architecture as set out in cognitive and affective science. Specifically, stories involve the usual appeal of simulation, combined with the usual folk psychology of characters following goals. (Folk psychology is the set of psychological principles that we rely on tacitly in understanding people’s minds.) Characters’ goals are defined by biologically given emotion systems, specified in relation to socially defined practices. The working out of the story prototypes results from the operation of ordinary principles that apply across emotion systems. For example, sexual desire and attachment systems produce goals—sexual union in the first case, enduring proximity with reciprocal attention and care in the second. These generate goals for characters and thereby produce genres. The sexual desire system gives rise to the seduction genre; the attachment system fosters the family separation and reunion plot. These emotion systems combine in romantic love. As it derives from a combination of motivation systems, we might expect the goal of romantic union to be particularly intense. This, and the frequency of the combination in real life, may lead us to expect in turn that the resulting genre would be particularly common. This is just what we seem to find in the proliferation of romantic narratives in different traditions. Similarly, pride and anger contribute to the generation of heroic narratives. Moreover, the dual story structure of this prototype derives from the combination of two types of emotions as well—those defined by individual aspiration or accomplishment (for the hero) and those defined by group identity (for the home society).

The definition of goals only gives us the largest trajectories of the genres, however. More specific structures are produced by some ordinary principles of emotional response. For example, an outcome emotion is generally enhanced by the gradient of change from a prior, contradictory emotion. If things turn out to be simply normal, that will give rise to joy (or at least relief) if the outlook had been grim, but it will foster sadness (or at least disappointment) if there had been a prospect of advancement. If I anticipate losing my job, then simply not losing my job is cause for celebration. But if I anticipate a promotion, then merely not losing my job may leave me melancholic. This is the reason stories commonly involve the opposite of their desired outcome in the middle. Tragedies end with the opposite of the desired outcome. But in full comedies the happy ending is intensified by the near-tragic middle. Thus, in romantic stories, lovers are separated, and sometimes thought to have died, before being reunited. In the heroic genre, the home society is devastated before defeating the enemy.

Other forms of emotion intensification include the manipulation of place attachment and what may be called “familialization.” Often the manipulation of place attachment appears as exile in the tragic middle, where lovers and deposed leaders are sent to alien places. The comic ending then involves not only romantic union or the restoration of rightful leadership, but also the protagonist’s return home. As to familialization, attachment bonds make conflict more painful. In consequence, the clash between the lovers and society is often a clash within the family and the heroic usurpation is often familial (e.g., fraternal) as well. Of course, these patterns to some extent reflect patterns in real life in addition to serving emotional intensification. But not all factors that recur in real life also recur in stories; the selection is guided by emotional functions.

Familialization provides a good example of the contrast between cross-cultural prototypes and Jungian archetypes. Jung identifies “the motif of the hostile brethern” as resulting from an archetype,⁷¹ which is to say, from one of the “pre-existent forms” that are “inherited” through a “collective unconscious” that is “universal” and “impersonal.”⁷² In contrast, the cognitive account just given views fraternal conflict as the result of emotional intensification repeated across individual stories, thereby yielding a standard or usual case (the prototype). As this should make clear, the concepts of prototype and archetype are no more closely related than the concepts of catalog and catastrophe. (Of course, the concepts of archetype and universal are related in that archetypes, should they exist, would be a form of universal.)

In short, emotion is a crucial factor in the generation and organization of stories. This is unsurprising. Stories involving human action are necessarily driven by motives. It would be incomprehensible if motivation systems did not play an important role in the development of tales recounting such action.