P ATRICK C OLM H OGAN Before Stories: Anna Karenina and the Emotional Structure of Lived Time Anna Karenina begins with a rift in a family. Dolly Oblonsky has discovered that her husband has been having an affair. One morning, a few days after the initial quarrel, Stiva Oblonsky, the husband, wakes up alone in his study. For a moment, he does not remember the rift, or even his own precise location in the home. For a moment, he is content, tacitly imagining his ordinary life, his ordinary bed. But then he remembers. “All the details” come rushing back, but they are not uniform. He particularly recalls, “the first moment when, on coming back cheerful and satisfied” he “saw her . . . holding the unlucky note that had revealed everything” (2). He goes on to reflect on the entire “event,” feeling particular torment over the “silly smile” with which he greeted his wife’s reproaches and the way, seeing this, “Dolly shuddered as though in physical pain” (3). He begins to feel “despair,” unable to answer the question “What is there to do? ” In Being and Time, Martin Heidegger draws a valuable distinction between the uniform, objective time of clocks and the subjective temporality of human experience; this is parallel with a distinction between the objective space of maps and the subjective spatiality of human activity. The opening of Anna Karenina brings home this distinction sharply. Spatiality is perhaps the more obvious here. In themselves, rooms are simply organizations of space. Objectively, Stiva’s location is simply a matter of a physical body located at a particular point relative to other physical bodies. But the spatial experience of Stiva is quite different from this. Stiva understands his location by contrast with where he should be, where he would like to be, where he would be if everything were right. Jean-Paul Sartre refers to this experience as nothingness. Stiva’s location is not only a matter of where he is, but equally of where he is not. My first contention here is that spatiality, the “existential” experience of location, is fundamentally an emotional experience. As my characterization of Stiva’s place already suggests, nothingness—the judgment of where one is not, should be, should have been—is, first of all, a function of what one feels about locations. In this case, there are two aspects to the feeling. The first is not precisely emotion per se, but rather forms the baseline from which emotions arise. This is normalcy. More often than not, emotions are a response to changes in what is routine, habitual, expected. We anticipate normalcy unreflectively. When our anticipations are violated, attentional focus is triggered (see, for example, Frijda 272-73, 318, and 386 ) and a sort of pre-emotional P ATRICK C OLM H OGAN 46 arousal occurs, an arousal that often prepares for a particular emotion (see, for instance, Simpson 692). It is only when Stiva puts down his feet toward the expected slippers and reaches out toward the expected robe, the moment when he finds the nothingness where the slippers and the robe should be, that his attention is focused. In this case, the focus is recollective; it is a matter of memory—and that increased attention carries in its train the entire sequence of happenings that pushed this body from his wife’s bed to the couch in the study. This leads to the second aspect of feeling that bears on our experience of space. Our experience of the world is not uniform. It is focused on particular areas. The center toward which we tend, and against which we experience all other places, is home. I am not simply referring here to the building we call “home,” as when we “go home” at the end of the day. Rather, I am referring to the location that, paradigmatically, both is home (in the sense of the origin and end-point of journeys) and, so to speak, “feels like home.” Thus it is a point of cognitive orientation (“Where is the restaurant? ” “About a five minute drive from your home”) and a point of emotional ease and security. The idea is not merely phenomenological. There are neurobiological reasons for “place attachment,” as it is called. Indeed, the same subcortical structures appear to be involved in place attachment as in attachment to persons, leading the affective neuroscientist, Jaak Panksepp, to suggest that perhaps “the ancient mechanisms of place attachment provided a neural impetus for the emergence of social attachments” (407 n.93). Deviation from normalcy or removal from home may have different valences. But leaving home and normalcy is always a matter of risk— specifically, emotional risk. That is what attentional focus responds to—risk, both threats and opportunities. Because leaving home and normalcy involves risk, it involves emotion as well, at least potentially. Conversely, being at home and surrounded by what is routine involves the avoidance of risk. This too can give rise to emotion, if risk is expected. In short, the spatiality of human being-in-the-world (to use the Heideggerian idiom) is a sort of emotional geography that develops out of fundamental human propensities toward organizing the world along two fundamental axes: attachment and normalcy. Normalcy in turn guides anticipation, which itself affects attentional focus and pre-emotional arousal, producing a state of emotion readiness. Similar points may be made about the temporality of our being-in-theworld. It too is organized - or, more precisely, encoded - emotionally. Encoding is the process whereby we select, segment (or “chunk”), and give preliminary structure to our experience. There are many different ways in which we encode aspects of our experience. These depend on current interests, expectations, contextual relations (e.g., figure/ ground relations in perception), and so forth. Moreover, there are different levels of encoding. Most basically, there is the perceptual encoding that gives us our sensory experience of the world. This is a function of the sensitivity of sensory neurons (e.g., visual Before Stories: Anna Karenina and the Emotional Structure of Lived Time 47 neurons, with their sensitivities to colors or particular orientations of lines), processes such as lateral inhibition (in which neurons surrounding a highly activated neuron are inhibited; this results in, for instance, the sharpening of lines in vision), etc. There are also higher levels of selection, segementation, and structuration. Some of these are self-conscious. In those, self-conscious cases, we may refer to the processes as “construal” or, even more broadly, “interpretation.” Though only partially recognized by theorists of emotion, it seems that our emotional encoding of experience too occurs repeatedly and at different levels. For example, there is a very basic level of emotional encoding that is bound up with perceptual encoding. Joseph LeDoux has argued that there are two perceptual streams, one of which goes directly to subcortical emotion-systems, while the other goes to cortical areas. Thus a potentially threatening movement in our vicinity might activate a circuit connected to the amygdala, thus generating fear. At the same time, a more informationally rich encoding of that experience may be sent to cortical areas, which may then inhibit or enhance the amygdala response. Put differently, in the “low road” (as LeDoux calls it), particular sorts of motion and proximity (commonly in relation to expectation) are selected by our sensory systems and given a tentative structure in relation to fear. At this fundamental level, emotional encoding is directly part of perceptual encoding. As we have seen, one aspect of emotional encoding is the encoding of space for home versus not home and familiar versus unfamiliar or expected versus unexpected. While the last has a geographical or spatial aspect, it should already be clear that it has a temporal aspect as well. When the expectation is a matter of slippers and a robe being absent, then the expectation is more spatial. When it is a matter of some movement in the nearby bushes, then it is more obviously a matter of time. Indeed, it seems plausible to assume that the encoding of occurrences as expected or unexpected is the first element or component of emotional temporality. I say “first element or component” because unexpectedness only excites attentional focus and gives rise to a pre-emotional state of increased arousal. The development of emotional experiences themselves involves further encoding of experiences as well. Consider, again, Stiva’s conflict with Dolly. As Stiva recalls this highly emotional experience (or sequence of experiences), a few components stand out sharply. The first is the image of Dolly with the letter, “looking at him with an expression of horror, despair, and fury” (2). What is striking here is that this immediately divides the experience into an almost atemporal focus and a broader periphery. The focal moment is in some way the key experience. It is the point that defines the emotion, or that crystallizes it. There is, of course, a larger event here. Dolly presumably moves and speaks; so does Stiva. The occurrences that constitute the event involve change. But this key moment is compressed, almost frozen. There is at least one other compressed, emotionally crucial moment as well--Stiva’s “silly smile” (3). Here, too, there must have been many occurrences surrounding the smile. But Stiva seems almost fixated on this one act, which he P ATRICK C OLM H OGAN 48 deeply regrets. In short, the temporal landscape here is uneven, and the unevenness is emotional. Some of the occurrences that constitute this sequence—itself forming a sort of whole or episode--are highly salient and consequential; others are downplayed. It is worth considering just what gives these particular moments their importance or, more precisely, what leads to this particular encoding of this sequence. Here, then, is the episode as Stiva recalls it. Stiva is returning home from the theater. His mood is “cheerful” and he carries “a huge pear for his wife.” He looks for her, sees her “in the bedroom holding the unlucky note” (2). She asks him, “What is this? This? ” (3). He responds with his smile. Her reaction follows. It is difficult to say how long this sequence took. There is a “torrent” of words that may have lasted ten seconds or ten minutes. Either way, the entire matter is reduced to the key emotional moments isolated above. We may refer to these moments as “incidents.” Incidents, as I am using the term, are the focal points of emotional response, the minimal units of emotional temporality. Moreover, they seem to operate through something akin to lateral inhibition. Specifically, the isolation of an incident appears to reduce the saliency of surrounding occurrences. (I will use “occurrence” as a neutral term, indifferent as to duration, structure, etc.) In this way, incidents stand out more sharply from the sequence of occurrences than is warranted by the occurrences themselves. Finally, incidents serve as the nuclei of “events.” An “event,” as I am using the term, is the next level of temporal segmentation, encompassing a cause and response to an incident. Events themselves compose episodes. An “episode,” in this sense, is a series of events which begins and ends in temporary normalcy. Above the episode, we have stories. A story begins and ends in permanent normalcy. To understand the segmentation of incidents here, and particularly its emotional nature, it is helpful to establish the context for Stiva’s meeting with his wife. One important aspect of emotional response is that our experience of emotion does not operate on an absolute scale. It is in part a function of the gradient of change from one moment to the next. To some extent, this is a matter of our prior mood. But it also crucially involves our anticipations, including—perhaps most importantly--our tacit anticipations (see my “Sensomotorische Projektion, Kontinuitätsbrüche und Emotion” and “Sensorimotor Projection”). In keeping with this, Tolstoy informs us about Stiva’s positive mood. Moreover, he suggests that Stiva anticipated enhancing his cheerfulness by presenting the pear to his wife and receiving a warm response from her. The oral, sensual nature of this particular piece of fruit also suggests that he envisioned at least a playful physical response from her. (Contrast, say, the gift of a book or theater tickets or a new hat.) This is not to say that we should envision Stiva as self-consciously thinking out all this. Rather, insofar as we imagine Stiva as a full person, we must imagine him as having a series of implicit anticipations. In this case, some such anticipations would have contributed to his motivation in bringing the pear. Given all this, we may begin to understand the isolation of this particular incident. When Before Stories: Anna Karenina and the Emotional Structure of Lived Time 49 Stiva sees Dolly holding the incriminating letter, he sees a face filled with “horror, despair, and fury” (2). This hostility directed at Stiva provides perhaps the most striking contrast with his own mood and, more significantly, his tacit anticipation of Dolly’s response to the pear. But why does Stiva (or Tolstoy) isolate just these aspects of the scene— Dolly’s expression and the letter itself? There is reason to believe that certain features of the world and of our own bodily experience are encoded almost immediately as emotion triggers (see chapter seven of my Cognitive Science and citations). I have already mentioned the encoding of, say, unexpected movement in the nearby bushes (thus occluded motion) and the direct activation of the amygdala, thus fear. This is a specific case of a more general tendency. We are ready or prone to encode certain sorts of experience as emotion-relevant. We have particular sensitivity to certain features of experience. When we encounter certain conditions or occurrences (or, to a lesser degree, when we imagine them), that encoding is triggered and the relevant emotion system is activated. Some of these sensitivities are innate. In particular, a good case can be made for considering one large set of emotion sensitivities as innate, and as particularly important for our emotional development. These are sensitivities to the “expressive outcomes” of emotion. Before continuing with this, we need to clarify the components of emotion. The most basic elements of an emotional experience are the following. First, there are eliciting conditions. These are the situations, occurrences, properties to which we are sensitive in emotional encoding and that serve to activate emotion systems. Second, there are expressive outcomes. These are the manifestations of an emotion that mark the subject as experiencing an emotion. They range from vocalization to facial gestures to postural changes to perspiration. A third component of emotion is actional response or what one does in reaction to the situation. This action may be undertaken to maintain a desirable situation or to alter an aversive situation. For example, when faced with an angry stranger, I might look for an escape route and begin to run away. A fourth component is the phenomenological tone, or just what the emotion feels like. This is important in part because it motivates us to sustain or change the situation by making us experience it as desirable or aversive. (We might isolate other components as well. However, these will suffice for present purposes.) A remarkable fact about expressive outcomes, briefly noted above, is that they are also eliciting conditions for emotion. Depending upon the nature of the emotion and its orientation, an emotional expression may affect us in a parallel or a complementary way. If I am out on a dark night in the woods and my companion gasps in fear, I will feel fear—a parallel emotion triggered by my companion’s emotion expression. When I meet a stranger in a dark alley and he looks at me with anger, I am also likely to feel fear, but in this case the fear will be a complementary emotion triggered by the expression of anger. Again, our emotional response to emotional expressions appears to be largely innate. Indeed, it is arguable that we do not need to posit P ATRICK C OLM H OGAN 50 much else in the way of innate emotional triggers. Our emotional sensitivity to the emotional expressions of others serves to guide us in just what we should fear, what should make us happy, and so on. As Margaret Atwood put it, “what to feel was like what to wear, you watched the others and memorized it” (131). For example, for some time, it was thought that fear of snakes is innate in monkeys. In fact, it may be that it is not innate per se. Rather, the fear of snakes is in effect “acquired” by monkeys when, at a young age, they experience their mother’s fear of snakes (see Damasio 47). It may still be that there is an innate fear-related sensitivity to snakes making this fear particularly easy to acquire. But, whether or not it is sufficient, the experience of someone else’s fear expression seems to be necessary in this case. On the other hand, there is something of a problem here. Suppose I am a small child, just “acquiring” emotional sensitivities. I see three people in a given situation. One is angry; one is afraid; a third is neutral. What governs my emotional acquisition in this case? There are probably several factors. For example, we may be more sensitive to emotions than to neutrality and fear may be more contagious than anger. Moreover, orientation of the emotion (signaled by line of sight, posture, and other expressive factors) is obviously crucial; anger oriented toward me is something very different from anger oriented toward someone else. But there is another important factor as well, already suggested in the preceding account. As I mentioned, monkeys acquire fear of snakes from their mothers. This indicates that there is some isolation of particular individuals whom we trust—or, more precisely, individuals with whom we have formed a basic attachment relation. We acquire our emotional sensitivities disproportionately from these individuals—most often, our parents and, of these, most commonly our mothers. For example, in the preceding scenario (of the three people exhibiting different emotional expressions), I am more likely to feel what is relevant to my mother in this context—fear, if she is afraid of something in the environment; anger, if she is angry at something; amusement if she is laughing, and so on. Thus it is no accident that Stiva experiences a particular emotional spike in connection with the facial expression of his wife, an attachment figure (on spikes in an emotion experience, see Smith). Note that, in this case, the response is partially complementary and partially parallel. Stiva feels “despair” in parallel with her “despair” (2, 3). But he feels fear and shame in the face of her “fury” and “horror” (2). No less importantly, Stiva remarks on her posture and movement. These too are expressive. He had tacitly imagined her “bustling about” in the sort of motion that would fit with his own “cheerful” mood. He is faced, instead, with her “sitting motionless” (2), thus in an attitude in keeping with her despair—the one parallel emotion that he experiences (2, 3). (The description of Dolly as motionless suggests that, at this point, despair is her predominant emotion, with the fury rather subdued.) So, the first incident of this episode marks a moment of sharp emotional change, due to a sharp change in expectations, a change triggered by emotional expressions from an attachment figure. It is becoming clear why so Before Stories: Anna Karenina and the Emotional Structure of Lived Time 51 much else in the larger occurrence fades into obscurity. However, we are still left with one unexplained detail—why the letter? There are several reasons why the letter is important. I will address only one. When we experience an emotional spike, our cognitive response is partially automatic. Crucially, we shift our attentional focus. But just what do we shift our attentional focus to? It is relatively simple—causes, or possible causes. Almost immediately upon experiencing an emotion, we attribute a cause. More precisely, we narrow our attentional focus to candidate causes, attributing a cause very quickly. This is crucial because causal attribution is a necessary prerequisite for any actional outcome. Put very crudely, if I am afraid, I need to know what I am afraid of in order to run away in the right direction. It may seem that this is just obvious. We just know directly what causes our emotion. Indeed, it may seem that we would not even have the emotion if we were unable to attribute a cause. But, in fact, our causal attributions are highly fallible. Later in the novel, Tolstoy makes the point that “Sometimes [Anna] had no idea of what it was that she was afraid of and what she wanted. Whether she was afraid of what had happened, and wanted that, or of what was going to happen, and wanted that, or just what it was that she wanted, she had no idea” (345). Research suggests that Anna’s state is far less remarkable than it might seem. The condition is unusual only in the fact that Anna recognizes that she does not know the cause of her fear or of her desire. We never directly know these causes. We must always infer them. Most often, however, we do not realize that we have to do this. We think that we know. As Frijda explains, “One knows, generally, that one has an emotion; one does not always know why, and what exactly makes one have it; and if one does know, it is a construction, a hypothesis, like those one makes about the emotions of someone else” (464). Moreover, even when “correct,” our causal attributions are only partial. I may feel sad for many reasons. But I will be inclined to isolate only one or two—the most obvious or salient ones, but not necessarily the most significant ones (see Clore and Ortony 27, Zajonc 48, and Damasio 75). Of course, we usually do get our causal attributions roughly right. Moreover, in most cases, they are relatively simple. Indeed, they are bound up with normalcy. I am going along in the usual way, then receive a letter explaining that I made a mistake in my income tax return and, rather than receiving a $392 refund, I actually owe $4,723. Later in the day, I feel unhappy. I attribute it, not to the ordinary things that occur every day, but to the unusual and unexpected occurrence. One important, simplifying aspect of our automatic and immediate causal attribution is that we tend to stop with the most proximate cause. We may elaborate on the causal sequence subsequently and self-consciously. But we tend not to do this spontaneously. The one clear and consistent exception to this is the case of emotions that result from other people’s emotional expressions. We may refer to these as “expression-triggered emotions.” In those cases, it is crucial that we attribute a cause, not only to our emotions, but to the emotions of the other person as well. If my friend shouts in fear, it is his P ATRICK C OLM H OGAN 52 shout that causes my fear. But I do not want to run away from his shout or from him. Rather, I want to run away from, say, the bear he has spotted. Indeed, in the case of expression-triggered emotions, causal attribution may in some cases precede the full communication of the emotion, for that causal attribution is often critical to the determination of whether one’s emotional response is parallel or complementary. Here we may return to the letter. The letter figures importantly because it is the salient object for causal attribution. Its salience is enhanced by Stiva’s expectation of handing Dolly the pear, an expectation that would draw his attention to anything she is holding, and of course by her following question, with its demonstrative repetition (“What is this? This? ”), presumably accompanied by appropriate movement. Indeed, the question itself is included in Stiva’s recollection primarily because it converges with Stiva’s attentional focus on the letter as the cause of Dolly’s emotion. But, of course, this cause does not in itself give the crucial information about the direction of Dolly’s emotion. Thus we need to add a further sort of inference here. In expression-triggered emotions, we need to isolate the target of the other person’s feeling as well as the cause. The two are often the same, but they need not be. We commonly isolate the target of an emotion by isolating the target of a trajectory of motion, a look, a verbal address. In the case we have been considering, the emotional target is, of course, Stiva, for Dolly was “looking at him” with her “expression of horror, despair, and fury” (2). This target-isolation also makes Stiva salient for himself. Indeed, recognizing oneself as the target of someone else’s attention regularly makes one self-conscious, makes one aware of oneself as an object. This self-awareness may have different foci. For example, it may be a matter of physical beauty or status markers (as in dress), racial group or religious affiliation (if this is indicated by visible signs), or something else. For our purposes, it may most importantly be a matter of one’s own emotional expressions, particularly in a context where emotional expressions are already salient, as here. In keeping with this, the second incident in this episode is Stiva’s experience of his own facial gesture, his “silly smile” (3). Before going on to this, however, we need to return to the hierarchical organization of temporality. The treatment of causal attribution begins to move us from the incident (Dolly motionless in the chair with the expression of horror, despair and fury) to the encompassing event. As I briefly indicated above, the event includes the proximate cause and what we might call the “situational response,” the immediate actional outcome. Consider a paradigmatic scenario for fear. A predator appears unexpectedly within my field of vision. I feel fear—attributing this (accurately) to the presence of the predator. Since this is a highly aversive situation, my unreflective response is to try to alter it. Since it is fear in particular, my response is to try to run away. I look around for a path that will take me away from the predator. The situation for Stiva is similar, though more complex. It is highly aversive. Thus his spontaneous impulse is to change the situation or his relation to it. This gen- Before Stories: Anna Karenina and the Emotional Structure of Lived Time 53 eral response is particularized by the specific emotions involved—the shame, despair, and fear or anxiety. The difficulty for Stiva is that this combination of emotional responses does not point toward any clear actional outcome. Shame counsels concealment. Shame is bound up with our acute sense that we are the target of someone else’s aversion, particularly their disgust, physical or moral. Our immediate response to this is the wish to hide. Note that this sense of shame further enhances Stiva’s self-consciousness about himself as the target of Dolly’s attention. This self-consciousness is likely to make him intensely aware of his own emotional expressions. The salience of these expressions increases the likelihood that he will isolate them in causal attribution for any subsequent incidents in Dolly’s emotional response. Thus Dolly’s subsequent shudder, “torrent” of words, and exit might in principle have been provoked by any number of things. Stiva’s self-consciousness at this moment makes his own emotional expression—in this case, the facial gesture—highly salient for him, thus a likely object for causal attribution. It may then, in turn, become isolated as the focus of his own intense regret, thus becoming an incident in a separate emotional event. In any case, Stiva’s experience of shame points toward an actional outcome of hiding. However, it may also point to an actional outcome of appeasement or submission. Many emotions involve more than one possible actional outcome. Typically, they are hierarchized—in fear, first try flight, but, if that does not work, try to fight (or, in certain cases, freeze). More generally, actional outcomes are bound up with spontaneous, rapid, unselfconscious possibility assessments. An assessment that one option is not viable shifts one’s response down the hierarchy of actional alternatives. In some cases, that shift can be repeated until no options remain. Tolstoy is particularly sensitive to our sense of such complete impossibility. Thus, when Stiva first begins to recall his situation, he thinks of “his impossible position” (2). He experienced the first shift in the direction of such impossibility in relation to shame when he initially saw Dolly holding the letter, for he could not hide. This left an appeasement response as his only option. We will return to this below. Stiva finds himself in a similar position with regard to the fear or anxiety. He could, of course, flee the immediate situation. However, even in the short term, that would not prove much of a solution. Where would he go? He could not simply remove himself to another part of the house, since Dolly could follow. If he left the house, where would he spend the night? I do not believe that this is a matter of long-term calculations, as the preceding questions might suggest. The exclusion of flight as a possible response is much the same as would occur in cases of physical threat when all escape routes pose dangers (the predator could catch me if I go right, but there is a cliff to left). A subsequent response option is to fight. But that too is not possible here, in part because Dolly’s “fury” has not manifested itself in physical aggression. The only possible fight would be a verbal conflict over Stiva’s guilt— P ATRICK C OLM H OGAN 54 “making denials or excuses” (3), as Stiva puts it retrospectively. But Stiva is guilty; the letter proves it. More importantly, his own sense of shame inhibits his fight response along precisely this axis of dispute. In addition, the fight response relies on a feeling of anger being connected with the anxiety. However, Stiva’s sense of his own guilt inhibits this response. Stiva “couldn’t manage to put on the right expression . . . now that his guilt was exposed.” There are several possibilities for that “right expression.” One involves “acting offended” (3), thus a form of anger, but, again, that is emotionally excluded. Freezing too is an option within the hierarchy of fear outcomes. However, it is preferential only in cases where one is not already the target of gaze, motion, and/ or address from the threatening agent. None of this is the case for Stiva. On the other hand, freezing tends to occur when no other actional outcome is initiated and this partially happens with Stiva. Finally, there is the option of appeasement. This opposes both flight and aggressive motion and thus may partially converge with freezing. Still, as in the case of shame, it is not clear that this provides a solution. The despair felt by Stiva only compounds the problem. Despair leads to just the sort of motionlessness that Stiva sees in Dolly. It is bound up with Stiva’s possibility assessments—or, more properly, impossibility assessments: “What is there to do? He said to himself in despair, without finding an answer” (3). The despair inhibits concealment, flight, conflict—or even much in the way of appeasement. It is most obviously compatible with submission, but submission of a very minimal sort—submission that does not really go beyond the absence of a refusal to submit. What, then, does Stiva do? What is the actional outcome of his emotion, his response to this provoking incident, the outcome that results from the interaction of these emotion systems and situational constraints? It is his silly smile. It is a form of emotional expression that serves generally to indicate benevolence. It is a “goodhearted” smile (3). The evolutionary reflections of writers such as Panksepp suggest that the smile is not entirely submissive, but incorporates a response to threat as well. As Panksepp explains, “The smile . . . probably harks back to ancient mammalian threat displays.” In humans, the probable evolutionary adaptation behind the display [of teeth] is that the potentially tense situation will require no further action if one smiles. The human smile may have evolved . . . to communicate that one is basically friendly but quite capable of dealing with any difficulties that may arise (287). Though this fits Stiva’s situation, I am skeptical about this analysis. Izard and Ackerman seem to me more accurate in stating that “The smile has the capacity to operate as a universally recognizable signal of readiness for friendly interaction” (258). Brody and Hall note that smiling “is socially useful in that it puts others at ease” (346). It is difficult to reconcile this with the idea that smiling communicates threat. In keeping with Izard, Ackerman, Brody, and Before Stories: Anna Karenina and the Emotional Structure of Lived Time 55 Hall, then, it seems that we are most justified in taking Stiva’s smile as something like an appeasement response. Tolstoy later makes this use of smiling explicit. Seryozha is a poor student, yet he “always conquered” his tutor with a “gay and affectionate smile” (625). Indeed, Stiva himself later has an unpleasant exchange with Karenin about Anna. In response to Karenin’s anger, Stiva’s face assumes “an embarrassed smile” as he says, “forgive me then, forgive me if I’ve wounded you” (867). Stiva’s response to Dolly is an expressive and actional outcome of the same sort. (Of course, it is problematic as such, since being an unfaithful husband is more emotionally divisive than being a bad student.) At the same time, it is more than this. First, it is a sort of default facial expression for Stiva. It is his “usual” expression (3), by implication an expression that his face assumes when there is no strong emotional reason for it to assume any other expression—or when there are conflicting emotional pressures that are not easily resolved. Perhaps most importantly, the smile suggests Stiva’s own attitude toward himself. The target of Stiva’s smile is in part Dolly, suggesting his goodhearted friendliness. But it is in part himself. Again, at this point, he is highly self-consciousness. With regard to himself, this smile suggests his sense of his own ludicrousness. He feels the smile to be “silly” because he feels himself to be silly. It may, then, be consistent with a sense of shame. Later in the novel, Tolstoy reports something of just this sort. Specifically, Levin smiles in such a way as to express the feeling that he is at once ridiculous and blameworthy: “Levin smiled at his own thoughts and shook his head at them disapprovingly; a feeling something like remorse tormented him” (581). One obvious problem with this as an appeasement response, however, is that a smile is ambiguous (as the preceding comments already indicate). Beyond a desire for reconciliation or a sense of one’s own ludicrousness, it could suggest self-contentment, amusement, or mockery. Here we need to shift to Dolly’s perspective. She anticipates some strongly negative emotional response on Stiva’s part—perhaps remorse, perhaps anger, but certainly something clearly indicating that he finds the situation emotionally aversive. In the context of this expectation, any sign of a positive affect is, at best, disconcerting. In Dolly’s case, it pushes her complex emotions in the direction of fury (rather than horror or despair). This is compounded by the fact that his smile would tend to frustrate Dolly’s preferences as well. Specifically, her preference—based in part on attachment—would be that he is not guilty (in which case, he would respond with anger) or that, if guilty, he begs forgiveness. Apparent amusement frustrates both desires—and frustration is a primary trigger for anger. Before turning to Dolly’s response, however, we need to conclude something about the occurrences discussed thus far. We have just outlined an event. We have the emotion-provoking incident (Dolly, motionless, looking at Stiva with horror, despair, and fury), causal attribution (the letter), target isolation for the emotional expression, and the expressive and actional out- P ATRICK C OLM H OGAN 56 comes. Stiva experiences this as an event; Tolstoy depicts it as such, and we presumably experience it as such as well. This event is causally linked to a second event. In this case, Stiva’s expressive and actional outcome (the smile) serves as an emotion-provoking incident for Dolly. This makes Dolly angry, leading to the expressive outcome of a shudder, followed by the actional outcome of shouting and leaving the room. Dolly plays out a more ordinary sort of anger response here. She experiences something akin to “physical pain” (3) and responds with aggression—verbal rather than physical. When the aggression is complete, she leaves, flight following fight in a fairly predictable way. Thus we have a second event. I will not consider this in great detail. However, it should be clear that the incidents arise in much the same way here as in the first case and that the linked event forms by the same principles as well. For Dolly, Stiva’s facial expression is a highly emotionally provocative violation of expectation and, if construed as mockery or even amusement, a violation of attachment-based preference as well. The target isolation is obvious (we can assume Dolly takes him to be smiling at her). Finally, the actional and expressive outcomes are clear. It is not at all necessary that two people should experience the same occurrences as incidents or the same complexes of occurrences as events. Indeed, there will invariably be some differences. However, in this case, the incident for Stiva is the same as the incident for Dolly (at least as Stiva infers this as an emotional cause)—his silly smile. Both seem to view the subsequent shudder and departure as the expressive and actional outcomes of Dolly’s response to that incident. (Tolstoy, presumably reflecting Stiva’s understanding, reduces Dolly’s words to their expressive aspect, noting their violence and bitterness, but not their semantic content.) Dolly’s departure, the end of this second event, marks an interruption in the causal sequence, if only temporarily. That is why this sequence of two events constitutes an episode. For example, that is why it makes sense for Tolstoy to end the first chapter with Stiva, “in despair,” thinking “But what can I do? What is there to do? . . . without finding an answer” (3). Prior to the first incident, we have ordinary life, normalcy. After Dolly leaves, there is a sense in which that situation is restored. For example, Stiva is able to go to sleep and to wake up expecting to find his robe and slippers. He would not be able to do this if Dolly were in the room, questioning him about the letter or berating him. But, at the same time, it is clear that the emotions aroused by these incidents and their encompassing events will have further expressive and actional consequences. In other words, it is clear that further events will follow in the same emotional/ causal sequence. Indeed, the incident that provoked Stiva’s regret has not been developed into a full event. The possibility of any satisfactory actional outcome (such as a plea for forgiveness) has been lost by Dolly’s departure. The temporary return to normalcy marks this as a structural unit of emotional history. But it is not a story, because the enc- Before Stories: Anna Karenina and the Emotional Structure of Lived Time 57 hainment of emotional causes initiated by the opening incident—and sustained most forcefully by the final, unacted incident--will continue. I indicated above that a story structure occurs when we move from normalcy to emotion-provoking incidents embedded in events and developed in causally-linked episodes, all followed by a return to normalcy. The difference between an episode and a story is simply that the normalcy is enduring in a story. For example, in a romantic plot, the lovers are married and live happily ever after. In these cases, there is no longer any need for further actional outcomes. In comic stories, the chain leads to a non-aversive situation, a situation that is sustained by normalcy and that we emotionally wish to be sustained. Alternatively, it can turn out that there is an aversive situation, but one that it is genuinely impossible to alter. In other words, it may turn out that no actional outcome will ameliorate conditions. In this case, we have a return to normalcy also. The causal sequence ends because an unalterable outcome has occurred. This is, obviously, the tragic version of a story. In a romantic plot, we find it when, for example, one of the lovers dies. Novels often have one or more extensively elaborated stories that span much of the work as well as many briefer stories that are more localized. In the case of Anna Karenina, the two extended stories are both romantic—the story of Anna and Vronsky (which ends with Anna’s death, thus tragically) and the story of Levin and Kitty (which ends in the couple’s ordinary life together, thus in a non-aversive return to normalcy—though there are complications here due to the intertwining of this story with another narrative regarding Levin’s spiritual self-realization). The various stories that comprise a novel are often related to one another in treating similar relations and themes. The rift between Stiva and Dolly that opens the novel is an obvious case of this sort, paralleling, as it does, some aspects of the two main plot lines. More exactly, despite Stiva’s feeling, at the end of chapter one, that his situation is impossible, it is not really impossible. It is apparently impossible in the short term, but not in the long term. Again, it is clear that the incidents will continue to provoke emotions, actions, and further incidents. The issue is whether there is anything that will produce a “resolution,” either a comic, non-aversive situation which demands no further action, or a tragic situation that could never be altered. In principle, there are several possibilities here. Tolstoy offers one option in particular. Indeed, it is named by Stiva in the course of this episode, and eventually taken up by him—“asking forgiveness” (3). A recurring motif in the novel is that human emotional conflict requires continual forgiveness—the guilty parties seeking forgiveness, the offended parties granting it. We are oriented toward this from the epigraph to the novel, “Vengeance is mine; I will repay, saith the Lord” (xv). Punishment should be in the hands of God alone. The tragedies of human life are often tragedies of emotion. In Tolstoy’s novel, forgiveness has the highest place among responses to emotion-inducing occurrences. Indeed, it is itself an emotional response, and one that is not unrelated to attachment—as we see when Dolly forgives Stiva later on, when Levin forgives Kitty for reject- P ATRICK C OLM H OGAN 58 ing his first proposal, and when Kitty forgives Levin for his pre-marital affairs. Forgiveness is what rounds out this particular episode, turns it into a story, brings the life of Stiva and Dolly back from this aberration and rift into normalcy. The normalcy is, of course, not untroubled. But that normalcy, and its troubles, are no longer part of this particular story. In sum, our experiences of both space and time are encoded nonhomogeneously. The principles by which objects and occurrences are selected, the principles by which they are segmented, and the principles by which they are structured, both internally and in embedded hierarchies, are crucially (though of course not exclusively) emotional. In relation to this, our sense of normalcy, our expectations, our experience of other people’s emotional expressions, our attachment relations with those people, and our tacit possibility assessments are of particular importance in our organization of emotional geography and, even more, emotional history. Finally, the hierarchy of that organization involves almost atemporal emotional causes or incidents, “thicker” structures of causal attribution and actional outcome (events), episodes or sequences of causally linked events that move from normalcy to temporary normalcy, and, finally, stories that conclude the sequence of episodes with enduring normalcy. Works Cited Atwood, Margaret. Surfacing. New York: Fawcett Crest, 1972. Brody, Leslie R., and Judith A. Hall. “Gender, Emotion, and Expression.” Handbook of Emotions. Ed. Michael Lewis and Jeannette M. Haviland-Jones. 2nd ed. New York: Guilford Press, 2000. 338-49. Clore, Gerald L., and Andrew Ortony.“Cognition in Emotion: Always, Sometimes, or Never? ” Cognitive Neuroscience of Emotion. Ed. Richard D. Land and Lynn Nadel, with Geoffrey L. Ahern, John J. B. Allen, Alfred W. Kaszniak, Steven Z. Rapcsak, and Gary E. Schwartz. Oxford: Oxford University Press, 2000. 24-61. Damasio, Antonio. Looking for Spinoza: Joy, Sorrow, and the Feeling Brain. Orlando, FL: Harcourt, 2003. Frijda, Nico H. The Emotions. Cambridge: Cambridge University Press; Paris: Editions de la Maison des Sciences de l’Homme, 1986. Heidegger, Martin. Being and Time. 1927. Trans. John Macquarrie and Edward Robinson. New York: Harper and Row, 1962. Hogan, Patrick Colm. Cognitive Science, Literature, and the Arts: A Guide for Humanists. New York: Routledge, 2003. —.“Sensomotorische Projection, Kontinuitätsbrüche und Emotion im Filmerleben.” Trans. Jens Eder. Audiovisuelle Emotionen: Emotionsdarstellung und Emotionsvermittlung durch audiovisuelle Medienangebote. Ed. Anne Bartsch, Jens Eder and Kathrin Fahlenbrach. Köln, Germany: Herbert von Halem Verlag, 2007. 361-81. —.“Sensorimotor Projection, Violations of Continuity, and Emotion in the Experience of Film.” Projections: The Journal for Movies and Mind 1.1 (2007): 41-55. Izard, Carroll E., and Brian P. Ackerman.“Motivational, Organizational, and Regulatory Functions of Discrete Emotions.” Handbook of Emotions. Ed. Michael Lewis and Jeannette M. Haviland-Jones. 2nd ed. New York: Guilford Press, 2000. 253-64. Before Stories: Anna Karenina and the Emotional Structure of Lived Time 59 LeDoux, Joseph. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New York: Touchstone, 1996. Panksepp, Jaak. Affective Neuroscience: The Foundations of Human and Animal Emotions. New York: Oxford University Press, 1998. Sartre, Jean-Paul. Being and Nothingness. 1943. Trans. Hazel Barnes. New York: Washington Square, 1966. Simpson, Joseph R, Jr., Wayne C. Drevets, Abraham Z. Snyder, Debra A. Gusnard, and Marcus E. Raichle. “Emotion-Induced Changes in Human Medial Prefrontal Cortex: II. During Anticipatory Anxiety.” PNAS 98.2 (16 January 2001): 688-93. Smith, Greg M. “Local Emotions, Global Moods, and Film Structure.” Passionate Views: Film, Cognition, and Emotion. Ed. Carl Plantinga and Greg Smith. Baltimore, MD: Johns Hopkins University Press, 1999. 103-26. Tolstoy, Leo. Anna Karenina. 1877. Trans. Joel Carmichael. New York: Bantam, 2006. Zajonc, Robert B. “Feeling and Thinking: Closing the Debate Over the Independence of Affect.” Feeling and Thinking: The Role of Affect in Social Cognition. Ed. Joseph P. Forgas. Cambridge: Cambridge University Press, 2000. 31-58.