THE CONCORDANT DISCORDANCE of emplotment is curiously and intricately correlated to the decentered, asynchronous temporality of the brain. One of the many ways in which the brain differs from a computer is that its temporal processes are not instantaneous and perfectly synchronized. Unlike electrical signals that discharge simultaneously at a fraction of the speed of light, action potentials at the neuronal level take more than a millisecond to fire, and different regions of the cortex respond at varying rates.1 For example, as Semir Zeki (2003, 215) observes, in the visual cortex “colour is perceived before motion” by approximately 80 milliseconds, and “locations are perceived before colours, which are perceived before orientations.” The integration of neuronal processes through which conscious awareness emerges may require up to half a second. As Zeki points out, however, this “binding” (as it is called) is itself not perfectly homogeneous: “The binding of colour to motion occurs after the binding of colour to colour or motion to motion” because “binding between attributes takes longer than binding within attributes” (216, 217). More time is needed to integrate inputs from vision and hearing, for example, than to synthesize visual signals alone. Although we typically don’t notice these disjunctions, the nonsimultaneity of the brain’s cognitive processes means that consciousness is inherently out of balance and is always catching up with itself. As Antonio Damasio (1999, 127) puts it, “We are probably late for consciousness by about 500 milliseconds.”
This imbalance is not a bad thing because it allows the brain to play in the ever-changing horizonal space between past patterns and the indeterminacies of the future, the space that plots organize into beginnings, middles, and ends. Concord with no trace of discord would be disabling. In waking life, as Gerald Edelman and Giulio Tononi (2000, 72) observe, “groups of neurons dynamically assemble and reassemble into continuously changing patterns of firing.” The synchronization of brain waves across the cortex makes possible the formation of neuronal assemblies and coordinates the workings of different regions of the brain (see Buzsáki 2006). As Bernard Baars and Nicole Gage (2010, 246) explain, “Normal cognition requires selective, local synchrony among brain regions,” “highly patterned and differentiated” oscillatory patterns in which “synchrony, desynchrony, and aperiodic ‘one-shot’ waveforms constantly appear and disappear.” But, as Edelman and Tononi point out, “if a large number of neurons in the brain start firing in the same way, reducing the diversity of the brain’s neuronal repertoires, as is the case in deep sleep and epilepsy, consciousness disappears” (36). In those conditions, “the slow, oscillatory firing of . . . distributed populations of neurons is highly synchronized globally” (72), and global hypersynchrony paralyzes normal functioning by disrupting the to and fro of synchronization and desynchronization. In contrast to sleep and epilepsy, “consciousness requires not just neural activity,” Edelman and Tononi observe, “but neural activity that changes continually and is thus spatially and temporally differentiated”—“distributed, integrated, but continuously changing patterns of neural activity . . . whose rich functioning actually requires variability” (73, 74–75; original emphasis).
This necessary tension at the neuronal level between pattern and change, synchrony and fluctuation, coordination and differentiation is the neural correlate of the ability of a plot to join concord and discord through temporal structures that order events while holding them open to surprise, variation, and refiguration. The ability to tell and follow stories requires much more than neural activity alone, of course, but if the temporal processes through which neuronal assemblies form, dissolve, and form again were not asynchronous in these ways, our cognitive apparatus probably could not support narrative interactions.
The capacity to play with temporal differences is a defining characteristic of narrative. As Christian Metz (1974, 18) observes, “Narrative is a . . . doubly temporal sequence. . . . There is the time of the thing told and the time of the telling (the time of the signified and the time of the signifier). This duality not only renders possible all the temporal distortions that are commonplace in narratives (three years of the hero’s life summed up in two sentences of a novel . . .). More basically, it invites us to consider that one of the functions of narrative is to invent one time scheme in terms of another time scheme.” The opposition between “discourse time” (sometimes referred to as “Erzählzeit”) and “story time” (“Erzählte Zeit”) gives rise to a series of other distinctions that narrative theorist Gérard Genette (1980) magisterially analyzes in his classic study of Proust’s In Search of Lost Time—differences in order, duration, speed, or frequency that authors manipulate to play with our sense of time. After describing the “narrative competence of the reader” as the ability to negotiate these temporal complications, Genette notes that “this very competence is what the author relies on to fool the reader by sometimes offering him false advance mentions, or snares,” constructing a “complex system of frustrated expectations, disappointed suspicions, surprises looked forward to and finally all the more surprising in being looked forward to and occurring nonetheless” (1980, 77, original emphasis).2 H. Porter Abbott (2002, 3) exaggerates only a little when he argues that “narrative is the principal way in which our species organizes its understanding of time.” Music and dance, to be sure, also serve this end.3 They do so, however, by invoking and manipulating the same disjunctions in the neurobiology of time and in our lived experience of time passing that make narrative possible.
The neurobiological processes underlying our experience of time are somewhat elusive, but the scientific consensus is that there is no central clock or timing mechanism in the brain, just as there is no single language module and no “homunculus” functioning as a central processor. Instead, as Valtteri Arstila and Dan Lloyd point out (2014a, 200), “the constituent processes of vision and hearing (and all the other senses) run on different schedules. Just as there is no one place where ‘it all comes together,’ there is no one time where cotemporal events are simultaneously represented.” Dean V. Buonomano (2014, 330) argues, indeed, that “the realization that temporal processing is not a unitary neural process” is key to understanding “the neural basis of timing.” As he explains, “The brain seems to have developed fundamentally different mechanisms for timing across different timescales” (337).
These mechanisms begin at the neuronal and synaptic level. As Buonomano notes, the responsiveness of neurons is “strongly dependent on their recent history of activity,” just as “the strength of the synapses between neurons” similarly “varies dramatically in a use-dependent fashion,” thereby providing “an ephemeral memory of what has happened in the past few hundred milliseconds” (334). Oscillatory coupling and decoupling of neuronal assemblies across the brain and between the brain and the body provide further mechanisms for processing durations at different scales. As György Buzsáki (2006, 174) observes, “Oscillatory coalitions of neurons can expand the effective window of synchronization from hundreds of milliseconds to many seconds.” This is why Francisco Varela, the foremost neurophenomenologist of time, proposes three “scales of duration” to distinguish different “windows” of integration, from “basic or elementary events” (10–100 milliseconds) beneath conscious awareness, to “large-scale integration” (250 milliseconds to a few seconds) underlying the experienced moment, to the “descriptive-narrative assessments” lasting several seconds that characterize working memory and longer-term syntheses (Varela 1999, 273; see also Thompson 2007, 330–38).
The borders between these “windows” are necessarily fuzzy, and there is some controversy about where to draw them, because the processes of synchronization and desynchronization across the brain, body, and world are always shifting. As Lloyd (2016) observes, “Brain activity, like consciousness, is a radically nonstationary process, constantly changing at every scale.” Or as Varela and his colleagues explain, “In the brain there is no ‘settling down’ but an ongoing change marked only by transient coordination among populations” of neurons (2001, 237). Not unified or centrally organized, the neurobiological bases of time are an array of distributed, fluctuating processes from the microsecond level up to the longer integrations supporting various embodied experiences of different temporal width.
According to Varela, “The fact that an assembly of coupled oscillators attains a transient synchrony and that it takes a certain time to do so is the explicit correlate” and “origin of nowness” (1999, 283). Because any “synchronization is dynamically unstable,” however, and “will constantly and successively give rise to new assemblies” (283), any “now” has what Edmund Husserl calls “horizons”—a retentional horizon carrying along traces of the immediate past and a protentional horizon pointing toward assemblies to come as synchronies organize themselves in recurring (but also changing) patterns that follow particular trajectories.4 The duration of this “now phase” may in turn differ according to the varying scales of integration, from hundreds of milliseconds to multiple seconds, that constitute what Varela calls the “window of simultaneity that corresponds to the duration of [the] lived present” (272).
Our lived experience of time is an emergent phenomenon that is based on but is not identical with or reducible to its neurobiological underpinnings. As Varela and his coauthors explain (2001, 237), “The large-scale integration of brain activity can be considered as the basis for the unity of mind familiar to us in everyday experience.” Beneath this felt unity, however, are a variety of asynchronous, disjunctive processes that support the brain’s ongoing balancing act between pattern and change, order and flux, synthesis and desynchronization. These disjunctions make “instability . . . the basis of normal functioning rather than a disturbance that needs to be compensated for” (Varela 1999, 285), but they are ordinarily invisible. As Bruno Mölder (2014, 222) explains, “Consciously represented temporal properties need not match with the temporal properties of the neural processes that underpin these representations”—but sometimes the disjunctions underlying them come into view. Ordinarily time seems to pass fluently and unproblematically, but occasionally the asynchrony of brain, body, and world may emerge from its cloak of invisibility in various subjective distortions, as when time seems to speed up or slow down according to our level of arousal, attention, or engagement, for example, or in certain temporal illusions like the oft-discussed phenomenon whereby a moving dot assumes the color of a location before it gets there (more on this in the next section). Distortions like these fascinate neurobiologists of time because they provide glimpses of the discrepancies in the processing of sensation that underlie our unified experience of the world.
Temporal experience is an emergent phenomenon with a neurobiological foundation. The characteristics of experienced time are not correlated in a one-to-one manner with their neural underpinnings, but they are not magical or mysterious. They have a material basis in the asynchronous, disjunctive but patterned, conjoined interactions linking brain, body, and world. Narrative configurations and refigurations of time set in motion interactions between our lived experience of the world and our embodied brains that can in turn reshape the cognitive processes of integration that give rise to our sense of time passing. According to Ricoeur (1984a, 54), in narrative “a prefigured time . . . becomes a refigured time through the mediation of a configured time.” The work of narrative in invoking and repatterning the reader’s sense of time through the temporal structures of stories and their ways of telling is accomplished by its enacting and manipulating various embodied cognitive processes of synchronization and desynchronization that also characterize the neural workings of time.
The circuit between anticipation and retrospection is a pervasive feature of narrative and embodied cognition. Just as we live forward but understand backward in our everyday experience of the world, so we comprehend a story by projecting expectations that its twists and turns then modify and revise until we arrive at an ending that may overturn and completely reconfigure what we had previously assumed.5 This is why Heidegger says that understanding has a fore-structure (Vor-Struktur) that always already guides interpretation, an implicit configuration of expectations that our explications (Auslegung) belatedly catch up with and revise, refine, and correct (see 1962 [1927], 188–95). How cognition entails retrospective adjustment of what we have already anticipated is evident even at the neuronal level, in the ways in which our brains respond to signals before we are aware of them, a gap that we then subliminally correct and smooth over so that we typically do not notice it (otherwise we would have the weird sense that our present experience had already taken place in the past—which it indeed has, in a sense, inasmuch as consciousness lags behind detection by up to half a second).
Stories play with the anticipation-retrospection circuit at various cognitive levels—at the subliminal level, for example, as we immerse ourselves unreflectively in their revelations and reversals, and at the level of conscious awareness, as we reflect on the implications of the adjustments and revisions we must make as our expectations are surprised. At whatever level, the twists and turns in our experience of stories are only possible because the temporality of brain functioning is asynchronous, ever shifting between states of equilibrium and disequilibrium. If our brains were temporally unified and homogeneous, everything firing simultaneously and in lockstep, we could not tell each other stories because there would be no temporal gaps and no disjunctions between anticipation and retrospection for their discordant concordances to play with. The ways in which stories confirm or surprise our expectations can have deep effects on the neuronal assemblies through which we know the world because those assemblies are formed through to-and-fro recursive processes of synchronization and desynchronization.
The temporal asynchronies of brain processing can cause our conscious awareness of a signal to lag behind our reaction to it in ways that can seem paradoxical and strange. Noting that “our quick reactions to sensory signals appear to be performed without any initial awareness of the signal,” Benjamin Libet (2004, 93, 33) observes that “the brain needs a relatively long period of appropriate activations, up to about half a second, to elicit awareness” of an event that we may already have responded to. Libet conducted a series of well-known experiments that documented this lag by implanting probes in the brains of patients undergoing surgery to treat a neurological disorder (procedures during which they remained conscious), and these single-cell probes enabled him to measure with exceptional accuracy the difference between neuronal reactions to various signals and a subject’s reported awareness of them (see Libet et al. 1979). Libet’s experiments showed that “the subjective ‘present’ is actually of a sensory event in the past” because “our awareness of our sensory world is substantially delayed from its actual occurrence” (2004, 88, 70).
Although this may sound odd, it is quite an ordinary and in many respects a beneficial phenomenon. As Libet notes, “All quick behavioral, motor responses to a sensory signal are performed unconsciously, . . . within 100–200 msec after the signal, well before awareness of the signal could be expected,” as when a tennis player returns a serve, for example, or a baseball player hits a home run (109). Marc Jeannerod (2006, 9) similarly observes that “during the playing of a musical instrument . . . finger alternations can, in certain instances, attain the frequency of 16 strokes /s[econd], which exceeds the possibility of any sensory feedback influencing the command system.” Even in an ordinary activity like speaking, Libet points out, we typically talk before we are fully conscious of what we are saying: “If you try to be aware of each word before speaking it, the flow of your speech becomes slow and hesitant” (108). Similarly, if you are driving and see a child chasing a ball into the street in front of your car, Libet explains that “you are capable of slamming on the brake in about 150 msec or less,” even though you would not be consciously aware of the danger for another 350 milliseconds (91). As Jeannerod observes, “We respond first and become aware later. . . . [W]e consciously see the obstacle [only] after we have avoided it” (47; also see 41, 48–49, and 60–61 on experiments confirming Libet’s results).
What is perhaps surprising is that Libet’s experiments showed that drivers in this last example “subjectively antedate” the experience and “report seeing the [child] immediately,” without the lag that in fact occurred between response and awareness (91). Libet makes this inference based on experiments that compared a stimulus applied to the skin and another stimulus applied intracortically. Both stimuli would require 500 milliseconds to reach awareness, but the external stimulus was reported to have happened earlier, half a second before it was consciously perceived, whereas the intracortical stimulus was reported to have occurred only after the half-second lag: “In our conscious experience of a sensory event, the event seems to occur when it actually happened, instead of 0.5 sec later (when we, in fact, became aware of the event)” (81). A primary electrical potential measurable on the skin when the stimulus is applied apparently becomes the reference point that the brain uses to establish the time of the event, even though we become conscious of the stimulus only 500 milliseconds later.
What happens, according to Libet, is a subjective referral that “corrects” the “distortions . . . imposed by the way in which the cerebral neurons represent the event,” analogous to how (in a well-known experiment) subjects who wore “prism spectacles that turned the visual image upside down” began after a week or so “to be able to behave as if the image were normal” (81, 82).6 Because of this capacity for recursive cognitive adjustment, “subjective timing need not be identical to neuronal time” (72). Our sense of the present moment—of the simultaneity of an event and our perception of it—is thus what Libet calls an “emergent property” that is not attributable to any particular element of the system but that develops out of the interactions of its parts. It results from a to-and-fro process of subjective referral that antedates what we perceive and constructs temporal unity that smooths over the asynchronies underlying it.
Although subjective referral and temporal antedating were surprising and quite controversial when Libet first reported his findings (see Libet 1993, 2002, 2003; Gomes 1998), they are evident in phenomena widely attested and extensively studied in the neuroscientific literature on time. For example, in an experiment conducted by Patrick Haggard and his group (2002), a situation where an action was perceived to cause a tone was compared with a control case where the action and the tone were presented separately. Analyzing this experiment and its significance, Kielan Yarrow and Sukhvinder Obhi (2014, 462) note that “when an intended action caused the tone, the action was perceived to occur later, and the tone was perceived to occur earlier,” almost as if they were temporally attracted to each other. As Yarrow and Obhi observe, “It seems as though actions cause their delayed sensory consequences to appear earlier in time, while the perceived time of the action is also drawn toward the sensory event,” a time-shift Haggard calls “intentional binding” that may “help conscious inferences of causality” (462). In this way, according to Haggard (2002, 385), “the brain . . . binds intentional actions to their effects to construct a coherent conscious experience of our own agency.”
As Thomas Fraps (2014, 273) observes, this is only one of “several studies [that] have demonstrated that causes and effects mutually attract each other in subjective time.” According to Fraps, this phenomenon supports the well-known claim of eighteenth-century philosopher David Hume that “our perception of causality is a mental construct, one only inferred from the sensory experience of temporal contiguity” (265). No doubt for evolutionarily understandable reasons, our brains seem to have developed a predisposition to bind cause and effect by temporally closing the gap between them through reciprocal subjective referral. This subliminal mechanism also predisposes us to link the events emplotted in a story.
The brain’s proclivity to bind temporally distinct events similarly underlies the oft-discussed color phi and cutaneous rabbit phenomena. As Mölder (2014, 220–21) explains:
The color phi phenomenon is the illusion of movement that occurs when the subject is presented with flashes of two objects of different colors at different positions (e.g., a blue spot and a red spot). It seems that the spot is moving from one location to another and changes its color midway through its movement. What is puzzling . . . is that the color seems to already change in the location that precedes the actual location where the spot with a different color is presented. . . . [T]he red color seems to be present before it is actually presented.7
The cutaneous rabbit experiment similarly entails temporal antedating that constructs a coherent perceptual experience by reorganizing retrospectively the stimuli that give rise to it. If five taps are made at intervals of 40–60 milliseconds first at the wrist, then at the elbow, and then again at the wrist, what the subject reports perceiving is not three discrete sets of stimuli; instead, as Mölder explains (221), “the subject feels as if something moves regularly with smooth jumps up the arm, . . . not only in the locations of the stimulation”—“as if the later taps have a backward effect” on the earlier stimuli and give rise to a feeling like a bunny hopping up the subject’s arm (hence the experiment’s name). In both of these cases, the brain subliminally reorganizes the temporal relations between sensory stimuli to construct a coherent pattern that uses information about what came later to restructure what came earlier. This backward-looking readjustment of previous stimuli to rearrange our sense of their order and relationship may seem counterintuitive, but this recursivity is based on processes of temporal antedating, subjective referral, and intentional binding that are constantly at work in the relation between brain, body, and world, closing the gap between late-arriving perceptual awareness and the signals to which we have already responded.
These recursive processes are often misunderstood. For example, Brian Massumi and other so-called affect theorists have used (or, better, misused) Libet’s experiments to posit a bodily, autonomous realm of subpersonal affective processes prior to cognition. “Thought lags behind itself,” Massumi (2002, 195) claims; “it can never catch up with its own beginnings. The half-second of thought-forming is forever lost in darkness. All awareness emerges from a nonconscious thought-o-genic lapse indistinguishable from movements of matter.” This is a misinterpretation of the implications of Libet’s experiment. Rather than uncovering some “dark” realm of autonomous affect beneath awareness (what Massumi notoriously calls “the mystery of the missing half-second” [28]), Libet’s findings demonstrate how the temporal asynchronies of the embodied brain form configurations at different scales and interact recursively—the subliminal perception of a situation prompting a driver to slam on the breaks, for example, before conscious awareness arises of the child chasing the ball across the street, the two moments interacting to antedate the now to the moment when the intuition of danger first arose. This recursivity is not a “nonconscious thought-o-genic lapse” (whatever that odd coinage might mean) but rather as Shaun Gallagher (2005, 239) argues, a “looping” process through which immediate perception and conscious awareness interact and shape each other. In phenomenological terms, this interaction is a recursive relation between two modes of meaning making—what Merleau-Ponty (2012 [1945], 441–42) calls “nonthetic” and “thetic” intentionality to distinguish unreflective perception from deliberate, consciously active meaning-creation—not a juxtaposition between an autonomous realm of subpersonal intensities and conscious thought.
Massumi errs by transforming ordinary neurobiological and epistemological processes into obscure metaphysical mysteries.8 These processes have a material basis, but they are also instances of emergence, whereby a phenomenon (like consciousness or life) is produced by underlying activities (neurons firing, chemicals interacting) to which it is not reducible (see Deacon 2012). Contrary to what Massumi claims, emergent processes are not “indistinguishable” from the “movements of matter” out of which they arise, nor are they evidence that we are somehow controlled by subpersonal affective processes or that a “nonconscious lapse” undermines cognition. Emergence, the creation of wholes that are more than the sum of their parts, is an ordinary and ubiquitous consequence of recursive processes that are everywhere at work in our embodied brain’s interactions with its world (see Kelso 1995).
Libet’s mind-time experiments and other similar instances of subjective referral and intentional binding suggest that the ability of stories to emplot disparate events into coherent temporal patterns and thereby to create concord out of discord is correlated with various timing processes in the embodied brain. Across the different tempos of relationship that join our brains and the world, we are constantly configuring story-like patterns that establish concord and “correct” (so to speak) the discord between the actual occurrence of events and our awareness of them. If we live forward but understand backward, the ability to retrospectively reconfigure past and present begins at the neuronal level as we smooth out the asynchronies in the temporality of perception just as the wearers of prism spectacles adjust the inverted shapes on their retinas.
A similar process of to-and-fro reconfiguration is at work when, as Ricoeur (1984a, 67–68) observes, we find a surprising and unforeseeable conclusion to a story to be nevertheless “acceptable” because it is “congruent with the episodes brought together by the story.” At such moments, he notes, “it is as though recollection inverted the so-called ‘natural’ order of time. In reading the ending in the beginning and the beginning in the ending, we also learn to read time itself backwards.” The trick of “reading time backwards” is not such an unnatural act after all, however, because our ability to reconfigure the shape of time is also evident in the brain-based process of subjective referral that antedates what we become aware of to the moment when it occurred or in the intentional binding that draws cause and effect to each other. These recursive temporal processes are at work in the color phi and cutaneous bunny phenomena, and they are also in play when we retrospectively reconfigure narrative beginnings to align with our later sense of the ending.
Establishing congruence retrospectively between endings and the beginnings whose meaning may shift after we discover what they led to is a to-and-fro process of reciprocal configuration that is integral to our ability to follow stories. This is the reader’s response (mimesis3) to the narrative act of “ ‘grasping together’ the detailed actions” in the configurative construction of a plot (mimesis2) that “draws from the manifold of events the unity of one temporal whole” (Ricoeur 1984a, 66). We can readily align disparate beginnings and endings because our lived experience of time (mimesis1) entails a continuous series of subjective referrals that adjust our sense of the timing of present and past. We can learn to follow stories by reconfiguring relations between surprising endings and their beginnings in events we hadn’t fully understood because the configurative readjustment of anticipation and retrospection is something our basic neurobiological equipment predisposes us to do.
One of the reasons why Charles Dickens’s novel Great Expectations is a classic of narratological theory is that it illustrates these temporal paradoxes and complications in so many interlocking ways (starting with the pun in its title, which alludes both to the narrator’s many epistemological errors and to the fortune he hoped to inherit). The retrospective first-person narration of older, wiser Pip recounts the many erroneous assumptions and anticipations of his younger self, and the gap between young Pip and old Pip narrows as the time of the story converges on the time of the discourse. This temporal gap is the source of much of the irony and the comedy of the novel that survives on rereading, even when we are able to identify the various snares (or false clues) that Pip misreads about his benefactor—and one of the pleasures of rereading is spotting these mistakes (misinterpretations of which we too had perhaps been guilty the first time we read the novel). Just as a magician misdirects our attention by separating the cause of an action from its effect so that we don’t see their interaction, which makes it possible for us to marvel at the trick (see Fraps 2014), so Dickens constructs a masterful plot that disguises, disjoins, and displaces cause and effect, thereby masking connections that only later emerge but whose origins we hadn’t previously understood. This is a temporal art that takes advantage of the asynchronies in the way our brains process signals and construct patterns that we assume are simply “natural” and “there.” We are ordinarily unaware of the recursive, to-and-fro balancing acts that create stability out of the instabilities of perceptual experience, and the patterns of temporal experience may consequently seem natural and inevitable because the constructive processes underlying them are invisible. Hence our surprise when what we had expected (about Pip’s benefactor or about the coin the magician pulls out of our ear) turns out to have been an erroneous construction.
Masking the contingency of the temporal connections we routinely make is also part of the narrative game Dickens plays. As we read forward and understand backward, just as Pip lives forward only to narrate backward, the links between events in the story he had told himself about his life give way to a reconfiguration that seems to have been lurking in the background all along, waiting to spring—a gestalt shift Pip recounts in a memorable comparison of his seemingly fated retrospective enlightenment to a tale in which a “heavy slab that was to fall on the bed of state” and kill an unsuspecting usurper was slowly and elaborately prepared:
All being made ready with much labour, and the hour come, the sultan was aroused in the dead of the night, and the sharpened axe that was to sever the rope [holding the slab] from the great iron ring was put into his hand, and he struck with it, and the rope parted and rushed away, and the ceiling fell. So, in my case, all the work, near and afar, that tended to the end, had been accomplished; and in an instant the blow was struck, and the roof of my stronghold dropped upon me. (Dickens 2008 [1861], 285)9
Pip’s revelations seem fated and inescapable when they finally arrive, but this sense of inevitability disguises the various contingencies in his life (and in Dickens’s narrative)—beginning with the accident of the convict Magwitch encountering him in the graveyard—that seem like fate only when they are ordered retrospectively into a particular pattern. The paradox of a good plot is that the connections between events must seem necessary, but the twists, turns, and reversals of the narrative are only possible because these links are variable and contingent.
This paradox is evident in the oft-discussed curiosity that Dickens was able to write two opposing conclusions to the novel (one where Pip gets his girl, the other where he doesn’t). He could do this because beginnings and endings are contingent, reciprocally constituting, retrospectively variable constructs, and each ending refigures previous events in the novel into a different pattern—emphasizing either its comic, often sentimental, affirmations of Victorian values or its dark recognition of the shortcomings and hypocrisies that give the lie to those platitudes (you can see which ending I prefer!). In reading, storytelling, and life, patterns of temporal connection between events are not independent, necessary givens but are recursively established in a to-and-fro manner as anticipation and retrospection mutually constitute each other. We may feel that time simply happens naturally in our experience of the world or that the order of events in the stories we follow simply had to be, but the construction of time is an ongoing, ever-shifting process at all levels, from the recursive construction of neuronal assembles to the anticipatory and retrospective interaction of beginnings and endings. The fact that time is not a line, either in experience or on the neuronal level, but a to-and-fro product of pattern formation and dissolution, is what makes Great Expectations possible in all of its temporal complexity.
Questions of time, fate, and freedom often go hand in hand.10 One implication of Libet’s mind-time experiments that sparked particular controversy was that acts we may think result from the voluntary, conscious exercise of free will are in fact initiated by cerebral activity of which we are unaware: “That is, the brain starts the voluntary process unconsciously” (2004, 93). Measuring neuronal activity through probes implanted in the motor cortex of patients undergoing brain surgery, Libet found that “the brain exhibited an initiating process, beginning 550 msec before the freely voluntary act; but awareness of the conscious will to perform the act appeared only 150–200 msec before the act” (123–24). The conscious awareness of an intention to act arose, then, approximately 350 milliseconds after the brain had already initiated it.
Libet rejects, however, the inference others were quick to draw that freedom is an illusion. Although “conscious free will does not initiate our freely voluntary acts,” he argues, “it can control the outcome or actual performance of the act” through a kind of “veto” (139). This is different from instances of automatic or uncontrollable behavior, as in Tourette’s syndrome where subjects “spontaneously shout obscenities” (142). Libet found experimentally that such automatic responses entail “a quick reaction to an unwarned stimulus” for which there is no readiness potential (142). That is, there is no temporal lag in such cases between an unconscious initiatory activity in the brain (the readiness potential his experiments measured) and subsequent awareness (after the time lag necessary for consciousness), and without this disjunction there is no possibility of retrospective intervention (the “veto power” of freedom). Once again homogeneity is disabling. The gap between a readiness potential and our awareness of an intention is what allows retrospective readjustment. Rather than evidence of our lack of autonomy, as Massumi (2002) claims, this lag is one foundation of our ability to exercise free will.
A temporal gap similarly characterizes many common experiences of action in the world that are free and intentional not because they are fully conscious and transparent but because they are capable of being recursively reoriented. As James J. Gibson (1979, 225) observes, “Locomotion and manipulation are neither triggered nor commanded but controlled. They are constrained, guided, or steered, and only in this sense are they ruled or governed.” Motor control is a process of ongoing prospective and retrospective adjustment, and this steering or guiding is made possible by the subliminal temporal disjunctions that Libet has identified (see Gallagher 2012, 109).11 “Motor perception is predictive,” as Ivar Hagendoorn (2004, 83) explains, as it must be to compensate for these lags. Hagendoorn notes that it takes 50–100 milliseconds to process visual signals after they first reach the retina, followed by further delays in coordinating these signals with various embodied motor processes. We steer and guide our actions by anticipating how these visual and motor patterns will come together. In order to catch a ball, for example, we need to predict and anticipate its arrival subliminally—and if we think about what we are doing and try to intend the action deliberately, we are going to drop it—but we are of course free to act differently (jumping out of the way or batting the ball down). We can also improve our skills through learning and instruction (improving our predictions and our anticipatory coordination of perception and action) as we could not if these behaviors were outside the realm of freedom. Citing a study that “compared the performance of astronauts catching a ball on earth and under zero gravity,” Hagendoorn suggests that “the brain uses an internal model of gravity-induced acceleration when predicting the trajectory of a falling ball” (83). When gravity is missing, the predictive processes underlying the coordination of action become visible because they don’t work and need to be adjusted—which we can do because of the temporal disjunctions that allow the brain to construct various kinds of subjective referrals.
As Buzsáki (2006, vii) observes, “Brains are foretelling devices,” and their “predictive powers emerge from the various rhythms they perpetually generate.” These rhythms give rise, according to Marcus Raichle (2001, 8), to “changes in brain activity preceding the appearance of a to-be-attended sensory stimulus,” cortical variations that are the neural correlates of anticipation. The brain’s ongoing fluctuations allow interactions at the neuronal level to produce the signals that steer our embodied responses to the world, making subliminal predictions about how patterns will come together in perceptual experience and then retrospectively tuning these invisible, unnoticed guesses in a never-ending, to-and-fro, recursive process. Free will is not an illusion, then, but the story we tell ourselves of being freely choosing, fully conscious agents may be easier to understand than these back-and-forth processes of fluctuation and synchronization at various levels of awareness.
Reciprocal processes of pattern formation are fundamental to the brain’s operation as a to-and-fro ensemble of neuronal assemblies that are constantly coming and going, waxing and waning. The technical term for this back-and-forth reciprocal processing is “reentry,” a term coined by Edelman and Tononi (2000, 48, 49) to describe “the ongoing, recursive interchange of parallel signals between reciprocally connected areas of the brain, an interchange that continually coordinates the activities of these areas’ maps to each other in space and time.” As they explain, “re-entry allows for a unity of perception and behavior that would otherwise be impossible, given the absence in the brain of a unique, computer-like central processor.” Varela and colleagues (Varela et al. 2001) call this decentered, interactive structure the “brain web.” As Stanislas Dehaene (2014, 137, 156) observes, such a structure requires “long-distance communication and a massive exchange of reciprocal signals”—what he calls “recurrent processing” through “short local loops” within particular areas of neural anatomy (like the different specialized regions of the rear visual cortex dedicated to orientation, motion, color, etc.) and long-distance “global loops” for more complicated cognitive activities like reading or listening to stories that necessitate interactions between far-flung regions of the brain (including, for example, the visual, auditory, and motor cortices, the areas connected to emotions like the insula and the amygdala, the memory functions associated with the hippocampus, and interactions across the brain-body divide through the thalamus and the brain stem). “Consciousness lives in the loops,” Dehaene memorably declares; “reverberating neuronal activity, circulating in the web of our cortical connections, causes our conscious experiences” (156). The processes of configuration, refiguration, and transfiguration that constitute narrative also characterize the interactions of the brain web, and that is one reason why stories can have such deep and lasting effects on the workings of our brains.
Some of these effects may occur subliminally, priming us to expect certain configurations that conform to our habitually established patterns of response. Narrative understanding often occurs beneath and before awareness, especially when the repetition of well-known stories draws on and reinforces familiar, expected configurations. Richard Gerrig (2012, 50) has shown (perhaps not surprisingly) that “preferences with respect to potential outcomes . . . affect the ease with which readers accept outcomes when they ultimately arrive.” His laboratory demonstrated experimentally, for example, that readers took “longer to read sentences that were inconsistent with their preferences” and didn’t align with what they expected or desired. This finding is consistent with experimental evidence that “we always perceive a predictable stimulus sooner than an unpredictable one” (Dehaene 2014, 127). Processing speeds up when “anticipation compensates” for the time lag of consciousness and facilitates the subjective referral of temporal attribution (Dehaene 2014, 127). Whether with stories or in life, we perceive faster what our habitual experiences have accustomed us to anticipate. But structuring unpredictability into a narrative pattern of suspense can counteract this effect. Gerrig’s collaborator and former student David Rapp (2008) found that readers took measurably longer to read factually inaccurate statements (like “Shirley Temple starred in The Wizard of Oz”) than sentences they knew to be true, but he also found that offering readers counterfactual information in the form of a suspenseful story significantly reduced these reaction times. Configuring counterintuitive, anomalous statements into predictable narrative patterns (the buildup and resolution of suspense) made it easier to assimilate discordant information.
This may be why Aristotle famously prefers probable if impossible plots to ones that are possible but improbable. Configuration matters for comprehension, as when the reversals in a story seem “right” because, as Ricoeur (1984a, 43) puts it, “strokes of chance . . . seem to arrive by design.” Hence the surprising sense of inevitability that Pip reports when the revelation of the identity of his benefactor causes him to reconfigure the patterns through which he understood his life—refigurations that we readers duplicate on our parallel track of anticipatory and retrospective pattern formation as we form and reform our understanding of the story offered by the narrator’s discourse. All of the connections Dickens weaves together, some defying probability—how unlikely, after all, that Magwitch would turn out to be Estella’s father and that his daughter would happen to end up as a ward of a foster parent in Pip’s hometown!—are acceptable and may even seem aesthetically inevitable because of how they fit together (although I have skeptical students every semester who find these links preposterous, despite my trying to convince them that they demonstrate Dickens’s artistry and testify to his moral sense of the hidden interconnections binding the human community). Such experiences of discord that are resolved into new structures of concord can teach us to recognize and accept new configurative patterns, whether aesthetically or morally or epistemologically, introducing us to new ways of grasping together or following events. The plot of a bildungsroman like Great Expectations can offer an education not only to its protagonist but also to its readers through these kinds of reconfiguration.
Plots please and instruct through their manipulation of probability and improbability, familiarity and unfamiliarity, novelty and surprise. Reversals like these are not only entertaining but also potentially cognitively formative because the narrative work of configuring beginning, middle, and end manipulates temporal processes that govern the brain’s perpetual balancing act as it revises its habitual syntheses and adjusts to novel, unexpected situations. As Nancy Easterlin (2015, 614) observes, “If our predilection for the new and for the knowledge that it can provide is tempered by a contrary disposition to avoid the unfamiliar, reading a literary work offers the experience of encountering novelty without imminent threat.” By playing with the time lags built into cognition, narratives can reinforce established patterns through the pleasures of recognition, providing reinforcement for the structures that build coherence across our temporal experience, or they can disrupt the expectations through which we build consistency and offer new possibilities of recurrent processing, new patterns of reentry. Or, most typically, narratives can do both in varying degrees and in different ways that constitute the temporal art of emplotment.
If the conjunctions that smooth over temporal discordances can facilitate configurative activity, the disjunctions inherent in these time lags can also be productive. This is the case both with everyday cognition and with narrative. That is not only because these disjunctions may give rise to reconfigurations in cognitive patterns of the sort that the surprising twists and turns in a good plot enact. The temporal discrepancies that require subjective referrals to correct for the time lag of consciousness also allow consciousness to duplicate itself and thereby give rise to self-consciousness. What Dehaene (2014, 253) calls our “unique faculty for thinking nested thoughts” is made possible by the recursive operations of the brain web, the to-and-fro doubling of top-down and bottom-up assemblies that continually, reciprocally interact with one another.
Doubling of this kind is fundamental to consciousness and self-consciousness and also to narrative. The most basic narrative correlative of this capacity for nesting is, of course, the relation between discourse and story—the sometimes concordant, sometimes discordant interplay between the order of events in the telling and their sequence in the told. Narratives are essentially double structures because of narrative anachronies, “the various types of discordance between the two orderings” of story and narration that are fundamental to narrative (Genette 1980, 35–36). Even the classic formula “Once upon a time” has the basic temporal structure of doubling one time over against another, nesting the time of the story in the time of its telling. A doubling of temporalities is also fundamental to the recursive operation of reentry in the looping operations through which subjective referrals antedate and organize the timing of conscious experience. Cognitive processes that were temporally homogeneous and globally hypersynchronized could not give rise to the doublings that make possible nested thought or the interaction of discourse and story.
These doublings can take a variety of forms and have a range of different consequences in cognitive experience and in narrative. When discourse and story reinforce each other, the process of subjective referral is facilitated; when they veer off and diverge, the possibility emerges of interruptions and disjunctions through which the ordinarily invisible temporal operations of recursive pattern formation can come into view. The discourse-story opposition has within it the potential of either promoting the construction of narrative syntheses or supporting metafictional reflection because the temporal gaps that begin at the neuronal level can either be smoothed over through subjective referrals or give rise to the recursive nesting of cognitive operations that is responsible for self-consciousness, the doubling of living forward and understanding backward.
The temporal disjunctions typical of modern fiction expose this doubling in ways that may provoke reflection about the cognitive paradoxes of time. What Joseph Frank (1991 [1945, 1977]) famously calls the “spatial form” characteristic of much modern literature experiments with the temporality of narrative configuration in order to lay bare recursive, to-and-fro cognitive processes that ordinarily pass unnoticed. Frank recalls Lessing’s classic argument in Laocoön (1962 [1766], 78) that painting and sculpture are spatial arts that should depict “only a single moment of an action” by juxtaposing parts in a simultaneous whole (“signs existing in space can express only objects whose wholes or parts coexist,” Lessing argues), whereas the temporal art of narrative should depict the sequential unfolding of an action (“signs that follow one another can express only objects whose wholes or parts are consecutive”). According to Frank, the disjunctions of Ulysses and other modern novels that interrupt sequential, temporal coherence are “based on a space-logic” that turns Lessing upside down: “The meaning-relationship is completed only by the simultaneous perception in space of word-groups that have no comprehensible relation to each other when read consecutively in time” (15). Spatial juxtaposition replaces temporal sequence as the principle of narrative organization: “Joyce cannot be read—he can only be reread” because “a knowledge of the whole is essential to an understanding of any part,” and “such knowledge can be obtained only after the book has been read, when all the references are fitted into their proper places and grasped as a unity” (21).
Frank’s argument unleashed a fury of responses by critics and theorists who pointed out (among other things) that temporal processes of construal underlie our apprehension of the pictorial arts and that the retrospective construction of meaningful patterns prompted by modernist juxtaposition is itself a temporal activity.12 These objections are of course valid. As neuroscientist Arthur P. Shimamura observes, even seemingly spatial works of art like a painting or a sculpture are comprehended temporally: “It takes time to experience art, as we cannot appreciate an artwork fully within a single eye fixation. We therefore interpret an artwork by successive glances, and this storyboard of experiences takes some time, about three fixations per second” (2013a, 125; original emphasis). Our cognitive habits for processing temporal and spatial works—how we “read” literature and the visual arts—are consequently capable of interacting and affecting one another. For example, although ordinarily “we seem to scan paintings from left to right, perhaps with an upward glance, thus starting from the bottom left and moving to the top right,” Shimamura points out that “such biases appear to be guided by reading behavior,” inasmuch as “Israeli and Arabic readers adopt a right-left scanning bias during picture viewing and show the opposite aesthetic preferences than those who read from left to right” (91, 92).
What was often missed in the controversy, however, are Frank’s many acknowledgments that spatial form strains the reader’s cognitive capacities and works against our natural, habitual practices of sensemaking, and that this is precisely because of the fundamental temporality of those processes. The disorientation and bewilderment that modernist formal experiments may cause are evidence that sensemaking entails temporal configuration, and so the question is not whether these works ever achieve “spatiality” but what purposes might be served by the incoherences produced by such an impossible quest. At least one answer is that they call attention to the processes of patternmaking that are impeded, interrupted, and blocked, promoting reflection by thwarting immersion in a coherent, self-consistent world. Spatial form does not escape or transcend time, then, but makes its otherwise invisible workings visible for inspection and contemplation—and one might even observe that the insistence of Frank’s critics on the omnipresence of temporality is evidence that it has quite effectively done so.
This, then, is the cognitive takeaway from the spatial form debate: the back-and-forth construction of coherence through subjective referral and temporal antedating ordinarily goes without notice (as it does in conventionally realistic fiction), but the doublings, nestings, and shiftings of the brain’s balancing act become visible when its recursive processes are interfered with (as in the experiments of modernism), and then the tension between stability and instability may come into view. In Faulkner’s Sound and the Fury, for example, the disjunctions between the four nonconsecutive sections of the novel, compounded by the temporal leaps that Benjy’s and Quentin’s minds repeatedly make, unsettle the reader’s sense of temporal coherence and expose the contingency and instability of our interpretive constructions. This unease may surprisingly if understandably make the reader feel relieved when the certainty of the bigoted, misogynist Jason announces itself: “Once a bitch always a bitch, what I say” (1994 [1929], 113). We finally know where we are, but the fixity of this self-certainty is unsettling not only because it is morally repulsive but also because it belies the contingency of pattern formation that the temporal disjunctions of the narrative expose. The sometimes incoherent fluidity of Benjy’s and Quentin’s narratives takes to disabling extremes the instabilities of the brain’s balancing act between pattern and openness, but the rigid prejudice and self-defeating, paralyzing resentment of Jason’s viewpoint give readers a visceral experience of the straightjacket of entrenched, immovable habit. The temporal disjunctions of Faulkner’s narrative may make the reader long for coherence and dramatize the cognitive need for pattern, but once we have it, we don’t want it, because it is presented in so debilitating a manner. The experiments of other great modernists may foreground different aspects of temporality—the malleability of subjective time as the years pass in Thomas Mann’s epic The Magic Mountain, for example, or the horizonality of past and present in Virginia Woolf’s elegaic To the Lighthouse—but a defining characteristic of spatial form is that it blocks temporal processes of cognitive pattern formation, not to overcome them, but to play with them and expose them to view.
One of the paradoxes of time is that it combines both segmentation and integration. Time passes as a series of identifiable perceptual windows that flow into each other and are seamlessly joined together. The now is a perceptual gestalt, different from the past and the future, even as it is inextricably linked to what was and what will be across its retentional and protentional horizons. Narratives are similarly a paradoxical combination of separate segments that are integrated into an overall pattern. What Ricoeur (1980b, 169, 171) calls “the illusion of sequence” requires the construction of an “intelligible whole that governs a succession of events in any story,” but this whole only does its configurative work if the elements it binds retain their distinctiveness. As Ricoeur (1980b, 178) notes, there is a fundamental, necessary tension between “the episodic dimension” of a narrative (the events of a story) and “the configurational dimension”: “the plot construes significant wholes out of scattered events,” but “the configurational dimension cannot overcome the episodic dimension without suppressing the narrative structure itself.” In a circular manner (the narrative equivalent of the hermeneutic circle), the parts of a narrative constitute a whole even as the whole configures the parts. The tension that holds together these two dimensions (the episodic and the configurational) without collapsing them into each other is one aspect of the concordant discordance that characterizes narrative.
Basic tensions between the elements of any narrative and their organization occur across different temporalities of integration and segmentation, different scales at which binding and framing happen, from the sentence level up through the construction of the work as a whole. These interactions take advantage of what Buonomano calls “the brain’s exquisite ability to process complex temporal patterns” (2014, 330). As Sylvie Droit-Volet (2014, 482) points out, “The mechanisms for processing short and long durations are different.” These mechanisms interact in complex, reciprocal ways in ordinary cognitive activity and in our processing of narratives—a pattern of “upward and downward causation” on “multiple spatial and temporal scales where on one hand, activity on a lower scale gives rise to an emergent phenomenon and on the other hand, the large-scale patterns have the potential to re-influence the small-scale interactions that generated them” (Bagdasaryian and Le Van Quyen 2013, 4). Merlin Donald (2012, 40–41) distinguishes between “three levels of temporal integration”: “local binding” that occurs “from fractions of a second to a few seconds”; the short-term “working memory range” that lasts “a few seconds or tens of seconds,” and “longer lasting neural activity” like “conversations that last for many hours and organized games of various kinds”—or, of course, telling and following stories. Each of these levels of cognitive processing combines segmentation and integration at a different temporal scale, and what happens at each level and scale interacts upward and downward with the others. This is similar to (and the condition of possibility for) the way different timescales play off of each other in narrative.
At the most basic level of cognitive processing, the tension between segmentation and integration is evident first of all in the temporal binding that occurs at the microsecond scale, beneath conscious awareness. As Niko Busch and Rufin VanRullen (2014, 163, 166, 161) observe, our sense of time as “a pseudo-continuous stream . . . is actually made up from a series of discrete snapshots”—“a continuously moving temporal integration window” that, like a camera, “integrates and fuses information for as long as the shutter is open.” A panel in a graphic novel is a similar “snapshot” of time passing. As Scott McCloud points out (1993, 95), “Each panel of a comic [that] shows a single moment in time” actually includes various not necessarily simultaneous events within a window of duration, a window that can be broadened or narrowed to varying temporal widths just as a comic panel can be elongated or compressed.
This framing begins in units of integration much smaller than a comic panel can register. What Ernst Pöppel (Pöppel and Bao 2014, 247) calls the “elementary integration unit” that is the smallest temporal window where “the before-and-after relationship of stimuli . . . is not defined or definable” has been experimentally “observed in the time domain of 30 to 60 ms [milliseconds].” For example, he reports, when “experimental subjects are asked to indicate in which temporal order stimuli have been presented, such as which ear was stimulated first when both ears are stimulated by acoustic signals with a short temporal delay,” the differences between these stimuli are invisible within a “temporal order threshold . . . of some tens of milliseconds,” and this holds equally “for the visual, auditory, and tactile modalities” (248). As Marc Wittman (2104, 512–13) explains, “These thresholds are indicative of elementary temporal building blocks of perception, because below such a temporal threshold a succession of events, their temporal order, is not perceived.” Varela (1991, 73, 75) similarly describes effects of perceptual framing that produce “a natural parsing” of visual stimuli, whereby “everything that falls within a frame will be treated by the subject as if it were within one time span.” For example, he points out, two lights flashing within a span of less than 50 milliseconds appear simultaneous but then seem sequential beyond a 100-millisecond threshold, and between 50–100 milliseconds they “appear to move” from one spot to another.13
The paradox however, as Wittman observes, is that “our experience is not temporally punctual, a static snapshot,” and “we do not perceive the world as a sequence of individual events, but as a temporally integrated whole” (513). Microsecond units of temporal framing are integrated into consciousness first at the level of half a second, the window of integration in which a sense of the subjective present begins to be available to awareness. Another level of what Wittman calls “temporal grouping” seems to occur at 2–3 seconds. As he explains, “effortless and automatic synchronization” of the repeated beats of a metronome is possible from “a lower limit of around 250 ms” to “an upper limit of approximately two seconds,” but “beat perception and subjective rhythmization” break down at longer intervals (514). According to Pöppel, whose experimental research on these questions is foundational, “It is easy to impose a subjective structure” on “click sounds [that] follow each other with[in] an interstimulus interval” of two to three seconds, but this process of automatic figuration breaks down if “the temporal interval between the stimuli becomes too long (for instance, five seconds)” and “they fall into successive integration windows” (251). The familiar temporal gestalts of Morse code, for example, exploit the mechanisms of integration at this timescale, but there also are upper and lower limits to synchronization here owing to how the dots and dashes must be grouped for them to be intelligible (an operator clicking too fast or too slow would be incomprehensible).
Ambiguous figures like the famous rabbit-duck gestalt provide further experimental evidence of automatic grouping at the 2–3 second level. For example, Pöppel observes that “if stimuli can be perceived with two perspectives (like the Necker cube, or a vase that can also be seen as two faces looking at each other) there is an automatic shift of perceptual content after an average of approximately 3 seconds” (251). This “perceptual shift also occurs with ambiguous auditory material, such as the phoneme sequence KU-BA-KU, where one hears either KUBA or BAKU; one can subjectively not avoid that after approximately 3 seconds the alternative percept takes possession of conscious content” (251). As he concludes, “Metaphorically speaking, every 2 to 3 seconds, an endogenously generated question arises regarding ‘what is new’ ” (251).
The neurobiological basis of this formation, dissolution, and re-formation of windows of perception at different temporal levels is what Varela and colleagues (2001, 235, 229) describe as an “alternation or balance between phase synchronization and phase scattering” of “neural assemblies” that are “transiently linked by dynamic connections.” As Evan Thompson (2007, 335) explains, there is “suggestive evidence that long-distance phase synchronization and desynchronization may subserve the temporal parsing of cognition into coherent and momentary acts.” This synchronization and scattering, fundamental to the brain’s balancing act between stability and instability, happens in windows of temporal grouping at different timescales—at the 30–60 millisecond level, at the 500 millisecond interval for consciousness, and in larger gestalts at the 2–3 second range—as well as in a to-and-fro interaction between these different temporal integration units. These fluctuations between different timescales of segmentation and integration make possible what Marcus Raichle (2011, 9) calls “the dynamic interplay between the brain’s ongoing rhythms and its ever-changing environment.”
Windows of perception, attention, and awareness operate at different timescales, and these windows interact, forming and dissolving, in ongoing, ever-shifting patterns of synchronization and scattering. These interactions occur across the entire range of the brain’s fluctuations, beginning with micro-scale neuronal spiking and the preconscious perceptual binding of stimuli and continuing up to the synchronization and dissolution of long-range assemblies across various sensory regions and between the brain and the body—all of the different interlocking, reciprocally formative, to-and-fro movements of what György Buzsáki (2006) memorably calls the “rhythms of the brain.” That neuroscientists would resort to a musical analogy to describe these interactions is not accidental. The temporalities of segmentation and integration in the brain’s fluctuations are less like the linear steps of a computer algorithm than the shifting, reciprocally interacting relations between musicians in an orchestra who all play different parts that resonate together in patterns that form, dissolve, and re-form over the course of a performance.
The patterns of segmentation and integration that characterize these rhythms are the neural basis of the episodic and configurational dimensions of narrative, which similarly entail reciprocally formative windows that group events at different interlocking, interacting timescales. Narrative, like other temporal arts, is what Tim van Gelder (1999, 252) calls “a dynamical system,” which, unlike a linear computational model, “is a set of quantities evolving interdependently over time,” bounded segments interacting as they become integrated into larger patterns of figuration at different temporal levels. As van Gelder observes: “Here’s the puzzle: if at any given time I am just aware of the part of the song playing now, how can I ever be aware of the song as an integrated piece that takes place over time?” (245). We temporally bind the unit of the melody that we hear at the moment (on the 2–3 second scale) because the sounds form a temporal gestalt that we recognize (think, for example, of the memorable opening four notes of Beethoven’s Fifth Symphony), but in and through these distinctive units we also perceive the melody as a whole: “When you listen to a melody, you do not have to wait around until the melody is finished in order to hear it as a melody” (247)—or to recognize the tune and change the channel on the radio or skip ahead on your playlist because it’s not what you’re in the mood to listen to.
Plots are similarly composed of subunits (events, episodes, conflicts) that horizonally project the larger configurations they suggest (a picaresque journey, a novel of education, a romance, a detective story, etc.), and these larger patterns, as they take shape, can in turn refigure our understanding of their components. This interaction of configurative patterns at different timescales enacts in narrative form the basic paradox of neurally based temporality—the inherent duality of segmentation and integration in the brain’s rhythms, discontinuous “snapshots” that give rise to a continuous stream of experience. What Ricoeur calls the basic tension in narrative between its episodic and configurational dimensions is possible because the brain processes time in a contradictory manner both as distinct windows and as an integrated, rhythmic flow.
The duality of segmentation and integration in the temporality of cognitive processing undergirds the brain’s perpetual balancing act between pattern and change—what Dehaene (2014, 189) calls its state of perpetually “unstable equilibrium.” As he explains, this neurobiological balancing act is what makes possible “exploratory behavior” as spontaneously generated “fluctuating patterns of activity” synchronize, scatter, and re-form (189). At a variety of temporal levels, windows of structuration impose gestalts on the flux of stimuli only to give way to other syntheses that in turn are just temporary equilibria in an ongoing alternation between stability and instability, ever-changing phases of synchronization and scattering and resynchronization. If brain time were simply a continuous flow, we would not alternate between patterns of assembly and disassembly, but if temporal gestalts did not regularly, automatically fade and re-form, we would be stuck forever in one shape and be unable to explore, adapt, and change.
Narrative can be an effective instrument for organizing experience and exploring the world because it too plays stability and instability off each other at different levels of temporal organization. Temporal patterns of segmentation and integration characterize both the telling and the told. As James Phelan (2002, 211) points out, there are “two main kinds of instabilities” in narrative: “Those occurring within the story, instabilities between characters, created by situations, and resolved through actions,” and “those created by the discourse, instabilities . . . of value, belief, opinion, knowledge, expectation” informing the pattern of narration and structuring the relation between the narrator and the recipient. In a plot, for example, a pattern is established only to be upset, overturned, and replaced by a new configuration in a temporary equilibrium that is open to subsequent dislodging. Todorov (1969, 75) has something like this in mind when he famously argues that “the minimal complete plot can be seen as the shift from one equilibrium to another, . . . two moments of equilibrium, similar and different, [that] are separated by a period of imbalance, which is composed of a process of degeneration and a process of improvement.” This alternation between equilibrium and disequilibrium is not just a once-and-done effect of the overall course of the story but may happen over and over again as the plot sets up conflicts, resolves them, only to introduce new complications. A plot is constructed by the repeated creation, dissolution, and re-formation of temporal windows of integration.
Shifting patterns of equilibrium and disequilibrium characterize not only the story but also the discourse. Here too our cognitive ability to process alternating windows of integration may be put into play. An extreme case is unreliable narration, where readers may come to wonder whether the narrator is misleading them, either deliberately or unknowingly. Unreliable narration is only an especially visible instance of the different degrees of balance and imbalance that are possible in the interaction between the telling and the following of a story over the time course of the narrative. At the level of discourse, in the way a story is presented over time, a manner of telling proposes a pattern of organization or a mode of seeing that may then in turn be confirmed or challenged and questioned by what is told, as when we begin to doubt a questionably reliable narrator’s version of events: is the supposition that ghosts are threatening to corrupt the children in The Turn of the Screw a more plausible configuration of the evidence, for example, than the suspicion that the governess who is telling us the story is beset by anxieties and unconscious desires that generate sexually charged fantasies? A shift in the reader’s attitude toward the discourse (believing or doubting the narrator) may then refigure the pattern he or she sees in the events of the story.
The equilibrium between the story and the discourse is potentially unstable and open to alteration and re-formation as one window of the telling gives way to another and these perspectives blend (or do not). At what point do we shift from one gestalt (ghost story) to another (hysterical governess), or do we go back and forth between them? In Lord Jim, similarly, each time Marlow gets a different view of Jim from another of the many informants he consults (for example, the French lieutenant’s take on Jim as a scoundrel who deserves condemnation for his breach of honor, or Stein’s perspective on the young man as a romantic who merits sympathy), the narration offers a temporary window of integration that the recipient may accept or question, only to have it dislodged by the next window the narrator offers. Is Jim simply guilty of criminal conduct, or is he an example of the nobility and fragility of the capacity for dreaming and imagining that all of us share? Marlow’s struggles to reconcile opposing interpretations of Jim in his narration are consequently as intriguing or disturbing (or both) as the story he is trying to tell. The shifting perspectives of the discourse challenge the reader’s attempts to integrate them, and these disjunctions make Conrad’s novel especially interesting for narrative theory by foregrounding what are usually invisible processes. Synthesis and scattering, segmentation and integration, units of interpretation and their interaction across the temporal unfolding of the narrative—these fundamental dualities of the temporality of cognitive processing characterize both the story and the discourse.
Just as the brain is constantly balancing the competing claims of stability and instability, pattern and change, so the shifting windows of integration in the tale and its telling entail an ongoing alternation between the imposition of patterns and their dissolution and reformation. According to Ellen Spolsky (2015, xxiv), stories and other “works of the imagination . . . help us keep our balance in a changing and often threatening world” and “contribute to the biological work of homeostasis. We are not stable in the way stones are stable, but we are evolved to maintain sufficient stability by sensing and repairing instabilities.” This is, however, only half of the story of narratives and of cognitive life. Instabilities and imbalances that keep us open to changes in the world’s patterns are also necessary as we adjust, explore, and create in an ever-varying environment. The temporality of cognition is characterized by ongoing rhythms of pattern formation, dissolution, and reformation, and these processes of stabilization and destabilization, synthesis and disruption, synchronization and scattering, are also integral to the construction of stories and their narration.
This necessary tension between equilibrium and disequilibrium is one reason why the endings of stories have received so much attention from narrative theory (see Kermode 1967 and Torgovnick 1981). As narratologists have often noted, closure in fiction is not a permanent imposition of stability because any ending is susceptible to reopening and may give rise to new beginnings. Closure may not necessarily be closed. Conversely, however, as Peter Rabinowitz (2002, 308) notes, “lack of closure does not mean lack of conclusion.” An ending may establish a state of equilibrium at the level of the discourse (“That’s all, folks”) while leaving much unsettled in the story (“Did they live happily ever after or not?”). A classic case is the ending of The Portrait of a Lady where, as James himself acknowledged, the uncertain consequences of Isabel’s decision to return to Rome leave him open to the “obvious criticism” that the novel “is not finished—that I have not seen the heroine to the end of her situation—that I have left her en l’air” (1987, 15). But as James memorably explained, “This is both true and false. The whole of anything is never told; you can only take what groups together” (15).
Ending is a temporal moment of integration that “groups together” elements in an inherently contingent, only seemingly stable equilibrium that is open to reorganization. This is why Phelan (2002, 214) distinguishes between “closure” and “completeness”: “Closure . . . refers to the way in which a narrative signals its end, whereas completeness refers to the degree of resolution accompanying the closure.” A narrative conclusion may be characterized by varying degrees of completeness, resolution, and openness, just as any temporal binding of a gestalt at whatever level of temporal organization is a more or less stable equilibrium in the ongoing alternation between segmentation and integration, more or less firmly fixed by previous cognitive activity and consequently more or less likely to give way to a different configuration. Similar tensions between equilibrium and disequilibrium characterize the different temporal structures of emplotment, narration, and ending because these are all ways in which narrative plays with the brain’s perpetual balancing act between stability and instability, its fluctuations between synchronization and phase scattering at local and global levels of binding and everywhere in between.
The temporality of narrative piggybacks on processes of segmentation and integration that characterize cognition in general and the reading process in particular. As Genette (1980, 34) observes, “The narrative text, like every other text, has no other temporality than what it borrows, metonymically, from its own reading.” The smallest units of temporal synthesis, from the microsecond level up to the 2–3 second integration window, correspond to the processing of letters and phonemes into words and words into sentences. Even at these levels of text processing, comprehension is configurative—a reciprocally formative grouping together of features into a gestalt. As Dehaene (2009, 49) explains, “Word decoding does not proceed in a strictly sequential manner, and the time needed to read a word is not related to the number of letters that it contains,” as it would be if reading were a linear, additive process. According to linguist I. M. Schlesinger (1968, 42), “Decoding proceeds in ‘chunks’ rather than in units of single words, and . . . these ‘chunks’ correspond to the syntactic units of the sentence.” Wolfgang Iser (1978, 110) points out, however, that these units in turn are not self-contained because “each sentence can achieve its end only by aiming at something beyond itself.” Reading a narrative (or any text, for that matter) entails an ongoing alternation between what Iser calls “theme” and “horizon” that “actively involves the reader in the process of synthesizing an assembly of constantly shifting viewpoints,” ever-changing perspectives that “not only modify one another but also influence past and future syntheses” (97).
This is not unlike ordinary, everyday perception where, as Gibson (1979, 66) notes, “a point of observation is never stationary, . . . and observation is typically made from a moving position.” What Iser calls “the continual switching of perspectives during the time-flow” of reading (97) is a special case of the general rule that “observation implies movement” (Gibson 1979, 72), but this switching can also create unique effects characteristic of narrative. As Iser points out, “the wandering viewpoint” of the reader is “situated in a particular perspective during every moment of reading, but . . . it is not confined to that perspective” and “constantly shifts between” angles of observation in a “process of reciprocal spotlighting” that “offsets and relates” the shifting viewpoints that the reader occupies (114). In a mimetic fiction, for example, “the accumulation of views and combinations gives us the illusion of depth and breadth, so that we have the impression that we are actually present in a real world” characterized by an event-like happening as perspectives change and interact (97). Sometimes, as in a modernist text like Ulysses or a postmodern work like Infinite Jest, the “kaleidoscope of perspectives” (113) may multiply so variously and resist integration so playfully (or stubbornly and annoyingly, depending on your sensibility) that they become foregrounded in their own right, not merely serving as means to representational ends, and the temporality of integration may then emerge from invisibility, transformed from a passive process into an active topic for contemplation because its smooth flow has been interrupted. Or the temporal interplay of theme and horizon may give rise to various effects of ironic doubling as “two perspectives throw each other into distinct relief” (117) and disclose one another’s defining limitations, whether for tragic or comic effect. Narrative irony is a temporal form that depends on the interaction between segmentation and integration as it sets alternative windows against each other in the time of reading.
These oft-discussed characteristics of narrative exploit the ability of the brain to form moments of integration of varying width and to relate them to one another in alternating patterns of synchronization and dispersal. A tension between segmentation and integration is pervasive in narrative, as it is fundamental to the temporality of cognitive life. The theme in focus at any given moment is a temporal gestalt (a window of integration), but it is then configuratively joined across its horizons to receding viewpoints immediately past and to perspectives anticipated ahead. What it would be like if these windows did not interact but instead existed in isolation is suggested by cases of motion agnosia, where the afflicted person “experiences the world as seemingly without motion, frozen in place, for several seconds. Things in the world may then suddenly seem to rearrange themselves in new positions” (Gallagher and Zahavi 2012, 78), making it difficult if not impossible to engage in routine acts like pouring a cup of coffee or crossing a street. Motion agnosia casts the segmentation of temporal windows into distressingly bold relief by disrupting their prereflective, ordinarily invisible integration.
The narrative equivalent of motion agnosia would produce disjointed episodes not connected by configurative emplotment or would offer up juxtaposed perspectives not organized by the discourse into coherent patterns. This kind of discordance without concordance is, of course, precisely the effect aimed for by the disjointed textual structures of many so-called unnatural narratives, disjunctions that subvert the reader’s expectations of mimetic consistency (the surreal disconnections of Djuna Barnes’ Nightwood, for example, or the shifting dreamscape of Calvino’s If on a Winter’s Night a Traveler). Just as motion agnosia is only understandable as a disturbance of ordinary processes of temporal integration, however, so the disruptions of unnatural narrative once again depend for their effect on the “natural” functions of cognitive patternmaking that they suspend, discordantly foregrounding the segmentation of narrative that is usually invisible because synthesized concordantly.
As narrative theorists have long recognized, the ability to follow a story depends on the connective glue of memory. Hence Aristotle’s (1990 [355 BCE], 9) understandable if perhaps overly rigid assertion that “the length of a plot should be such as can be easily retained in memory.” The cognitive intuition behind Aristotle’s claim is that narrative integration requires the capacity to make and remember connections between temporally distributed segments. Although much narrative comprehension is subliminal and occurs beneath awareness, forging these connections, especially over longer timescales, would be impossible without explicit, conscious memories. As Dehaene (2014, 103) points out, “Subliminal thoughts last only for an instant,” whereas “a temporally extended working memory requires consciousness.” According to Dehaene, consequently, “we need to be conscious in order to rationally think through a problem” with multiple steps that have to be held in memory (108).
For example, solving a complicated problem in arithmetic (adding or subtracting a sequence of numbers) is something we can only do because we are conscious and can therefore remember its various steps. But this is also true of a story that is composed of different episodes and that, as it is told, passes through different phases of equilibrium and disequilibrium. Consciousness is consequently not an epiphenomenon irrelevant to the deep structures of narrative as some versions of structural narratology suggest (recall Ann Banfield’s claim, for example, that we can understand stories only because of homologies at the level of “universal grammar” between narratives and our preconscious minds). Although many of the effects of narrative occur subliminally, beneath awareness, it is also true that only a being endowed with consciousness and the capacity it supports to hold together multiple elements in memory would have the ability to tell and follow a story. Without consciousness we would not have narratives at all. One answer to the oft-asked question, “Why have we evolved to have consciousness?” is that consciousness allows us to tell each other stories, and that is something that gives our species all sorts of competitive advantages (see Humphrey 2006 and Boyd 2009).
Perhaps paradoxically, the ability to forget is also essential to forging temporal syntheses in everyday cognition and in narrative. As Yadin Dudai (2011, 36, 37) observes, “Memories too robust are a potential disadvantage, as they may not fit anymore to guide the proper action and reaction in a changing environment”; “too rigid a memory may lead to poor imagination, one that plays scenarios of the future that are only similar to the past.” What Daniel Schacter (2002) famously calls “the seven sins of memory”—“transience, absent-mindedness, blocking, misattribution, suggestibility, bias, and persistence”—are not necessarily flaws but may be cognitively beneficial. His list of “sins” reflects the brain’s never-ending balancing act between stability and instability—the recurrence of patterns from the past (“persistence,” “bias”) versus their contingency and susceptibility to variation (“transience,” “absent-mindedness”). As Alcino J. Silva (2011, 49) argues, “Memory mechanisms are designed, not for accuracy and permanency, but instead, for constant editing and fine-tuning of information with experience.” The defects of memory may be functional and not maladaptive because they support this “editing and fine-tuning.” The temporality of the brain once again reflects a double imperative. If memories did not preserve cognitive configurations, we would be lost in the flux, but a perfect memory would lock us into past patterns and make us resistant to change, whereas our fallible, unstable recollections support cognitive flexibility.
The shifting patterns of equilibrium and disequilibrium characteristic of narrative comprehension similarly require an ability not only to remember but also to cast aside and revise previous configurations. Following a story is a balancing act between memory and forgetting as patterns form and re-form in shifting relations of stability and instability. In Great Expectations, for example, Dickens relies on our forgetting Pip’s youthful act of kindness toward an escaped convict on the marshes in order to lull us into sharing his hopes about his presumed benefactor Miss Havisham and her ward Estella—but then our surprise when Magwitch announces himself as the true source of Pip’s fortune reminds us of what we may have forgotten. The snares we fell for took advantage of our forgetfulness—otherwise we would not be surprised by Magwitch’s revelations—but we can then edit and revise our understanding by recalling and readjusting our memories of Pip’s past (just as he has had to do) and see how he (and we) had misled ourselves by unjustified expectations. The “sins” of memory make us vulnerable to the snares Dickens uses to misdirect our attention, but they also give us the flexibility to balance and rebalance shifting patterns of integration in response to changes in life and in stories, as we do when we revise and reedit our sense of the connections between events.
The usefulness of a fallible memory has led a growing number of cognitive scientists to recognize that our ability to remember the past is closely linked to our capacity to imagine the future. As Terence Cave (2016, 74) notes, experimental evidence convincingly shows that “memory and imagination are not, as earlier faculty psychologists believed, separate domains, but different aspects of the same cognitive processes.” Schacter and Addis (2007, 773) observe, for example, that the liability of memory to “various kinds of errors and illusions” also allows it to support the imaginative simulation of future events, which “requires a system that can draw on the past in a manner that flexibly extracts and recombines elements of previous experiences.” Citing lesion studies that show “an overlap between deficits of remembering and envisioning the future,” Randy L. Buckner and David C. Carroll (2007, 51, 49) posit “a core brain network” or “common set of processes by which past experiences are used adaptively to imagine perspectives and events beyond those that emerge from the immediate environment.” This has come to be called the “default mode network” (DMN) because it was first discovered by analyses of fMRI scans of the brain at rest, when it was not engaged in any specific attentive activity (its “default” condition). As Robert Stickgold (2011, 90) explains, the “resting or ‘default’ brain state” entails “envisioning the future, . . . remembering the past, . . . conceiving the viewpoint of others, . . . and spatial navigation.” These interactions between remembering and imagining suggest, he argues, that “narrative construction is the default mode of the brain” (90).
The neuroscientists may have found these links surprising, but cognitive literary critics were quick to point out that they are not news to anyone familiar with the history of aesthetics. As cognitive theorist and historian of romanticism Alan Richardson (2011, 670, 665) was among the first to observe, the “Janus hypothesis” that sees “retrospection into the past and prospection into the future” as two sides of the same coin is widely attested by “centuries of literary scholarship on imagination.” Ignoring this history, he argues, “has made for an impoverished scientific agenda” (665). Our ability to engage in what neuroscientists call “mental time travel” (Suddendorf and Corballis 2007)—our dual capacity to recall the past and imagine possible futures—will be familiar to anyone who has read “Tintern Abbey” (1798) or any of Wordsworth’s many other poems about memory and imagination.
This coupling of retrospection and prospection extends across our cognitive life, from everyday daydreaming to aesthetic experiences of various kinds, and these connections suggest why it is a commonplace of discussions of literary imagination to relate daydreaming to creativity as Freud (1958 [1908]) does, for example, in his classic essay on the psychological sources of poetry. Indeed, neurobiologically trained literary theorist Gabrielle Starr (2013, 23), who has had an active collaboration with brain scientists at New York University, conducted experiments that showed measurable increases in the activity of the default mode network not only in internally-oriented acts of daydreaming but also in response to “intensely powerful aesthetic experiences” provoked by external stimuli (such as seeing a painting, hearing a musical composition, or reading a story). This stands in marked contrast to the DMN’s response to ordinary tasks requiring attention to external stimuli, during which its “activity generally decreases,” and this evidence “suggests that powerful aesthetic experience calls on the brain to integrate external perceptions [our response to the artwork] with the inner senses” by drawing on the ordinarily inner-directed capacities of the DMN (Starr 2013, 23; for the original experiment, see Vessel, Starr, and Rubin 2012). Such integration requires to-and-fro interactions between cortical areas devoted to memory and imagination and those involved in attention and perception.
Just as narrative comprehension occurs on various timescales, so memory itself is not uniform, either neurobiologically or experientially. As Shaun Gallagher and Dan Zahavi (2008, 70–71) point out, “Memory is not a single faculty of mind. Rather, it is composed of a variety of distinct and dissociable properties,” and fMRI scans show that “different parts of [the] brain appear to be particularly active depending on the type of memory task [we] are engaged in.” Neuroscientists distinguish, for example, between short-term memories that may last from seconds to minutes to hours and long-term memories that may extend over days and weeks and may then become consolidated over a lifetime (see Baars and Gage 2010, 324–25). The differences between short- and long-term memories have to do with how engrams, the anatomical basis of memories, are formed and reshaped and reinforced. Engrams are patterns of neuronal assembly that become established through repeated Hebbian firing and wiring—how neurons that fire together, wire together—in short-term processes of varying duration. These patterns may then become more enduring as they are reinforced through repetition and consolidated by various metabolic processes (during sleep, for example) and by specialized parts of the cortical anatomy like the hippocampus (see Bear, Connors, and Paradiso 2007, 725–59).
Cognitive scientists further distinguish between different kinds of memory based on the abilities they support—for example, between declarative memory, on the one hand, which can be semantic (that is, memory of facts) or episodic (memory of events), and nondeclarative memory, on the other, which is typically not fully available to consciousness, and that includes subliminal priming effects, emotional dispositions, habits, and procedural memories (skills like swimming, typing, or riding a bicycle) (see Baars and Gage 2010, 326). As well-known cases in the neuroscience literature attest, it is possible to lose one form of memory but retain others. A patient who suffers a stroke that blocks the retrieval of semantic or episodic memories, for example, may still be able to perform various skilled tasks that draw on procedural memories.14
These categories are somewhat fuzzy and the boundaries between them are fluid, and necessarily so. Memories pass from one form into another as different processes of integration and consolidation interact. For example, working memory is “a temporary form of information storage that is limited in capacity,” usually encompassing five to seven items, but it can change from short- to long-term memory through repetition and rehearsal (Bear, Connor, and Paradiso 2007, 729). The distinction between episodic and semantic memories is similarly both fixed and fluid. Episodic memories typically “have a specific source in time, space, and life circumstances” and “are often autobiographical in nature,” whereas semantic memories have to do with “facts about the world, about ourselves, and about other knowledge that we share with a community” (Baars and Gage 2010, 325–26). But these two kinds of memory are often related and combined. As Baars and Gage point out, for example, “semantic memories may be the neocortical residue of many episodic memories” (328) as particular instances are forgotten and “only the semantic knowledge remains.” Experiments have also shown that “performance on semantic tests is better if the participant also has some episodic memory associated” with the question—an image of a celebrity from a film or a TV show one has seen, for example, as opposed to a famous name that evokes no particular episodic recollection (Baars and Gage 2010, 329; see Westmacott et al. 2004). Even though I can only remember a few of the tennis matches I played in my youth, those episodes inflect my semantic memory of tennis, and neurons in my motor cortex associated with the acts of hitting a tennis ball are also probably activated in response to viewing matches or even reading about the sport (see Jeannerod 2006, Rizzolatti and Craighero 2004, and Rizzolatti and Sinigaglia 2008). And so when I watch a Wimbledon final, my procedural, episodic, and semantic memories all interact.
One of the functions of narrative is to coordinate the different processes of memory and thereby to cultivate our ability to engage in processes of binding and integration across different timescales. Episodic, semantic, and procedural memories are all involved, for example, in the five narrative codes that Roland Barthes identifies in S /Z (1974), his classic structural analysis of Balzac’s story “Sarrasine.” For example, the code of actions (ACT) is intelligible only because our procedural memories interact with specific episodic recollections and learned semantic categories for various kinds of action. The deployment of enigmas and the revelation of “truth” in the hermeneutic code (HER) is primarily semantic, but it also requires an interaction between short-term and long-term memory across the ever-shifting boundaries of working memory, and these processes of integration will range from subliminal priming effects to explicit, conscious acts of episodic and semantic recollection. The three codes that Barthes claims are exempt from temporal sequence—the designation of “semes” (SEM), or signifiers of people, places, and things; the “cultural codes” of conventional wisdom (REF); and the antitheses (SYM) structuring the oppositions in the text—all draw on memories of different kinds and so are not really outside of the time of textual processing.
The “weaving of the codes,” as Barthes describes it, that configures a particular text creates a temporal object that is a structure of memory and anticipation. Like a melody, a narrative consequently has a peculiar mode of existence that is neither completely reducible to the moments through which it passes nor entirely independent of them. Neither autonomous from its transient, ephemeral manifestations nor merely dependent on them, a narrative or a melody is a “heteronomous” object that we initially create in and through temporal experiences of anticipation and retrospection and that we are then able afterwards to recollect in various forms and from different perspectives (see Armstrong 1990, 20–43). This temporal heteronomy is especially visible, for example, in the paradoxical existence of a literary character like Isabel Archer or Marlow or Pip, whom we can designate and discuss without having to reproduce all of the sentences out of which they arose. These memorable characters would not exist without those sentences, but we can analyze them and talk about them without having to recite these word by word. Like a literary character, a narrative is a temporal object with a heteronomous existence because it is a construction of memory.
The ability to integrate temporal processes across various timescales—from microsecond binding at the neuronal level up through long-term declarative memories linking brain, body, and world—is necessary to create complex heteronomous objects like a narrative, a fictional character, or a symphony. Merlin Donald (2012, 39) argues that our species has developed a “slow process” of “intermediate-term governance” between the moment-to-moment binding of short-term working memory and the consolidations of long-term memory that uniquely equips us to produce and exchange such heteronomous objects. Longer than the moment-to-moment problem-solving of short-term working memory but more flexible and open-ended than long-term memory, this “slow process” of temporal integration, as he calls it, involves extended to-and-fro synthesizing activities that “can operate over long time frames” and require “a vastly extended working memory system” whose “prime function is to enable the mind to comprehend and navigate the multifaceted social-cognitive world that human beings inhabit” (38).
These mid-range powers of temporal integration undergird the human capacity for engaging in socially coordinated transactions like telling and following stories. Donald observes that “the capacity to achieve temporal integration on this scale seems to be absent in apes” (38) and our other close primate relatives, and he points out that it is necessary for “the comprehension of extended human social scenarios that engage several agents in interaction” (37) over hours at a time beyond what either short- or long-term memories typically entail—distributed, collaborative cognitive activities like exchanging stories, performing or listening to a symphony, or conducting a conversation. Activities like these require the coordination of “lengthy episodic experiences” that exceed the “pre-existing primate capacity for temporal integration” (36) in visual, aural, and haptic perception. According to Donald, “We have no good neural model of the activation or localization of such a long-lasting process” at an intermediate scale and (at least as yet) no firm empirical evidence of “a class of active neural traces that can last for hours on end governing decisions and maintaining the general direction of behavior and thought”—“yet this class of trace must exist,” he argues, “given the overwhelming evidence of autonomous sustained imagination, thought, and planning in human social life” (39).
The limitations of available brain-scanning technologies make it impossible to map the locations and interactions that characterize these mid-range, transactional processes, and so this gap in the evidence is not surprising. The technologies would need to have more temporal sensitivity and precision than fMRI measures of changes of blood flow in the brain and more anatomical precision than EEG readings of intracranial electronic activity off of sensors attached to the skull. Scanning methods would also need to map longer durations than these instruments can easily accommodate (imagine spending hours on end in the noisy, claustrophobic confines of an fMRI machine) and, additionally, would have to be designed to compare and calibrate these processes not just within a single brain but across two or more participants in an interaction.15
Whether any measuring equipment can do complete justice to the brain-body-world interactions through which we create and exchange stories is doubtful because quantitative measures inherently miss “what it is like” to have an experience that the as-if quality of narrative figuration is uniquely able to convey (see chapter 4). Still, if the technology were ever to develop sufficient sensitivity and sophistication to overcome its current limitations, a heteronomous temporal object like narrative would be a promising candidate for studying the “slow process” of mid-range temporal integration. We can tell and follow stories because our species has evolved the cognitive capacity to pass back and forth between small-scale temporal binding and long-term memories in an intermediate range of sustained, intersubjectively coordinated temporal activity. The ability of narratives to cultivate and coordinate these “slow processes” is in turn fundamental to the social and distributed cognition that characterizes the life of our species. Stories facilitate processes of temporal integration along a range of timescales and across a distributed social network of interacting, embodied subjects, and this is an important source of their social power.
Our sense that we share time with others is integral to the lived experience of intersubjectivity, even as our inability to fully fathom someone else’s passing moment is evidence of the solipsism we can never fully overcome because we cannot know what it is like to inhabit another embodied consciousness with its own perspective on the world. The paradoxes of social time embody and enact what Merleau-Ponty (2012 [1945], lxxvi) memorably calls “the paradox of the alter ego,” the peculiar combination of indubitable community and insuperable isolation that characterizes our relations with others. As he explains:
Of course, another will never exist for us as we exist for ourselves: . . . we are never present at the thrust of temporalization in him as we are in ourselves. But unlike two consciousnesses, two temporalities are not mutually incompatible, . . . because they can intertwine. . . . Since my living present opens up to a past that I nevertheless no longer live and to a future that I do not yet live, . . . it can also open up to temporalities that I do not live and can have a social horizon . . . [and a] collective history that my private existence takes up and carries forward. (457)
As Heidegger (1927) similarly observes, being-in-time and being-with-others are correlative existential structures. The felt analogy in our experience between the horizons of the past and future and the horizons of other worlds allows these aspects of existence to intertwine, and these interactions are crucial to the social work of narratives.
Just as we experience the passing moment as a paradoxical structure of presence and absence where past and future intersect, so we experience other embodied subjectivities as horizonal to our world, simultaneously present and absent in ways that inform and are informed by our experience of time. Others are undeniably “there,” present in our worlds, inextricably related to us, with a perspective complementary to what is open to our view. But the presence of others is also anticipatory, based on expectations about how their perspectives would probably complete what is disguised from our standpoint. The horizon of the past includes experiences with others that give rise to these expectations, but this future is inherently indeterminate, not yet completely specified, and our confidence in what it may hold is based on assumptions about what others experience that necessarily lie beyond what we can ever fully sense and know.
Narratives exploit the paradoxes of intersubjective time to various effects—drawing on and reinforcing or subverting and playing with our social experience of temporal horizonality. A multiple-plot novel like Anna Karenina, for example, relies on the reader’s sense of the overlap and interconnection between the temporalities of different worlds in order to persuade us to coordinate the developments emplotted on implicitly correlated time lines. We assume that the development of Anna and Vronsky’s adultery parallels in time the starts and stops of Levin and Kitty’s courtship even though these stories actually cross only occasionally. Our assumption of the horizonal intertwining of these worlds and their temporalities is so compelling, indeed, that most readers do not notice that Tolstoy got it wrong. As critics who have carefully studied the structure of the novel point out, “the timeline of the Anna and Vronsky plot is out of sync with the Kitty and Levin plot by an entire year, while Dolly’s story lags behind even more” (Auyoung 2018, 17; see Alexandrov 1982). When Anna and Levin finally meet toward the end of the novel, their encounter is strictly speaking impossible because of this temporal incongruity—but its absurdity is invisible to most readers because our immersion in a coherently developing illusion draws so effectively on our everyday assumption that the gears of time join worlds together. This assumption usually operates silently in Tolstoy’s novel, but sometimes it becomes explicit, as in the well-known repetition of the steeplechase scene, first from Vronsky’s perspective as a participant and then from Anna’s as she observes the race from the stands, a repetition that makes visible the otherwise tacit workings of intersubjective temporal horizonality by separately depicting correlated experiences of the same sequence of events.
This is also the temporal principle on which Ulysses is structured, with Stephen and Bloom traversing separate paths over the course of a day (Bloom first briefly spotting the young man during the drive to the cemetery) until they meet at the maternity hospital. The coordinated horizonality of intersubjectivity and time is a principle the novel makes explicit and dramatizes in the “Wandering Rocks” chapter, which maps the interactions of various characters’ intersecting worlds during the regal procession, and some critics consequently view this chapter as a small-scale model of the novel as a whole (see Blamires 1996). We do not hesitate as readers to make these temporal linkages because the horizonal parallels assumed by Tolstoy’s and Joyce’s narratives invoke our lived experience of time as an intersubjective intertwining of distinct but coordinated temporalities.
The temporal experiments of modern fiction dramatize and play with the paradoxes of social time, but these sometimes “unnatural” effects depend once again on the “natural” cognitive processes they invoke. For example, Virginia Woolf’s representation in Mrs. Dalloway of urban life as an intersubjective intertwining of separate, opaque worlds is reinforced by the ringing of Big Ben and the other bells of London’s clocks as the narrative passes from one world to another. This “natural” effect is echoed by the artificial poetic refrain “the leaden circles dissolved in the air,” which the narrator repeats as the ripples of sound cross the horizons that paradoxically join but also separate the worlds of Clarissa, Peter, Septimus, and Rezia. This recurring metaphor is both unnatural and natural, rendering in textual, figurative terms how waves of interaction (like dissolving circles) connect passing moments and intersecting worlds. It is a poetic device, created and repeated by the narrator over the time of the discourse, that has the function of forging textual connections across the different stories joined in the narrative, but it is an artifice with a natural basis because it is an attempt to suggest a physical, embodied correlative for how the temporalities of disparate, otherwise unconnected worlds are joined in intersubjective experience.
Faulkner’s games with time in Sound and the Fury are more disruptive and subversive, but they too are a metafictional staging of the interaction between social and temporal horizonality. The challenge of aligning the temporalities of this novel’s different worlds begins with the disjunctions between the dates marking the four chapters, which are titled “April Seventh, 1928,” “June Second, 1910,” “April Sixth, 1928,” and “April Eighth, 1928” (the last of which also happens to be Easter Sunday). Dates on a calendar coordinate the passing of time socially, and the intersubjective horizonality of time is especially marked on collectively celebrated holidays (even more so a holiday with symbolic resonances of death and rebirth). Because each of these chapters anchors the narrative present in a date, the reader assumes that the time they narrate is intersubjectively shared—but how these temporalities are linked is not only thrown into question by the fact that the dates are out of sequence but also (as the reader soon discovers in the initial pages of Benjy’s narration) by the narrative’s jumping back and forth across time without the guideposts provided by a clock or a calendar (in narrative terms Quentin’s act of smashing his watch is a redundant joke). Making sense of this disjointed narrative requires coordinating the time shared by differently experienced events repeated across the various chapters (Caddy’s muddy drawers, the older Quentin’s suicide, the younger Quentin’s absconding with Jason’s stash, etc.).
This work of intersubjective and cross-temporal consistency building is, of course, a routine aspect of following even the most ordinary story. Such invisible, everyday cognitive processing, however, is precisely what Faulkner both invokes and subverts. These disruptions in turn call attention to the coordination of differently experienced moments in our intersubjective sense of time passing—horizonal intertwinings that we ordinarily do not notice in life or in mimetic narratives because they are so familiar, but that Faulkner denaturalizes and defamiliarizes and thereby brings into view.
There is a good deal of experimental evidence demonstrating that, as Sylvie Droit-Volet (2014, 496) points out, “people who interact with one another experience a more similar passage of time than people who do not interact in the same way.” Shared emotions are one basis of such interactions. As William James (1950 [1890], 1:618) observes, our feeling of time “harmonize[s] with different mental moods.” A primary example of the intertwining of temporality and intersubjectivity, our emotions attune us to one another by coordinating our sense of time. When stories invoke emotional responses, they intertwine us with each other by coordinating the temporal processes through which our experience of the narrative unfolds. The social power of narrative emotions derives from their ability to intertwine our experience of time.
Whether we perceive time to be passing faster or slower, for example, will vary according to the emotion we are experiencing, and this subjective phenomenon is intersubjectively shared. Experiments have shown that anger, fear, and other threatening situations cause time to be “judged longer than normal,” producing a slowdown in our sense of time passing that enhances “the automatic preparation of the organism to act or to move” (Droit-Volet 2014, 488). Not all arousal affects our sense of time in the same way, however. For example, as Droit-Volet observes, “the perception of ashamed faces has been found to produce not an overestimation but rather an underestimation of time” (491). Some of these effects are odd and counterintuitive. For example, although disgust might seem to interfere with our sense of time, experiments show that viewing disgusted faces has no effect on the ability of subjects to report time spans accurately (see Droit-Volet).
The explanations psychologists give for why time seems to pass faster or slower in these experiments are speculative, but there is impressive evidence that variations in what is called “subjective time” are not merely phantoms. Experiments have demonstrated that the emotions generated by our embodied interactions with each other and our worlds have predictable effects on our sense of how time passes, and these subjective feelings are grounded on empirically measurable “changes in bodily states” (Droit-Volet 2014, 494; see Barsalou 1999). Someone experiencing anger, fear, or a threat will predictably report that a two-minute interval felt much longer, and someone undergoing shame or guilt will more often than not sense that a five-minute span was much shorter. These phenomena are not perfectly uniform across everyone (you may be reading this and saying to yourself “not me!”), but they are not random, and they are broadly shared. These similarities are compelling evidence of how subjectively felt experiences can intertwine in shared time.
There has been considerable debate about how to categorize emotions and about whether they are culturally variable or uniform across our species. Beneath these disputes, however, the scientific consensus is that emotions are embodied cognitive processes that guide our attitudes toward the world and that coordinate our relations with one another (see Barrett 2007, 2017). The ability of emotions to do this work depends on their power to orient us temporally, and that is why the future is the tense most frequently associated with emotions. Heidegger describes emotional attunement (our mood, or Stimmung) as fundamentally futural because it orients our anticipatory understanding of people, texts, and other states of affairs (on the fore-structure, or Vor-Struktur, of understanding, see Heidegger1962 [1927], 172–95). Emotions are configurative cognitive structures that attune us to our worlds, and this attunement is intersubjective and futural. Narrative emotions coordinate our reactions to the storyworld by orienting and manipulating our expectations.
Emotion and cognition are inseparable. As Lisa Feldman Barrett (2007, 390) observes, the “brain structures” and “neural circuitry” for emotion and cognition are interconnected, overlapping, and intertwined. Unlike, say, vision and hearing, which operate through distinct, anatomically identifiable neural systems (the visual cortex as opposed to the auditory cortex), the cortical structures supporting emotions are inextricably interwoven with other cognitive processes. Consequently, as Barrett observes, “No one would ever mistake seeing for hearing (although one sensory representation might trigger another), but the same cannot be said for feeling and thinking” (390). Not simply a private, internal state of affairs, “an experience of emotion is an intentional state—it is an affective state that is about something” (379)—attuning our attention and modulating our responses to situations we encounter. Emotions have meaning and value because of their intentionality—their directedness toward people, places, and things. This orientation is inherently temporal because, like other kinds of intentionality, it anticipates a certain kind of continuation of its perspective in experiences beyond our immediate horizon. Such horizonality characterizes the emotions we experience in life as well as those invoked by the stories we hear and read.
Consider, for example, the so-called basic emotions that Paul Ekman (1999) identified and that he controversially claimed are culturally universal: anger, sadness, fear, surprise, disgust, contempt, and happiness. Ekman compiled this list by showing pictures of facial expressions to members of a tribe in a remote region of Papua New Guinea whose exposure to outside cultures had been limited, and he found that faces expressing these emotions were immediately recognized—evidence, he concluded, that they must be biologically based universals, not cultural constructs, because they had not been learned from representations in external media. Recent research about the neurobiology of emotions has cast doubt on Ekman’s claims that emotions are species-wide, neurobiologically based constants (see Colombetti 2014, 25–52; Barrett 2017). It turns out that there are no “neural signatures” that clearly demarcate different emotions, and there is considerable variability in the relation between particular emotions and the neuroanatomical structures on which they are based. Fear involves many parts of the brain and the body besides the amygdala, for example, and the amygdala is involved in a range of different emotions associated with novelty and uncertainty and not just responses to threats (see Lindquist et al. 2012).
Current research suggests that emotions are not products of universal characteristics of our neural anatomy but rather that emotional responses develop as our bodies interact with our worlds, and that it is consequently impossible to separate cleanly what is learned and what is biologically inherited (see Gross 2010, Colombetti 2014, and Barrett 2017). As Merleau-Ponty (2012 [1945], 195) argues, for human beings “everything is constructed and everything is natural, in the sense that there is no single word or behavior that does not owe something to mere biological being—and, at the same time, there is no word or behavior that does not break free from animal life.” As evidence for this claim, Merleau-Ponty observes that “the gesticulations of anger or love are not the same for a Japanese person and a Western person. . . . When angry, the Japanese person smiles, whereas the Westerner turns red and stamps his foot, or even turns pale and speaks with a shrill voice” (194–95). Merleau-Ponty concludes: “Having the same organs and the same nervous system is not sufficient for the same emotions to take on the same signs in two different conscious subjects. What matters is the manner in which they make use of their body, the simultaneous articulation of their body and their world in the emotion” (195). As embodied expressions, emotions are biocultural hybrids.
If Ekman’s “basic emotions” are indeed widely evident around the globe, they would have become established through a combination of biological and cultural mechanisms because, in a Darwinian manner, various comparative advantages were attached to them. These advantages could have accrued only because emotions are intentional structures—that is (recalling Barrett’s description), “affective states about something” in the social and natural world. In an important critique of ideas about emotion that have been very influential in the humanities although they are scientifically dubious, Ruth Leys (2011, 437) notes that many so-called affect theorists have embraced Ekman’s model because, in their view, it “assumes that affective processes occur independently of intention and meaning.” As Leys explains, these theorists regard “affects” as “noncognitive, corporeal processes or states,” intensities that are “capable of discharging themselves in a self-rewarding or self-punishing manner without regard to the objects that elicit them”—“rapid, phylogenetically old, automatic responses of the organism that have evolved for survival purposes and lack the cognitive characteristics of the higher-order mental processes” (437–38; also see Wehrs 2017, 34–41). If emotions were “noncognitive” intensities that lack “intention and meaning,” however, they could not possibly have evolutionary value, and they would not have survived. Claims for the “autonomy of affect” (see Massumi 1995, 2002) are ill equipped to account for the social work of emotions, whether over the long duration of evolutionary history or for the shorter spans of narrative interactions through which we are made to feel together about a story.
Basic emotions could proliferate and endure only if they conferred a survival advantage by helping us respond to recurrent, typical problems or opportunities, and a powerful advantage would be how they helped us and our conspecifics coordinate our responses to states of affairs in our shared worlds. These states of affairs may be objects or situations about which we feel (say) fear, surprise, or disgust, or they may be other members of our species with whom we experience (citing Ekman’s other categories) sadness or happiness or for whom we feel contempt. Such emotional responses are not only intersubjective and intentional (in the sense that they are directed toward something in the world); they are also temporal, because they predispose us to act one way or another in the future (different kinds of behavior would align, for example, with anger, happiness, disgust, etc.). By intertwining intersubjectivity and temporality, emotions may convey powerful survival benefits because they can facilitate collaborative interactions. The reverse side of this coin, of course, is that emotions can also fuel destructive conflicts through this same capacity for coordinating our attunement to the world and aligning aggressive and violent responses to rivals or perceived threats. The emotions evoked by narratives can similarly have either pro- or antisocial consequences because they can intertwine our worlds toward different ends, promoting compassion or stoking fears and anxieties about matters beyond the horizons of our worlds (see Armstrong 2013, 131–74).
Whether for good or for ill, emotions are embodied intersubjective structures that attune us to our worlds, and this attunement is a temporal phenomenon that orients us toward the future. This is the case not only with so-called basic emotions but also with what Suzanne Keen (2011, 6n) calls “complex emotions” that are indisputably (and even by definition) cultural constructs: “moral sentiments (empathy and compassion) and social emotions (shame, embarrassment, envy, guilt and hatred as well as the positive calm and amae, or comfort in belonging).” “Moral sentiments” and “social emotions” have intentionality because they are “affective states about something”—attitudes directed toward others whose experiences lie beyond what we can fully share but to whom we respond across our horizons. This attunement to others is also inherently futural, orienting us toward a horizon of anticipated interactions. Even the entries “calm” and “comfort” in Keen’s list, which might seem to be about nothing other than themselves, project us in an anticipatory manner toward future involvements with others (as in the slogan “Keep Calm and Carry On”). Not purely subjective and internal, these emotions entail attitudes that respond to and configure our relations with others and the circumstances we find ourselves in. This capacity to respond to and configure situations we share with others is one reason why emotions of all kinds, whether “basic” or “complex,” can be patterned into stories.
As intentional structures, emotions are characterized by the to-and-fro motions of recursivity that are fundamental to narrative. This is evident, for example, in the cycle that Jenefer Robinson (2005) describes between “noncognitive affective appraisals” and subsequent “cognitive monitoring.” Drawing on the work of Joseph LeDoux (1996), Robinson (2005, 45) describes affective appraisals as “swift and automatic” physiological responses, mediated through the amygdala and the thalamus, “without any conscious deliberation or awareness.” According to Robinson’s model, these intuitive, bodily-based reactions are then followed by “a slower, more discriminating processing system which operates through the cortex and figures out whether the thalamo-amygdala ‘affective appraisal’ is appropriate or not” (50). This model is better able to account for the social work of narrative emotions than theories that view affect as autonomous and outside the realms of intentionality and judgment. It is preferable not only to the affect theory that derives from Ekman but also to Hogan’s [2011b, 46] influential account of “what literature teaches us about emotions” that wrongly claims that “emotions are not produced by the evaluation of situations or events in relation to goals.”
Robinson’s term “noncognitive appraisal” is an oxymoron, of course, because an appraisal necessarily entails cognition. She resorts to an oxymoron, however, in order to do justice to the nonsimultaneity of the brain’s cognitive processes. Affective appraisals are part of the cognitive apparatus that makes it possible for us to respond to situations in the world before consciousness has a chance to form. The delay between prereflective perception and the arrival of consciousness is not a “thought-o-genic lapse” (to recall Massumi) but an interval filled by affective appraisals. Emotions play an integral role in the to-and-fro processes that couple anticipatory attunement and retrospective revision as we review and rearrange what we have subliminally predicted across the gap that Libet identified in his “mind-time” experiments. This temporal doubling of reflection on unreflected experience is not a static, unbridgeable divide between affect on one side and meaning or cognition on the other. It is, rather, a temporal to-and-fro between two different kinds of intentionality—between what Merleau-Ponty (following Husserl) calls “non-thetic, operative intentionality” at a primary perceptual level and “thetic, act intentionality” when we take up a posture of judgment or deliberation (see 2012 [1945], lxxxii). Whenever we reflect, as Merleau-Ponty points out, we discover not a meaningless, autonomous realm of impulses or intensities but rather a world of significations already there in our prereflective perceptual experience. Reflective monitoring may come after emotions have already projected preliminary attitudes, orientations, and assessments, but affect is (once again) not an autonomous system that operates apart from consciousness and evaluation.
The recursive interaction between anticipatory affective appraisals and retrospective, revisionary monitoring is the experiential correlate of neurobiological processes that are constantly forming and dissolving at different levels and rates. Recall, for example, Merlin Donald’s tripartite description of different processes of integration: short-term and working memory in immediate experience, the consolidation of sequences of events in long-term memory, and the intermediate range interactions of conversations and narrative exchanges over longer durations. Similarly, Varela (1999, 300) proposes that we “distinguish three scales for affect,” roughly corresponding to three different “scales of temporality,” with specific terms for each level: “The first scale is emotion: the awareness of a tonal shift that is constitutive of the living present. The second is affect, a dispositional trend proper to a coherent sequence of embodied actions. The third is mood, which exists at the scale of narrative description over time.” These scales are not exactly equivalent to Donald’s levels or to other ways of marking the temporality of cognitive integration (recall the distinctions, for example, between the microsecond level of subliminal processing [100–150 milliseconds], the delayed emergence of conscious awareness [at 500 milliseconds], and the multisecond parsing of temporal windows [at 3–5 seconds]). But Varela’s suggestion that a vocabulary for emotions might be based on temporal processing levels is intriguing because it follows from the recognition that emotions are temporal cognitive structures.
More than in most other experiential domains, however, coming to agreement about the definition of terms in emotion studies is difficult. Varela’s distinctions are an interesting attempt to create order in an inherently fuzzy, ambiguous area, but his definitions themselves demonstrate why emotional categories are unstable. The processes of affective appraisal he is trying to pin down cross back and forth across the temporal boundaries he describes. His emotion categories blur into one another because of the recursivity and interconnectedness of the temporal processes on which they are based. For example, Heidegger’s notion of mood (Stimmung) refers to the anticipatory attitude of attuned expectation that characterizes not only the third scale in Varela’s scheme to which he assigns this term but the other two as well (“emotion” and “affect”). Gearing us toward the future is something emotion does across all of the various scales of temporal integration, even if this attunement may feel differently (if it is perceived at all) according to the level at which it operates (experiential differences that Varela’s terms attempt to capture). The terms used by emotion researchers criss-cross these domains, as do the embodied, temporally fluid processes Varela’s distinctions try to pin down. It may be useful and informative to ask what temporal level an emotion term refers to and what kind of attunement it describes, but agreement on a strict definitional taxonomy is unlikely, and fixed, clear-cut categories would reify and misrepresent this fluidity.
The play of emotions in reading and narrative enacts in different ways the basic temporal duality between affective appraisal and cognitive monitoring. For example, David Miall (2011, 330) has found considerable experimental evidence for “the role of emotion in initiating and directing cognitive processes while reading.” Miall has attempted to measure what he calls “the anticipatory role of feeling” in the comprehension of stories, and he has reported that test subjects respond to the emotional valences of words “early in verbal response, around 250 msec after a word is first seen, or possibly earlier”—in any case before the processing of word meaning is complete (336, 327). There is also interesting experimental evidence that aligning our affective appraisals with the emotional valences of a text facilitates comprehension. Lawrence Barsalou (2008, 629) reports, for example, that “when facial emotion matched sentence emotion, comprehension was better than when they mismatched.”
Text comprehension is not a static correspondence of affect and word meaning, however, but an interactive, temporally dynamic process of adjustment and readjustment. As Miall observes (2011, 327), “Several studies focused on the temporal unfolding of the first responses to verbal presentations” document a to-and-fro interaction between emotionally oriented anticipation and retrospective cognitive activity, consistent with the recursive cycle of affective appraisals and cognitive monitoring that Robinson describes. Emotions can set this loop in motion because they are configurative structures, projecting in their appraisals a part-whole gestalt that guides comprehension—a shape that is futural in orientation because it waits to be filled in, revised, refined, or overturned.
How a narrative invokes and manipulates the reader’s or listener’s emotional responses is integral to how it builds, breaks, and re-forms patterns of concordant discordance out of the events it configures. For example, Meir Sternberg’s (1987) triad of affective attitudes fundamental to narrative—curiosity, suspense, and surprise—are all horizonal phenomena through which a story sets up expectations that it then confirms, revises, or overturns. Curiosity, suspense, and surprise are attitudes of affective appraisal that guide the configurative work of narrative comprehension by attuning us to future developments.
Emotions project a structured but open, incompletely determined horizon of expectations that orient the temporality of understanding. This horizon of expectations extends from the emotions invoked by a text’s genre—will this be a comedy or a tragedy?—to the twists and turns of events in the plot—are we headed for a happy ending or a good cry? As Robinson explains, “Emotions focus attention” (126). This focus is both directed and open, oriented but not fully specified, in the manner of a horizon beyond which we anticipate certain kinds of continuation of the perspective available from our standpoint—but not all kinds. We can be surprised by some developments because they are not what we had expected, even if what we anticipated was not completely determined in advance (as when we react to the surprising ending of a novel or a film: “I couldn’t tell you exactly what I thought would happen, but not that!”).
A work’s characters are obviously a primary focus of the emotions generated by a work. What we may feel about a particular character matters, however, because of how it may orient our attitudes toward much larger concerns. Our changing responses to a work’s characters—the pity and fear we may feel for the tragic hero, for example—are central to our developing sense of the narrative’s overall emotional tonality, and this tonality has broad hermeneutic implications. As Robinson explains:
Emotional understanding “regestalts” the world in a global way: in responding emotionally to Anna Karenina, I see the whole world of the novel through the prism of that emotion. My feelings for her affect my feelings about the harshness of the marriage laws, the difficulties of Vronsky’s professional position, the heartlessness of polite society, and so on and so on. Feeling compassion for Anna is not just a response to her but a response with wide implications for my understanding of the novel as a whole. (128)
Our changing emotional reactions to the fate of a particular character can have such broad ramifications because emotions entail affective appraisals that orient and attune our attitude toward what lies beyond our horizons. This will include the whole of the text of which the character is a part, but it need not stop there. The “beyond” of our emotional response to a character starts with the immediate future of our reading, and it may then extend across the horizon of our engagement with this particular story to other involvements that await us in our worlds. Because emotions are horizonal configurations that orient us to the future and to other people, manipulating our emotions is a powerful way for narratives to attune us toward the world at large—to “regestalt” our sense of the world’s patterns far beyond the particular events, characters, and concerns in the story.
Narrative emotions intertwine temporality and intersubjectivity and connect us to other worlds by aligning our horizons toward what lies beyond our view. These emotions connect us not only to fictional characters whose desires and struggles we may sympathize with as we respond to their stories but also to other listeners and readers with whom we share narratives. The experience of telling and following stories coordinates subjectivities in time. As Hanne De Jaegher, Ezequiel Di Paolo, and Ralph Adolphs (2016, 5–6) explain, neuroscience has begun to acknowledge the importance of a range of phenomena that entail a kind of “inter-brain phase synchronization”—processes of “mutual coordination” and “participatory sense-making” that require “an interactive sharing of socio-cognitive processes” (also see De Jaegher, Di Paolo, and Gallagher 2010). What De Jaegher and her colleagues call the “interactive brain hypothesis” also pertains to the synchronization of teller and listener in the exchange of stories (see my discussion of brain-to-brain coupling in chapter 4). As Ian Cross (2003, 48) points out, these processes of intersubjective synchronization have their earliest instantiations in “the rhythmicity of caregiver-infant interaction” that in turn is “central to the development of human significative and communicative capacities.” This to-and-fro coordination of subjectivities in a temporal rhythm of collaborative sensemaking carries over into adult experiences in music and dance and other instances of sharing patterned time and harmonizing emotionally with others.
As De Jaegher and her coauthors (2016, 8) explain, these kinds of “socially coupled interaction” give rise to “emergent dynamical patterns that do not reduce to the activity of” a single agent or brain. Hans-Georg Gadamer (1993 [1960], 101–10) has something of this sort in mind when he describes a game as a collaborative activity that seems to have a trans-subjective agency of its own that paradoxically arises from the activity of its players, even as it goes beyond what they can individually command. What Gadamer calls “the primacy of the game over the players engaged in it” is experientially evident in the control the game seems to have over the players’ actions as they respond to movements generated by their agency but not entirely subject to their control. Games, music, dance, and the exchange of stories are all emergent intersubjective experiences of participatory sensemaking that would not come into being without the contributions of individual subjectivities who strangely but compellingly are not fully in charge of what they co-create but who find their responses guided and coordinated by these interactions. These kinds of experiences of mutual coordination are evidence of the emergent dynamical systems that the interactive brain hypothesis attempts to explain.
Exactly what is happening in the bodies and brains of participants in such interactions is not entirely clear because of the limitations of scanning technologies. But this kind of mutual entrainment of separate but linked dynamical systems is not unusual in the physical world. Karl Friston and Christopher Frith (2015, 391) compare it to the synchrony of swinging pendulums, a well-known instance of “the action at a distance among coupled dynamical systems.” Two or more swinging pendulums miraculously seem to coordinate their movements, but this “emergence of generalized synchronization” is a common phenomenon that occurs whenever two oscillatory systems “are coupled to each other through action” (391). Experiences of rhythmical, temporal coordination between people suggest that this kind of coupling also takes place when two embodied brains interact. The rhythms of an embodied brain—“the oscillatory dynamics associated with binding, attention, and dynamic coordination” through which neuronal assemblies are formed, dissolve, and re-form (391)—are not unlike the to-and-fro movements of a swinging pendulum.
The capacity of these oscillations to become synchronized across two or more embodied brains is the neurobiological basis of the temporal coordination of collaborative meaning making processes in music, dance, or stories. These coupled oscillations are what make it possible for embodied brains to intertwine intersubjectively in time, and they are the neurobiological means by which aesthetic experiences synchronize the temporalities of different worlds. We have the ability to share rhythms and emotions in time because the oscillatory processes of our horizonal interactions with each other can synchronize. Coupling the dynamical systems through which cognitive patterns are formed and dissolved in narrative interactions, waves of synchronization can join the worlds of readers not only with the worlds projected by stories but also with the worlds of other readers, listeners, and storytellers through the to-and-fro rhythms of narrative exchange.
The bodily resonances preserved and transmitted through these rhythms of concord and discord, synchronization and desynchronization, give stories the power to align worlds across the horizons of history. As Gallagher and Zahavi (2012, 95) observe, “human time is the time of our life stories,” and “the story of any individual life is not only interwoven with those of others (parents, siblings, friends, etc.); it is always embedded in a larger historical and communal meaning-giving structure.” Our sense of embeddedness in communal time comes in large part from the stories we are told and learn to tell and that circulate in our culture. As members of historical communities, we inherit, preserve, transform, and pass on stories and other artifacts of our shared culture, and this is how our lives intertwine intersubjectively with others beyond our horizons. We participate in history by passing these stories on to those who come after us, some whom we know but many others whom we will never directly encounter. As Ricoeur (1980b, 188) argues, “Narrativity . . . open[s] any meditation on time to another horizon than that of death, to the problem of communication not just between living beings, but between contemporaries, predecessors, and successors.” In this sense, he contends, “narrative time [is] a time that continues beyond the death of each of its protagonists” inasmuch as the exchange of stories enables “transmission between generations” (188). These processes of temporal transmission are horizonal and intersubjective, an intertwining of my time with the time of others beyond the limits of my life.
The temporal asynchrony of the embodied brain means that our horizons are always shifting and that we are never at one with ourselves for as long as we live, but the horizonality of temporal experience also makes possible the intertwining of my temporality with the time of others, and these open-ended interactions may reverberate long after we are no longer there to experience them ourselves. Death is not necessarily the end of the story, then, because we can live on beyond the horizons of our own mortal lives by the stories that we learn, create, and exchange in collaborative processes of cultural meaning making. There is life after death in the stories we pass on, and the neurobiological basis of this capacity for generativity is the asynchronous temporality of the embodied brain.