THE TEMPORALITY of the decentered brain makes mimesis possible because imitation is not a static correspondence of a sign to a thing but a dynamic configuration of an action. Mimesis is a kind of action (a linguistic making) that produces an organization of events (an emplotment of actions) that the reader or listener follows and reconstructs (the activity of comprehension). Action is central not only to mimesis but also more broadly to cognitive life. This is why the building and breaking of narrative patterns can powerfully affect the formation and disruption of patterns across many different cognitive domains. The paradox of concordant discordance in narrative traverses the circuit of action that joins narration, story comprehension, and everyday embodied cognition. The ever-shifting balance and tension between concord and discord within, across, and between these modes of action is what allows stories to play with the brain’s competing requirements for order and flexibility, organization and openness to change.
Recent work on enactive narratology by so-called second-generation cognitive literary theorists has brought renewed attention to the relation between action and narrative, but this has been a central topic of narrative theory since its beginnings.1 Aristotle (1990 [335 BCE], 7; original emphasis) famously claims that “tragedy is an imitation not of men but of action” and, further, that “performers act not in order to imitate character; they take on character for the sake of [imitating] actions.”2 This priority placed on action, according to Ricoeur (1984a, 34), “excludes any interpretation of Aristotle’s mimesis in terms of a copy or identical replica. Imitating or representing is mimetic activity inasmuch as it produces something, namely the organization of events by emplotment.” Because narrative and action are inextricably intertwined, Ricoeur argues that “there is no structural analysis of narrative that does not borrow from an explicit or an implicit phenomenology of ‘doing something’ ” (56). As he explains, “the composition of the plot is grounded in a pre-understanding of the world of action, its meaningful structures, its symbolic resources, and its temporal character” (54). Understanding narrative consequently requires understanding the role of action in our cognitive lives.
Contemporary neuroscience suggests that an action-perception circuit makes action fundamental to many cognitive processes that might seem remote from motor control. In Andy Clark’s (2016, 7) evocative words, “Perception and action are . . . locked in a kind of endless circular embrace.” There is a growing consensus among neuroscientists that “action is rich in cognitive resources” (Schulkin and Heelan 2012, 224) and, even more, that “action influences perception at its very source” (Berthoz and Petit 2008, 46). As Alain Berthoz and Jean-Luc Petit (2008, 49) explain, it is impossible to demarcate “a frontier between sensation and motoricity, because action is already to be found in perception.” Recent experimental evidence on the responsiveness of the brain to imagined action and even to action words suggests that the brain is primed to respond to linguistically staged configurations of action and that these can have a profound effect on our cognitive processes because perception in many different modalities (vision, hearing, smell, touch) depends on embodied action. Plots can play a central role in structuring our understanding of the world because action is thoroughly implicated in perception.
Perception and action are interdependent in many ways. According to Alva Noë (2004, 8), the neurophilosopher who has studied this interdependence most thoroughly, “the basis of perception . . . is implicit practical knowledge of the ways movement gives rise to changes in stimulation.” As he points out, “the world makes itself available to the perceiver through physical movement and interaction” (1). The classic statement of this view is James J. Gibson’s Ecological Approach to Visual Perception (1979, 66): “A point of observation is never stationary, except as a limiting case. Observers move about in the environment, and observation is typically made from a moving position.” For all modes of perception, exploratory activity provides ever-changing information about regularities and irregularities in the environment, and it is these differences to which the organism responds. Noë consequently claims that “all perception is . . . touch-like,” even vision: “As in touch, the content of visual experience is not given all at once. We gain content by looking around just as we gain tactile content by moving our hands” (17, 73). As he notes, for example, “in normal perceivers, the eyes are in nearly constant motion, engaging in saccades (sharp, ballistic movements) and microsaccades several times a second. If the eyes were to cease moving, they’d lose their receptive power” because “optic flow contains information that is not available in single retinal images” (13, 20). It has long been experimentally established, for example, that fixed images on the retina gradually weaken and vanish, just as we don’t notice the clothes we wear, constant background noise, or other unchanging stimuli to which we become habituated.3
Seeing, hearing, and touching are all active processes that are especially attuned to difference and change. In vision, for example, the workings of opponency make the retina more sensitive to changes in light than to a uniform, constant illumination. As Margaret Livingstone (2002, 54–55) explains, retinal neurons have a center/surround organization that “respond[s] best to sharp changes,” which set off an opposition between these two parts of the cell, “rather than to gradual shifts in luminance” or to “the overall level of the illuminant,” which do not much if at all perturb the center-surround balance. As she notes, “it is much more efficient to encode . . . changes or discontinuities than to encode the entire image” because “the most information in an image is in its discontinuities” (also see Armstrong 2013, 65–66). Similarly, as Mark Bear and his colleagues (2007, 420) observe, the responsiveness of “warm and cold receptors” on the skin is “greatest during, and shortly after, temperature changes”; “with thermoreception, as with most other sensory systems, it is the sudden change in the quality of a stimulus that generates the most intense neural and perceptual responses.” In all modalities of sensation, the organism’s changing relation to its world produces differences to which our sensory equipment responds, even as changes in how we direct that equipment toward the world (moving our eyes or hands or the direction of our ears) can generate differences that are rich in information. A circuit joins action and perception because perception is an exploratory activity that produces and responds to differences.
There is perhaps no more vivid dramatization of the connections between action and perception than the opening of James Joyce’s novel Portrait of the Artist as a Young Man, where the infant Stephen Dedalus is depicted as coming to consciousness through various kinds of exploratory activity that register sensory differences and attempt to configure them into meaningful patterns. Joyce portrays the dawning of an aesthetic sensibility by calling attention to the circuit between action and perception and shows how their intertwining is fundamental to embodied consciousness. The novel famously opens with words Stephen hears from a story about “a nicens little boy named baby tuckoo” (2007 [1916], 5). This story stages a configuration of actions and relationships (“a moocow coming down along the road” [5] to meet the boy) that in turn initiates a series of attempts by Stephen to organize perceptual differences by acting in the world. First he listens, then he sings (“He sang that song. That was his song” [5]), and then he dances to a “sailor’s hornpipe” (5) that his mother plays on the piano. These modalities of action parallel a series of perceptual differences that he notices and tries to align. Moving from sound to vision, he first responds to the words he hears by looking at the speaker and noticing distinguishing facial features (“his father looked at him through a glass” and “had a hairy face” [5]), differences between speaker and listener that lead Stephen to identify himself: “He was baby tuckoo.” The relation between self and other is figured to Stephen through the relation of different perceptual modalities (what he hears, what he sees) that then move from sound and vision to touch (“When you wet the bed first it is warm then it gets cold” [5]) and even to smell (the oilsheet had a “queer smell” but “his mother had a nicer smell than his father” [5]). Stephen comes to consciousness of his world by experimenting with different kinds of action (listening, seeing, singing, dancing) and exploring the relations between perceptual differences (sight, sound, touch, and smell).
Action is meaningful and can be a resource for perception because it is figurative. Marc Jeannerod (2008, 103), a leading authority on motor cognition, points out that “human infants as young as three months old are visually sensitive to the difference between the biological motion of dots produced by a walking person and the random, artificially produced, non-biological motion of similar dots.” Infants understand biological movements as meaningful actions because they perceive them as a gestalt, a figurative arrangement of elements in a pattern. As Shaun Gallagher (2012, 175–76) observes, however, “the meaning of any action is not purely intrinsic to its motoric aspects.” Gregory Hickok (2014, 136) similarly notes that “the meanings simply aren’t in the movements,” inasmuch as “the very same motor program . . . could mean” any number of things depending on the context and purpose of the action (is a raised hand a vote, a request to speak, a signal to attack, or simply a stretching motion?).4 Context and purpose matter because an action is figurative. To perceive a movement as an action is to understand the pattern of part-whole relations that form its gestalt, its configuration as a means directed toward an end in a particular situation.
The idea that all consciousness is consciousness of something—phenomenology’s classic definition of intentionality (see Armstrong 2012)—also applies to action. As Gallagher explains, “Actions have intentionality because they are directed at some goal or project, and this is something that we can see in the actions of others” (77). Our actions are not empty and random but are aimed at states of affairs we engage with and point toward ends we anticipate. This does not mean that we are always aware of this directedness—intentionality is not the same as intention—but rather that an action is meaningful because it has what Heidegger (1962 [1927], 188–95) calls an “as-structure” and a “fore-structure.” We understand action X as signifying Y because it is projected toward Z.5 What a raised hand means depends on what we perceive it as directed toward—the pattern to which it belongs and the ends it projects and aims to produce.
The gestalt processes at work in understanding and producing action are evident in oft-cited fMRI experiments that measured the brain activity of dancers observing a dance style they have or have not learned to do (classical ballet, for example, versus capoeira) and that then compared these scans to novices and choreographers. Dance lays bare the configurative structure of action that ordinarily goes unnoticed in our everyday lives but that allows us to interpret actions differently according to our familiarity (or unfamiliarity) with their underlying patterns, the as-relations that shape them into meaningful gestalts (see Sheets-Johnstone 1966, 2011). As Catherine Stevens and Shirley McKechnie (2005, 247) point out, in these experiments “brain activity was affected by whether observers could do the action or not,” and these differences turned up not only in the motor cortex but also “in middle temporal areas suggesting semantic categorization of dance movement by experts” into the “vocabularies of ‘steps’ that constitute classical ballet and capoeira repertoires” (also see Calvo-Merino et al. 2004).
The brains of dancers responded differently to actions they could perform as opposed to those they could not because the dancers interpreted these movements as a configuration that is either familiar or strange. These patterns of response show up not only in the motor cortex but also in other areas of the brain because the steps have meanings that are not reducible to our physiological equipment for generating bodily movements. For the same reasons, scans comparing dancers to novices and choreographers showed that these steps are invisible to novices, whose brains do not respond to them, and that they have a different significance to choreographers, who imagine and create the patterns that dancers must then assimilate and reproduce and whose fMRI images differ accordingly, with more activity in cortical areas involved with planning and reflection. Not simply a one-to-one mirroring process, these experiments show that understanding actions entails configuring movements as something according to our past experiences and future expectations about their purpose, shape, and direction.
Both action and perception consequently have a virtual dimension that makes them a paradoxical combination of presence and absence, immanence and transcendence—a duality that Merleau-Ponty describes as the “intertwining” of the “visible and invisible” (see 1968 [1964], 3–49, 130–55). Any situation of perception is riven with absence, according to Merleau-Ponty, because its perspective is partial and incomplete, with invisible sides and undisclosed dimensions that we anticipate could be filled in by future experiences or by other, differently positioned observers. These indeterminacies make perception horizonal not only spatially (the hidden sides of an object) but also temporally (the not yet evident aspects we anticipate experiencing) and intersubjectively (our assumption that we inhabit a shared world with others whose perspectives disclose aspects concealed from us and who could confirm our hypotheses about what lies beyond our view). Action is similarly horizonal, defined both by presence and absence—not only by its immanent, embodied engagement with objects, instruments, and other actors but also by its transcendence, its directedness toward goals, purposes, and future events that lie beyond its immediate situation. Actions and perceptions are anticipatory in structure, projected toward states of affairs beyond their immediate horizons, and this virtuality is part and parcel of their “of-ness” and “as-ness” as patterned, intentional gestalts. There can be a circuit between action and perception because they share this as- and fore-structure.
Because actions and perceptions have an intrinsic virtual dimension, they are ready to be rendered into fictions—as-if narrative structures through which the imaginary transfigures and transforms the real. As Wolfgang Iser argues in his monumental and underappreciated late work The Fictive and the Imaginary (1993, 12–13), in fictional narratives “the incorporated ‘real’ world is, so to speak, placed in brackets to indicate that it is not something given but is merely to be understood as if it were given.” This is a process of doubling that “turns the whole of the world organized in the text into an ‘as if’ construction” (13). As Iser explains, “every staging lives on what it is not” (301), and this is true not only of mimetic but also of nonmimetic and even (or especially) antimimetic texts of the sort foregrounded by proponents of “unnatural narrative” (see Alber, Nielsen, and Richardson 2013 and Richardson 2015). The capacity to tell stories and project fictional stagings of ourselves and our worlds is made possible by the virtuality of action and perception because their dual structures of immanence and transcendence can be refigured fictionally into what Ricoeur calls patterns of “split reference” (1977, 265–302)—narrative doublings of absence and presence classically rendered by fairy tales that begin “It was and it was not.” We can double ourselves in stories that stage versions of ourselves because embodied action is already configurative and projective, characterized by both presence and absence, immanence and transcendence.
Fictionality and figuration go hand in hand because they are as-structures. The virtuality of the as if, according to Iser, allows the fictional world to “be taken to figure something other than itself” (19–20). The figurative processes of fictionality and narrative can take different forms across history and in different cultures because the as is variable and can set up any number of relationships. Indeed, Iser claims, “staging in literature makes conceivable the extraordinary plasticity of human beings, who, precisely because they do not seem to have a determinable nature, can expand into an almost unlimited range of culture-bound patternings” (297). As an evolved species with constraints on our cognitive and bodily potentialities that derive from our inherited biological makeup, we are perhaps more limited in how we can configure ourselves than Iser suggests here (see Armstrong 2013, 35–39, Boyd 2009, and Easterlin 2012). Human beings are biocultural hybrids, however, whose nature it is to experiment with and explore different versions of ourselves. This “almost unlimited” plasticity is an anthropological constant across historical and cultural variation and one of the defining characteristics of our species (see Malabou 2008 and Armstrong 2018). Our capacity to emplot actions into narrative configurations is both biologically based and culturally open ended, and this paradox is an important example of our biocultural hybridity.6
This hybridity is also on display in the opening of Portrait of the Artist. The virtual dimension of the actions and perceptions through which Stephen configures his world enables them to stand for meanings other than themselves. The split reference of doubling is evident in the juxtaposition of the time of the story—“and a very good time it was” (2007 [1916], 5)—and the time of Stephen’s hearing of it (which soon turns out to be not such an ideal time after all), and this doubling allows him to take on a role that he also was not (“He was baby tuckoo,” an identity that he can assume only because it is not who he is). Sensations similarly have a virtual dimension that allows them to be doubled and configured into as-relations, with differences in smell signaling mother versus father (supplementing visual differences like the father’s glasses and “hairy face”). Singing and dancing similarly set up family relations (father reciting the song, mother playing the piano, Uncle Charles and Dante clapping) that are biologically based (including age differences: “Charles was older than Dante” [5]) and also culturally charged. The political resonances of the perceptual landscape emerge as Stephen visualizes the colors of Dante’s two brushes and understands their references (maroon for one figure in Irish politics and green for another). The political problems of Ireland that will become crucial for Stephen’s (and Joyce’s) struggles to find an artistic voice are already signaled by his early response to colors as he configures perceptual differences into meaningful relations and imagines what various sensations might stand for. Actions and perceptions can take on these semiotic functions only because they have a virtual dimension and an as- and fore-structure that can be molded into patterns suggestive of aims, purposes, and directions beyond young Stephen’s immediate horizons.
As this scene suggests, the many links between action and perception are crucial for language acquisition. Stephen hears words and reproduces them in song, using rhythmic patterns that are reinforced by music and dance. These intertwinings of action, perception, and language arise from cortical connections that have been documented experimentally. Citing extensive evidence from brain-imaging studies, Friedemann Pulvermüller and Luciano Fadiga (2010, 358) report that “language processing is based on neuronal circuits that reciprocally connect action systems of the brain with perceptual circuits.” As they explain, “Rich links between articulatory and auditory neurons are required to learn the precise mapping between acoustic patterns and the motor programmes necessary for successful word repetition” (352). They point out that “these connections are strong only in humans and are weak in non-human primates,” which may in part “explain why language has not emerged” in those species (352). Speech sounds elicit responses in the motor cortex according to their phonetic qualities, with tongue- and lip-produced phonemes generating measurable reactions in the corresponding topographical areas of the brain that control the tongue and the lips (353). Although the evidence is mixed (as I explain), Pulvermüller and Fadiga note that lesions affecting the “frontal and premotor areas” and interrupting the action-perception circuit have been shown to “compromise the patients’ ability to comprehend meaningful words” (354).
Action, sensation, and perception are richly linked in language comprehension, just as they are in everyday experience. Brain-imaging studies have shown that word recognition is extensively and exquisitely connected to the cortical areas associated with different perceptions of sensation; “words related to odours (for example ‘cinnamon’) activate olfactory brain areas more strongly than do control words,” and “words that are semantically related to sounds (for example ‘telephone’) . . . strongly activate superior temporal auditory areas even if presented in written form” (Pulvermüller and Fadiga 2010, 355). Similarly, words referring to a particular action (for example, kicking or throwing) activate motor areas of the cortex associated with them, so much so that left- and right-handed subjects evince responses in opposite sides of their brains (see Hauk and Pulvermüller 2004, Boulenger et al. 2006, Willems et al. 2010). There is also evidence that tool words, which entail action, produce responses in the motor cortex, whereas animal words do not (Puvermüller and Fadiga 2010, 355).
The linguistic resonances between action words and real actions are part of a more general association between real and imagined action. As Jeannerod (2006, 28, 39) points out, many different experiments have shown that “imagining a movement relies on the same mechanisms as actually performing it,” and that is because “imagined actions are indeed actions in their own right: they involve a kinematic content, they activate motor areas almost to the same extent as executed actions, they involve the autonomic system as if a real action was under way.” Imagining an action stimulates both the brain and the body in a manner very similar to real action, even down to the level of muscular response. For example, Jeannerod notes, measurements of “the changes in . . . the motor pathways during various forms” of imagined action have shown the presence of excitability “in the muscular groups involved in the simulated action”; ample experimental evidence demonstrates that, “during motor imagery, motor commands to muscles are only partially blocked, and . . . motoneurons are close to the firing threshold” (31).
Although imagining an action is not by itself sufficient to master it, many studies have shown that “mental training” by athletes and musicians can improve actual motor performance, enhancing speed, accuracy, consistency, and even “strength of muscular contraction” (41). These findings are consistent with brain-imaging studies that have disclosed that “learning a motor task by using motor imagery induces a pattern of dynamic changes in cortical activation similar to that occurring during physical practice” (41). These results have had important consequences not only for athletic training and observational learning but also for the rehabilitation of stroke patients suffering from motor impairments. Because of the physiological links between real and imagined movement, mental rehearsal of an action can have demonstrable effects on the ability of athletes, musicians, or patients with lesions to actually perform the action.
It consequently stands to reason that reading or listening to a narrative that imitates an action would have practical effects on the brains and bodies of the recipients and on their capacities for acting in the world. If the motor cortex and even muscle tissue can be excited by mental rehearsal of an action, that should also be true of linguistic simulations of actions, and there is experimental evidence that this is so. Lawrence Barsalou (2008, 628) reports, for example, that “when reading about a sport, such as hockey, experts produce motor simulations absent in novices.” This is consistent with an fMRI study by Nicole Speer and her group (2009, 989) that showed correlations between six different kinds of changes represented in stories and the brain regions activated by “analogous activities in the real world” (changes in the location, cause, goal, character, timing, or the object involved in an action). “Different brain regions track different aspects of a story,” Speer and her colleagues conclude, “such as a character’s physical location or current goals,” and “some of these regions mirror those involved when people perform, imagine, or observe similar real-world activities” (989).
There is also evidence that actions and sensations figured in metaphorical language evoke embodied, cortical responses correlated to literal, real-world perceptual experiences. A brain-imaging study conducted by Krish Sathian’s group (Lacey, Stilla, and Sathian 2012, 417, 416) demonstrated that areas of the brain “previously shown to be texture-selective during haptic perception” were also “activated when processing sentences containing textural metaphors” from the same sensory domains. Even very familiar, unsurprising metaphors like “a rough day” or “a slimy person” that “have negative connotations because they refer to attributes that may be particularly unpleasant to touch” evoked responses in the somatosensory cortex associated with these bodily experiences (418). These and many other similar studies point to some of the neurobiological processes set in motion by the feigned figuration of action and sensation in narrative, brain-body interactions in response to imitation that could have the power to reinforce or reshape the recipient’s embodied, configured experience of the world.
Action performs a fundamental role in coordinating different modalities of cognition, and this organizing role is crucial not only for language but also for narrative and our ability to construct and follow plots. The anatomical region of the brain central to these interactions is Broca’s area, a region of the inferior frontal cortex adjacent to the sections of the motor cortex that control the mouth and the lips (see Bear, Connors, and Paradiso 2007, 620–23). According to Pulvermüller and Fadiga (2010, 351), “Studies have shown this area to be active in human action observation, action imagery and language understanding.” Impairments in Broca’s area have long been known to result in difficulties producing and comprehending grammatical sentences. Patients with lesions in this part of the brain can understand and pronounce single words, “but they have great difficulty in aligning scrambled words into a sentence or in understanding complex sentences,” and these deficiencies are “paralleled in non-linguistic modalities” (357). A number of brain-imaging studies have shown, for example, that musical syntax is processed in Broca’s area and that listening to musical rhythms activates the motor cortex (Maess et al. 2001, Chen, Penhune, and Zatorre 2008, Patel 2008).
This region of the brain is also apparently crucial for narrative. An intriguing experiment by Patrik Fazio and his group (2009, 1987) revealed that “a lesion affecting Broca’s area impairs the ability to sequence actions in a task with no explicit linguistic requirements.” His laboratory showed patients with Broca’s aphasia “short movies of human actions or of physical events,” and they were then asked to order, “in a temporal sequence, four pictures taken from each movie and randomly presented on the computer screen” (1980). Curiously, although these patients could still recognize before-after relations between physical events, they had a harder time reconstructing the order of human actions. Their ability to remember and compose a sequence of represented actions was impaired. This result suggests that the patients in Fazio’s study suffered a deficiency in the capacity for emplotment, the ability to produce and follow configurations of action. Such an inference is consistent with Fazio’s claim that “the complex pattern of abilities associated with Broca’s area might have evolved from its premotor function of assembling individual motor acts into goal-directed actions” (1987). This capacity for organizing action into meaningful sequences makes the brain ready for language, but it also prepares the brain for narrative.7
Action has a fundamental coordinating role not only in perception but also in our response to imagery of various kinds. Neuroscientifically trained literary critic Gabrielle Starr (2013, 75) claims that motor imagery and motor processes are involved in the aesthetic pleasures offered by all three of the so-called sister arts. As she explains, embodied action underlies and coordinates imagery and perception in the different sensory domains on which literature, painting, and music draw. After observing that “investigations of imagery” offer “very good evidence that across sensory modes, when people experience perceptual imagery, areas of the brain involved in actual perception are active and function in similar patterns for imagined sensation as during actual perception,” Starr points out that action underlies imagery across different sensory domains and that “imagery of motion” is “at the heart of our capacities for both simulation and aesthetic experience” (75, 81). For example, visual imagery typically has an implicit proprioceptive dimension, Starr observes, because “people imagine motion along with sight—the experience of moving one’s body is, for sighted persons, both visual and kinetic” (82). Auditory imagery similarly has fundamental connections to movement because the perception of sound depends on patterns of repetition and variation. For imagery related to music and sound, Starr observes, action matters because “timing is everything and the tension between what is new and what is expected is ever present” (128). Multisensory imagery that invokes and interrelates all of these different perceptual areas can associate vision, sound, touch, and even smell, she argues, because of their bases in embodied action. If narrative as the emplotment of action similarly has the power to coordinate processes of configuration in different sensory domains, that is due in no small measure to the underlying organizing role of motor processes in perception.
Following a story, whether by listening or by reading, is not passive absorption or one-to-one mirroring. Here too action is fundamental to cognition. Making sense of a narrative involves an interaction between different kinds of action—the organization of story elements in patterns of discordant concordance (mimesis2) and the work of pattern formation, gap filling, and illusion building in the process of reception (mimesis3). The comprehension of a story requires active participation by the recipient, who must project relations between the parts that are told and their probable configuration in the whole that seems to be forming—this entails a to-and-fro work of anticipation and retrospection through which illusions are built and then broken, consistent patterns formed and then disturbed, disrupted, and reassembled as expectations are invoked, disappointed, modified, and fulfilled.
Reciprocal, to-and-fro pattern formation is involved in all cognitive processes, whether visual, auditory, or haptic, and across the multiple sensory domains that are bound together in embodied experiences of cognition. The construction of narrative, story comprehension, and embodied cognition are mutually formative because they all entail to-and-fro, reciprocal processes of configurative activity. The activity of construing a text or understanding a story has the power to change the way our brains and bodies respond to the world because the processes of configuration across these domains are not separate but congruent, mutually formative, and reciprocally interactive. This is why the interchange between the emplotment of action in narrative (mimesis2) and the recognition and reconstruction of a plot in story comprehension (mimesis3) can interact with the action-perception circuit in cognition (mimesis1) in a potentially transformative manner.
These interactions play an important role in the brain’s incessant balancing act between pattern and change, the tension between the need for order and the requirement for flexibility and openness to variation that is fundamental to successful mental functioning. Repeated patterns of activity form into habits, and the process of making and breaking habits is crucial to emplotment, story comprehension, and embodied cognition. It is a widely acknowledged paradox of cognitive life that habits are both good and bad, and necessarily so—habits are indispensable to skillful coping and efficient functioning but at the same time also an obstacle to innovation and adaptation to novelty, change, and unexpected irregularities (James 1890, 1:104–27; Noë 2009, 99–128). Habit formation is basic to cognitive life, even at the neuronal level, as Hebb’s law (2002 [1949]) famously observes: “Neurons that fire together, wire together.” The capacity of repeated neuronal activity to establish enduring cortical connections prevents the recursive interactions of “reentry” (Edelman and Tononi 2000) from being purely random, chaotic, and accidental, and this in turn allows learning to occur. The usefulness of habit is similarly evident in the conventions of storytelling and the repetitions of narrative patterns that facilitate the construction and recognition of plots.
But a brain locked forever in repeated patterns of neuronal assembly would lack the capacity to respond to changing stimuli, to explore an unpredictable environment, or to try out new cortical configurations that might eventually develop into new cognitive capacities (the joining up of the motor cortex to auditory, visual, and sensory areas of the brain, for example, that occurred as our species developed the capacity for language and reading). Similarly, although there is pleasure in recognizing a familiar story, even the retelling of a well-known tale typically varies the received patterns it deploys in order to keep the listener interested. Excessive homogeneity is once again bad for the brain. If the value of concord in repeated stories is to reinforce and teach habitual, socially shared cognitive patterns, the value of disruption, irregularity, and surprise in the concordant discordance of emplotment is to preserve and enhance cognitive flexibility, our openness to change, variation, and novelty. By invoking and disrupting the habits through which we pattern the world, the experience of telling and following stories has the power to bring about potentially transformative exchanges between different modalities of figuration. The neurobiological source of this power is the coordinated interaction between far-flung areas of the brain and the body that join together action, perception, and cognition.
The many experimental findings on the links between cognition and embodied experiences of perception, sensation, and action have given support to the theory of grounded cognition most notably propounded by Barsalou. This theory rejects the view that “knowledge is represented in abstract codes, distinct from the sensory modalities through which the knowledge was acquired” (Lacey, Stilla, and Sathian 2012, 416). “The idea that the conceptual system draws on sensory and motor systems,” Rutvik Desai and company note, “has received considerable experimental support in recent years” (2013, 862). And as Michael L. Anderson (2010, 257) observes, “It is quite easy to find studies reporting that the neural implementations of higher cognitive functions overlap with those of the sensorimotor system”; as proof he points to the twenty-seven-essay volume Sensorimotor Foundations of Higher Cognition (Haggard, Rossetti, and Kawato 2008).
So overwhelming is this evidence that the debate has recently turned and the question has become, how do we think abstractly? Warning against “some of the hyperbolic rhetoric used by supporters of embodied cognition,” Guy Dove (2011, 3, 5) argues that “what we need to explain is our ability to go beyond embodied experience.” As he observes, “Sensorimotor simulations seem ill-suited for representing conceptual content that is not closely tied to particular experience,” and “some concepts appear to require what we might call ungrounded representations” (5). Similarly worried that “a quick acceptance of embodied accounts runs the danger of ignoring alternate hypotheses and not scrutinizing neuroscience data critically,” Anjan Chatterjee (2010, 79) suggests that “the question of whether or not cognition is grounded is more fruitfully replaced by questions about gradations in this grounding. A focus on disembodying cognition, or on graded grounding, opens the way to think about how humans abstract.” How narratives stage figurative versions of experience similarly requires an understanding of the relation between the abstract and the concrete. We can tell each other stories that reconfigure embodied experience and transform it into narrative patterns because action and perception have a virtual dimension and are both grounded and ungrounded, embodied and disembodied.
In ways that are not always recognized and understood, this duality is already evident in the key term “simulation” in Barsalou’s theory: “Grounded cognition proposes that modal simulations, bodily states, and situated action underlie cognition” (2008, 617). According to Barsalou, “Simulation is the reenactment of perceptual, motor, and introspective states acquired during experience with the world, body, and mind” (618). Simulation is often a bit of a black box in cognitive and psychological theories, however (see Mumper and Gerrig 2019), and it is frequently invoked in ways that misleadingly imply causation or one-to-one correspondences. To a literary critic or theorist, however, this term is almost self-evidently an as-structure.
By definition, after all, a reenactment both is and is not what it reproduces. In terms familiar to literary theory, the equivalences established in a simulation are of the paradoxical Nietzschean variety—“das Gleichsetzen des Nicht-Gleichen,” or “the setting equal” of things that are not the same (see Nietzsche 2015 [1873]). In a simulation, an as-if rendering reproduces a state of affairs as something that it is not, and this entails a doubling of the like and the not like (Gleich and Nicht-Gleich). The as and the as-if enable conceptual and narrative reenactments of perceptual and sensory experiences in simulations that both are and are not what they reproduce. The virtuality of the as and the as-if also makes possible graded transitions from the concrete to the abstract.
Barsalou’s explanations of simulation again and again presume its as-structure without ever naming it as such or fully recognizing the complications it introduces to the model of grounded cognition. Barsalou acknowledges, for example, that “simulations rarely, if ever, recreate full experiences” but instead are “typically partial recreations of experience that can contain bias and error” (620). Because cognitive simulations are partial and incomplete, they necessarily entail what he calls “a pattern-completion inference mechanism” (622). As the very term “mechanism” itself implies, sometimes Barsalou unfortunately uses causal language to explain how this process works. For example, he states that “bodily states are not simply effects of social cognition; they also cause it” (630). An inference is not a causal mechanism, however, but a hypothesis about a pattern and a guess about how an incomplete perspective should be filled out.8 Simulations entail variable, open-ended processes of pattern formation and gap filling because they are not straightforward, mechanical productions based on cause and effect but as-relations that make connections based on inference and association.
Simulations are products of selection and combination that can use the materials they draw on in a variety of ways. As partial re-creations, they do not retain all of the original experience, only traces of it or some aspects rather than others. These selections can then be combined differently as new situations arise, which is why past sensorimotor experiences can be used to configure new cognitive situations. Grounded cognition can only respond to unprecedented and unpredictable states of affairs because its simulations are not completely identical to the experiences these simulations draw on. New perceptions can be “grounded” in past cognitive experiences because of this variability—because, that is, they are not mechanical, one-to-one repetitions of previous sensations but as-relations, both like and not like the experiences they reenact, partial and incomplete in ways that make possible bias and error but that also enable novelty and creativity. The processes of selection and recombination in a cognitive simulation are much more variable and open ended than the language of mechanism and cause suggests, and that is a good thing because it makes cognition more flexible and more open to new and unexpected situations than it would be if grounding were seamless and ungraded. The paradox of grounded cognition is that its simulations work only because they are to some extent ungrounded.
The interpretive activity of figurative pattern construction and gap filling is evident in many of Barsalou’s examples of how inference and simulation work even when he invokes a mechanical model. Here is a typical instance: “Representations of familiar situations that contain embodiments become established in memory (e.g., receiving a gift, feeling positive affect, and smiling). When part of this situation occurs (e.g., receiving a gift), it activates the remainder of the situational pattern, producing associated embodiments (e.g., smiling)” (630). This is a classic example of what Richard Gerrig (2010, 2012) calls “memory-based processing,” which is evidence of the brain’s recursive work of constructing cognitive patterns of seeing-as—here, seeing a new situation as associated with learned conventions from previous interactions. Traces of those associations may be embodied, but this is not a case where a bodily state simply causes a social cognition.
Not a cause-effect mechanism, this is an interpretive situation in which signs must be read as standing for something that must be construed. As Umberto Eco (1976, 6–7) memorably notes, one can tell signs are at work whenever there is the possibility of lying. Here, it is possible to feign a polite smile at a gift one hadn’t really wanted—or that one knew was well intentioned but misfired—and the neurobiological basis of this possibility is that the reactivations of a simulation are partial and can be configured in different ways. For example, one can smile with a wink and a nod that suggests to someone in the know an ironic response (“Thanks a lot!” can be intoned in a lot of different ways). It is not always possible to tell whether someone is smiling sincerely or simply politely because they are following a social convention, and the as-if of simulation is responsible for this ambiguity. A polite smile is a biocultural hybrid, a bodily response and a conventional expression, and this duality is possible because simulations are both grounded and ungrounded.
A process of doubling characterizes the as-if of simulation because it is a reenactment of an experience that both is and is not like the original it reinvokes. The role of the as-if is evident, for example, in a prominent article in Science on “Embodying Emotion” by Paula Niedenthal (2007) which is based on Barsalou’s model. Her account is couched in the language of mechanical causation, but her terminology interestingly and subtly shifts in ways that reveal the variability of simulative as-relations: “The embodiment of emotion, when induced in human participants by manipulations of facial expression and posture in the laboratory, causally affects how emotional information is processed. Congruence between the recipient’s bodily expression of emotion and the sender’s emotional tone of language, for instance, facilitates comprehension of the communication, whereas incongruence can impair comprehension” (1002, emphasis added). The slippage between “causally affects” and “facilitates” and “can impair” is telling, moving from billiard-ball mechanical certainty to the language of probability and contingency. The result may occur—but maybe not—because a simulation is both like and not like the original experience it reenacts.
In the reported experiment, subjects who were made to smile by holding a pen in their teeth (thus simulating smiling) rated cartoons as funnier than subjects holding a pen “between their lips,” which “prevented them from smiling” (1005). Niedenthal explains that “using knowledge—as in recalling memories, drawing inferences, and making plans—is thus called ‘embodied’ because an admittedly incomplete but cognitively useful reexperience is produced in the originally implicated sensory-motor systems, as if the individual were there in the very situation, the very emotional state, or with the very object of thought” (1003; emphasis added). Note the as-if quality of this reexperiencing, which is incomplete in that only a subset of the neurons fire that were originally active, setting in motion only partial traces of the original physiological processes in the sensory-motor system that is reactivated by the simulation. The placement of the pen (between the teeth or the lips) recreates an embodied experience as if one were smiling or frowning. This simulation (which both is and is not like actual smiling or frowning) has associations (happy or sad, funny or not funny) that connect and fill out what is missing (how to evaluate the comedy of the cartoon, itself a blank not contained in any of its features, which is why it can be evaluated differently whether one is smiling or frowning).
This response is not a causal mechanism but an inference that completes a pattern and fills in a blank. This experiment illustrates the fore-structure of grounded cognition, in line with Heidegger’s claim (1962 [1927], 172–88) that mood (Stimmung) provides an anticipatory understanding of a state of affairs.9 A simulated smile or frown can prime a response, making one more or less likely to perceive something as funny, because it orients the perceiver’s attitude toward a situation. As an as- and fore-structure, an embodied simulation is a hermeneutic figuration of something as something depending on how we are directed toward its various unspecified horizons (patterns that our inferences complete). Grounded cognitions are structures of intentionality, not deterministic processes.
Similar doublings and ambiguities are evident in well-known disputes about the relation between embodied experience and our ability to understand another’s actions and gestures. The basis of these disputes is the much-discussed discovery by Giacomo Rizzolatti and his Parma group of “mirror neurons” in the motor cortex of macaque monkeys that fired not only when the animals performed an action but also when they observed the same action or even an object (like a cup) associated with the action.10 The existence of mirror neurons, also subsequently found in humans, seemed to provide a straightforward neurobiological explanation for the grounding of cognition on embodied experiences and the relation between action and perception. But this explanation soon raised further questions. Mirror-neuron skeptic Gregory Hickok (2014, 43, 49) observes, for example, that “we don’t need to be able to perform an action to understand it,” and he also notes that “deficits in the control of action [because of lesions or physical impairments] do not uniformly result in deficits in understanding of action.” These counterexamples cast doubt on the claim that we understand an action by reenacting (simulating) its performance.
The clinical and experimental evidence about the relation between performance deficits and cognitive ability is indeed mixed. On the one hand, some studies have shown that lesions in the motor and somatosensory cortices “can lead to the loss of multiple categories that share perceptual properties”; for example, “a selective deficit in action word processing has been found in patients with motor neuron disease” (Dove 2011, 3; see Bak et al. 2001). On the other hand, Gilles Vannuscorps and Alfonso Caramazza (2016, 86–87) have demonstrated that “individuals born without upper limbs can perceive, anticipate, predict, comprehend, and memorize observed actions with the same accuracy and speed as the controls,” and they consequently conclude that “efficient perception and interpretation of actions can be achieved without motor simulation.” Jeannerod (2006, 16) summarizes this ambiguous state of affairs: although apraxic patients who cannot perform a particular action “fail in action representation tasks,” they “usually remain unimpaired in tasks evaluating their conceptual knowledge about objects or tools. . . . This dissociation between ‘actions’ and ‘objects’ stresses the limitation of theories based on motor simulation to explain conceptual knowledge.” Despite his skepticism about mirror neurons, Hickok (2014, 153) offers a sensible compromise view: “We should not confuse the fact that action is an important part of perception with the idea that motor representations alone are the basis of perceptual understanding.” As he explains, “This is not to say that the ability to act in the world doesn’t make some contribution to perceiving and therefore understanding. On the contrary, action contributes a great deal” (152; original emphasis)—but it is not the whole story because cognition is both grounded and ungrounded, embodied and disembodied.11
These contradictions are based on a fundamental paradox of cortical functioning. Cognition is both anatomically localized and a function of network connections across the “brain web.” On the one hand, some functions depend on specific parts of the brain that are active during a particular cognitive experience, and these abilities therefore get knocked out when these areas are damaged (as when, for example, impairments in various regions of the rear visual cortex may cause particular kinds of blindness to motion, colors, or faces). On the other hand, many cognitive functions set in motion reciprocal, cross-cortical interactions between different anatomical locations and assemble neurons into particular patterns that may even change what a specific area does (so that the visual cortex of a blind person can be repurposed to support touch in reading braille). Some localized somatosensory and motor areas would consequently be expected to fire in response to words, cognitive experiences, or even imaginings that trigger them (as happens, for example, when action words set off particular parts of the motor cortex). But other cognitive functions depend on network interactions that are not reducible to any anatomical location (as when we understand actions we have not experienced or that our bodies cannot perform). Someone who has not had a particular experience might not be able to understand it as completely as someone whose brain and body would be able to simulate and reenact a somatosensory, embodied cognitive act (recall that novices do not register reactions to the steps of ballet or capoeira that the brains of experts resonate to), but by combining various cortical areas across the brain web, we can understand actions outside our particular motor repertoire (and appreciate a dance performance of which we are ourselves incapable).
Cognition is not simply a linear, cause-effect, bottom-up mechanism of stimulus and response. It is, rather, a reciprocal, to-and-fro process of assembling cortical regions into patterns that may stabilize through repeated Hebbian firing but that can also shift and change and that can, in a top-down manner, mold and shape our understanding of what is “out there” in the world. As Hickok (2014, 238) observes, “There are more top-down neural connections (from higher to lower levels) than bottom-up connections (from lower to higher levels), about an order of magnitude difference (10:1 ratio).” This is anatomical evidence of the ability of network interactions to produce cognitive experiences that go beyond those generated by receptor neurons in local somatosensory regions. Noting that “action understanding is the interaction of many things,” Hickok consequently proposes “a hybrid model of conceptual knowledge” that is based on network connections between motoric and somatosensory areas (180, 168; see 159–81).
Anderson (2010, 246) found similar empirical evidence for a network model: “An empirical review of 1,469 subtraction-based fMRI experiments in eleven task domains,” including “action execution, action inhibition, action observation, vision, audition, attention, emotion, language, mathematics, memory, and reasoning,” showed that “a typical cortical region is activated by tasks in nine different domains.” What a particular anatomical region of the brain is doing depends not only on its own physiological characteristics, then, but also on the areas it is linked with. Anderson also found that “language tasks activate more and more broadly scattered regions than do visual perception and attention. This finding was corroborated by a larger study . . . which found that language was the most widely scattered domain of those tested, followed (in descending order) by reasoning, memory, emotion, mental imagery, visual perception, action, and attention” (247). These findings support Anderson’s theory of neural reuse, which emphasizes the plasticity of the brain web and the ability of interactions between cortical regions to repurpose particular areas (as when, in an example he cites, an area of the visual cortex devoted to invariant object recognition is “recycled” into what Stanislas Dehaene [2009] calls a “visual word form area” that supports reading). As Anderson explains, “Modularity advocates are guided by an idealization of functional structure that is significantly at odds with the actual nature of the system. Instead of the decompose-and-localize approach to cognitive science that is advocated and exemplified by most modular accounts of the brain, neural reuse encourages ‘network thinking’ ” (249). A dual model of cognition as both anatomically localized and a function of network connectivity has become widely accepted by the neuroscientific community.12
A combination of localization and network effects is necessary to explain the simulations of grounded cognition as well as the openness of grounding to various gradations. Simulations are partial, incomplete reactivations of particular cortical areas as well as reenactments of these experiences in new configurations. They are selections as well as combinations—partial traces of the original activity on which they are based, without the same fullness or intensity, that are then reassembled with different cortical areas in patterns that generate new inferences and that complete and fill out what is missing according to the demands of the new cognitive situation. This is the neurobiological basis of Barsalou’s pattern-completion inference mechanism.
Because simulations are as-if re-creations that differ in the degree to which they activate a specific cortical area, they can be variously concrete and abstract. These degrees of grounding also depend on the interactions a simulation sets in motion across the brain web. Traces of sensory reactivation could be expected to occur in the most far-flung assemblies, as when specific motoric responses to action words are part of widely diverse, cross-cortical linguistic patterns (somatotopic reactions in the motor cortex to the words “kick” or “throw” in a narrative about football, for example), and these selections and combinations may result in different gradations of abstraction in the simulations they construct. Because the brain is a paradoxical web of local, anatomically based capacities and network connections that can activate and even repurpose particular areas in different ways, it can support simulations that are variously concrete and abstract, with gradations in how grounded they are in specific motoric and somatosensory regions and in how much these local reactions are transcended and transformed by the effects of cross-cortical interactions.
The paradoxes of simulation and graded grounding help to explain some of the contradictions and controversies related to the theory of embodied metaphor. As Raymond Gibbs and Teenie Matlock (2008, 162) explain, the theory that “many abstract concepts . . . are understood, at least partly, in embodied metaphorical terms” is based on simulation: “The recruitment of embodied metaphors . . . is done imaginatively as people re-create what it must be like to engage in similar actions. The key to this imaginative process is simulation, in this case the mental enactment of the very action referred to in the metaphor.” As suggested by the title of George Lakoff and Mark Johnson’s seminal text Metaphors We Live By (1980, 5), embodied metaphors like “affection is warmth,” “important is big,” or “happy is up” are as- and fore-structures that “understand . . . one kind of thing in terms of another.” We “live by” these metaphors because they guide our engagements with the world according to the various patterns they project. Lakoff and Johnson’s later formulation in Philosophy in the Flesh (1999, 47) of metaphors as “experientially grounded mappings” similarly implies both the as and the fore, inasmuch as a map establishes correlations and orients our expectations about pattern.
Rather than conceiving the relation between signifier and signified as purely arbitrary and conventional, as in Saussurean theory, proponents of embodied metaphor claim that our experiences of the world, constrained by our bodily-based abilities and dispositions, structure many, if not most (and the slippage here is where the controversy starts), of the relationships of similarity and difference through which we categorize states of affairs and (accordingly) reason, perceive, and act. As Lakoff and Johnson explain:
Experiencing the More Is Up correlation over and over should lead to the establishment of connections between those neural networks in the brain characterizing More in the domain of quantity and those networks characterizing Up in the domain of verticality. In the model, such neural connections would carry out the function of a conceptual mapping between More and Up and make it possible (though not necessary) for the words of verticality (such as rise, fall, skyrocket, plummet, high, low, dip, and peak) to be used conventionally to indicate quantity as well. (1999, 54)
It is not clear that there are specific regions or networks in the brain that encode More or Up or quantity or verticality as such, but otherwise the key claim here—that repeated correlations that “arise out of our embodied functioning in the world” establish “neural connections” that then become reactivated in simulations in new situations (54, 46)—matches up well with core principles of neuroscience. This is, after all, classic Hebbian wiring and firing.
As with simulations in general, however, a key question is how fixed or variable the connections established by the as-if might be. Here Lakoff and Johnson describe them as “possible (though not necessary),” although elsewhere they waffle between claiming that “the architecture of your brain’s neural networks determines what concepts you have and hence the kind of reasoning you can do” and asserting merely that “the body and brain shape reason,” or even less strictly that “our sense of what is real begins with and depends crucially upon our bodies” (16–17; emphasis added). These slippages are not surprising, however, because they reflect the contingencies of the as-structures that make possible gradations in grounding and variabilities in the concreteness and abstractness of our categories.
The experimental evidence on embodied metaphors is mixed, and its contradictions reflect these contingencies. Vittorio Gallese, a member of the Parma mirror-neuron group, joined up with Lakoff (2005, 456–57) to thresh out the connections between the theories of action understanding and embodied metaphor, based on their shared assumption that “imagining is a form of simulation—a mental simulation of action or perception, using many of the same neurons as actually acting or perceiving”:
For example, in the case of the concept of grasping, one would expect the parietal-premotor circuits that form functional clusters for grasping to be active not only when actually grasping, but also when understanding sentences involving the concept of grasping. . . . A further prediction of our theory of concepts is that such results should be obtained in fMRI studies, not only with literal sentences, but also with the corresponding metaphorical sentences. Thus, the sentence He grasped the idea should activate the sensory-motor grasping-related regions of the brain. (472)
Interestingly, in fact, subsequent brain-scanning studies have not uniformly born out this prediction. Although some fMRI experiments have found sensory-motor resonances to nonliteral uses of action words (grasping an idea as opposed to grasping an object), others have not.
For example, a study by Shirley-Ann Rüschemeyer and her group (2007) comparing brain scans of responses to German action verbs like “grasp” (“greifen”) and conceptual verbs based on them (“begreifen,” “to understand”) found premotor activation only for the action verb and not its conceptual derivative. This finding is in line with other experimental results that have shown responses in the premotor cortex to literal but not nonliteral uses of action verbs (for a summary, see Willems and Casasanto 2011, 7–8). In one case, an fMRI study by Lisa Aziz-Zadeh and associates in the Parma group (2006) found activation in the corresponding section of the premotor cortex in response to phrases describing literal hand, foot, and mouth actions like “grasping the scissors,” “pressing the piano pedal,” or “biting the peach” but not to nonliteral phrases that used action words in a conceptual or metaphorical manner (“handling the truth,” “chewing over the details,” “kicking off the year”). Similarly, a brain-scanning study by Ana Raposo’s group (2009) documented motor cortex responses to individual action verbs like “kick” or action sentences like “kick the ball,” but not to idiomatic phrases like “kick the bucket.”
Still other experimental evidence has documented a gradual decline in sensory-motor activation along a continuum from literal to nonliteral usages, with metaphors sometimes eliciting responses, but less so as they become more abstract and conventional. For example, a study conducted by Cristina Cacciari and colleagues (2011, 149) showed greater “excitability of the motor system” when experimental subjects “were presented with literal, fictive, and metaphorical motion sentences than with idiomatic motion or mental sentences.” Her experiment found gradations in somatosensory and motor response to linguistic simulations depending on the degree of the involvement of the original experience in the simulation. As she explains, “These results suggest that the excitability of the motor system is modulated by the motor component of the verb, which is preserved in fictive and metaphorical motion sentences” (149), but not in idiomatic expressions that have submerged the original action in a conventional expression or in an abstraction (like “begreifen” for grasping a concept [“Begriff”]) at a distance from the concrete motor act. An fMRI study by Rutvik Desai’s laboratory (2013, 862) similarly “showed involvement of sensory-motor areas for literal and metaphoric action sentences, but not for idiomatic ones. A trend of increasing sensory-motor activation from abstract to idiomatic to metaphoric to literal sentences was seen,” which suggests “a gradual abstraction process whereby the reliance on sensory-motor systems is reduced as the abstractness of meaning as well as conventionalization is increased.”
These experiments do not support Gallese’s strong prediction about the embodied grounding of nonliteral usages (literal versus conceptual grasping), but they are consistent with a graded view of grounding. These kinds of variations would indeed be predicted by a model of simulation as an as-if process of figuring relationships that can vary in their degree of abstractness and embodiment. According to such a model, as conventional associations take over and metaphors once alive become dead, traces of the bodily origins reenacted by a simulation would become less and less visible, dulled by repeated use in routine expressions because of the well-known desensitizing effects of habituation. The sensory-motor aspects of the simulation would consequently become less pronounced as the equivalences established by the as-relation become more customary and the abstract semantic elements come to the fore. Only an odd, unexpected use of the term—a defamiliarizing gesture that de- and recontextualizes it (as is typical, for example, in poetry)—might revivify those sensory-motor associations and reactivate the parts of the simulation that have gone dormant (see Armstrong 1990, 67–88). These variations are possible, however, only because a simulation is an as-relation that may activate local sensory-motor resonances in different and shifting degrees. Nonliteral usages may seem more or less abstract as they are more or less grounded in perceptual experiences, more or less a product of long-distance, cross-cortical network connections, more or less habitualized through repetition and conventionalization.
Narratives typically deploy metaphors to help organize the action they set in motion, and these figures similarly range along different gradations of grounding, from the concrete to the abstract. In a well-known passage from Père Goriot, Balzac’s classic novel in the realist tradition, the proprietress Madame Vauquer figuratively embodies her boardinghouse:
The fat old face . . . , with the nose jutting from the middle of it like a parrot’s beak; the podgy little hands; the body, plump as a churchgoer’s; the flabby, uncontrollable bust—they are all of a piece with the reeking misery of the room, where all hope and eagerness have been extinguished and whose stifling, fetid air she alone can breathe without being sickened. Her face, with the bite of the first autumn frost in it; the wrinkles round her eyes; the expression in them that can quickly change from the set smile of a ballet dancer to the embittered scowl of a bill discounter—her whole person, in fact, explains the house, as the house implies her person. The prison cannot be run without a jailer; you can’t imagine one without the other. The unwholesome corpulence of this little woman is the product of this life, as typhus is the product of a hospital. . . . When she is there, the exhibition is complete. (2004 [1834], 10–11)
As Erich Auerbach (2003 [1953], 470–71) explains in his famous analysis of this scene, “the motif of the harmony between Madame Vauquer’s person on the one hand and the room in which she is present, the pension which she directs, and the life which she leads . . . is not established rationally but is presented as a striking and immediately apprehended state of things, purely suggestively, without any proof.” Vividly embodied in Madame Vauquer’s physical presence, the as-relations in this simulation seem so concrete and immediate as to be irrefutable even though they are purely figurative and metaphorical. The equivalences established by this configuration are offered as an intuitive demonstration of their truth, even as their hyperbolic strangeness and lingering discordance (how might one combine a ballet dancer, a bill discounter, a jailer, and typhus?) remind us of their rhetorical constructedness (a paradox consistent with Balzac’s combination of melodramatic exaggeration and historical verisimilitude). The figurative coherence proposed by these embodied equivalences grounds this simulation in the concrete.
By contrast, the artificiality and even peculiarity of the equivalences at play in Henry James’s typically elaborate metaphors are graded toward the abstract, so much so that many readers have complained that they are awkward and difficult to visualize. The best-known, most frequently commented example is perhaps Maggie Verver’s invocation of a pagoda as a figure for her dawning intuition at the beginning of the second volume of The Golden Bowl that all might not be well in the arrangement that has thrust her husband and her best friend (and unbeknownst to her, his former lover) so much into each other’s company while she and her father (now wed to this woman) cultivate the intimacy they enjoyed before their marriages:
This situation had been occupying for months and months the very centre of the garden of her life, but it had reared itself there like some strange tall tower of ivory, or perhaps rather some wonderful beautiful but outlandish pagoda, a structure plated with hard bright porcelain, coloured and figured and adorned at the overhanging eaves with silver bells that tinkled ever so charmingly when stirred by chance airs. She had walked round and round it—that was what she felt; she had carried on her existence in the space left her for circulation, a space that sometimes seemed ample and sometimes narrow: looking up all the while at the fair structure that spread itself so amply and rose so high, but never quite making out as yet where she might have entered had she wished. (2009 [1904], 327)
Concrete and embodied, this figure simulates Maggie’s first faltering attempt to understand bewildering signs by rendering it as an exploratory stroll around a strange construction. At first glance the foursome seems not at all like a pagoda, however, and it is not surprising that many readers have found it a strained, unsuccessful figure.13 This incongruity is nevertheless epistemologically appropriate because it is a sign of Maggie’s groping effort to find new patterns of like and not like to make sense of her world. The very complication and ornateness of the pagoda image as James extends and elaborates it over several pages dramatize Maggie’s inability to make her world cohere—her struggle to fit its parts into a consistent whole inasmuch their formerly familiar arrangement now seems incongruous and strange. Maggie’s stroll around the garden is a concrete, embodied simulation whose disjunctions and abstractness call attention to the cognitive processes whereby she attempts to understand her world.
These discrepancies and complications also foreground the way in which metaphors construct patterns of relationship by joining the like and the not like. The very ornateness of the metaphor displays and plays with the as-if quality of a figurative simulation as an experiment with relations of equivalence (its as-structure). Rather than trying to submerge and naturalize these functions in an illusion of verisimilitude, James foregrounds and dramatizes them by elaborating the pagoda image with almost baroque complication. This unnatural, deliberately artificial display of the workings of figuration playfully exposes the as-relations set in motion by an as-if simulation instead of seeking, as Balzac does, to cover them up with a cloak of intuitive immediacy.
A further gradation in grounding, moving from the concrete to the abstract, characterizes the various figurations of embodiment in that most unnatural of narratives, James Joyce’s Finnegans Wake. The most famous chapter in the novel, where washerwomen gossip about Anna Livia Plurabelle (ALP), is replete with bodily actions and bawdy figures, but the overall effect of Joyce’s linguistic play is to foreground and call attention to the infinitely variable contingencies of “as if” simulations. Imagine the women talking (or Joyce reciting their speech in his Irish lilt in the remarkable recording he made of this section):
tell me all about
Anna Livia! I want to hear all
about Anna Livia. Well, you know Anna Livia? Yes, of course, we all know Anna Livia. Tell me all. Tell me now. You’ll die when you hear. Well, you know, when the old cheb went futt and did what you know. Yes, I know, go on. Wash quit and don’t be dabbling. Tuck up your sleeves and loosen your talk-tapes. And don’t butt me—hike!—when you bend. Or whatever it was they threed to make out he thried to two in the Fiendish Park. He’s an awful old reppe. Look at the shirt of him! Look at the dirt of it! He has all my water black on me. (2012 [1939], 196)
These lines are very physical and corporeal—evocative back-and-forth exchanges spoken by women at work washing clothes by the riverside—but abstract verbal play (not so visible when Joyce recites these lines) begins with their inaugural typographical games, with the enunciated “O” also signifying the dominance of the feminine in the chapter, an allegorical resonance reinforced by the inverted triangle it sits atop, which in turn is not only another reference to the female anatomy but also the sigla (or hieroglyph) that Joyce used during composition to identify ALP. A pun on the French word “eau,” this letter also announces the chapter’s preoccupation with water, a theme suggestive again of the feminine (recalling Molly’s flowing monologue in Ulysses), and notoriously echoed by hundreds of indirect references to rivers throughout the chapter (Roland McHugh [1991] finds allusions in this passage to the river Repe in Germany and the Blackwater River in Florida).14
This is self-consciously both speech and writing, and that doubling is reiterated and reinforced in the further as-if games Joyce plays in this elaborate simulation. Even as the women talk and wash, their acts suggest the sexual behavior (“don’t butt me”) that they avidly discuss, but these allusions come through portmanteau words and puns that are products of the writer’s games rather than their actual speech—“Fiendish Park” for Phoenix Park, for example, where ALP’s husband, HCE (the “chap” indicated by his initials in “cheb”), is said to have “tried” to “do” whatever his notorious but ambiguous crime was. The suggestive but indeterminate word-play here suggests yet another triangle (“threed,” “thried,” “two”) where HCE was seen by a third person trying to do some evil deed to a second—a sexual act of some kind? The grime that befouls their washing water figures the foulness of his sin as well as the dirt with which they enjoy besmirching their supposed friend ALP. The physical rhythms of their speech give the language a concreteness that the abstract allegorical references and the playful puns of the writing counter with a lively virtuality. This juxtaposition in turn has a paradoxical effect, producing at one and the same time a very physical, natural, and bodily scene and a very artificial, abstract, and symbolic set of games, a duality that once again foregrounds the work of the as in as-if simulations.
The paradoxes and contradictions of graded grounding and as-if simulation are evident in the conflict over whether bodily-based metaphors are universal or culturally and historically variable. “When the embodied experiences in the world are universal,” Lakoff and Johnson argue (1999, 56), “then the corresponding primary metaphors [those that derive from basic sensorimotor experiences] are universally acquired. This explains the widespread occurrence around the world of a great many primary metaphors.” For example, a study by Benjamin Wilkowski and colleagues (2009) demonstrated that the metaphor of heat for anger is based on bodily experiences that transcend cultural differences. As they explain, “many parallels between heat and anger provide an embodied basis for anger-heat associations. . . . For example, just as boiling fluids expand and threaten to escape their container, anger threatens to erupt in the form of an aggressive ‘explosion’ ” (475). Wilkowski’s study concludes: “These conceptual parallels between anger and heat are embedded in the embodied experience of human existence and know no cultural boundaries” and, consequently, “are responsible for the robust cross-cultural belief that anger and bodily heat are systematically related” (475). These claims are supported by a series of experiments involving 438 participants that tested in various ways their tendency to associate anger and heat. Among the reported results: “visual depictions of heat facilitated the use of anger-related conceptual knowledge” (464), “priming anger-related thoughts led participants to judge unfamiliar cities and the actual room temperature as hotter in nature” (469), “participants were faster to categorize an angry facial expression when it was presented in a context suggestive of heat” (469), and “cues to heat led individuals to identify a greater degree of anger in blended emotional expressions” (475).
By contrast, however, a wide-ranging study by historian Joanna Bourke (2014, 484) provides compelling evidence that “the physiological body is not a culture-free object.” As she observes, “Even a cursory look at the world’s languages reveals a formidable number of non-universal metaphors” for even as fundamental an experience as pain—so much so that “the McGill Pain Questionnaire (an extensive list of pain-descriptors that was developed in America in the 1960s) could not always be translated straightforwardly into other European languages” (482). These discrepancies are even more acute, she points out, with non-European examples: “The Sakhalin Ainu of Japan complain of ‘bear headaches’ that resemble the heavy steps of a bear; ‘musk deer headaches,’ like the lighter galloping of running deer; and ‘woodpecker headaches,’ as if pounding into the bark of a tree” (483). Similarly, comparing modern medical notions of homeostasis to eighteenth-century humoral theory that figured pain through “rich figurative languages of ebbs and flows,” Bourke argues that what the opposition of balance versus imbalance means in these different systems is not identical: “Of course, the most basic schema survive: ones based on sensorimotor bodily experience of lying down and standing up (HAPPY IS UP and SAD IS DOWN), for instance, or those based on physical comportment relating to vigour versus infirmity (HEALTH/LIFE ARE UP and ILLNESS/DEATH ARE DOWN). But others take on such a different meaning in the humoural scheme of physiology [as] to be radically different” (486)—disequilibria or blockages, for example, involving “phlegm, black bile, yellow bile, and blood” that are connected to “personality types (sanguine and melancholic)” and “three kinds of spirits, which acted on the humours: the natural, the vital, and the animal” (485). “Eighteenth-century bodies-in-pain felt different to modern ones,” Bourke concludes, because “the figurative languages of humoural bodies reveal different ways of being-in-the-world” (488).
Because different cultural worlds have a shared biological basis, however, they are not entirely opaque to each other. It is possible to translate these analogies into different as-relations, as Bourke herself does—explaining what these kinds of pain or imbalances resemble (or are like), with what sorts of as-if connotations. As-relations point both ways—toward the nonequivalence of the terms of an analogy that are not like each other, and also toward the equivalences that make the analogy meaningful and useful because of ways in which the terms are like. Metaphors configure bodily experience as something it is not (anger both is and is not heat, a headache is both like and not like the steps of a bear or the pounding of a woodpecker, pain is both similar to and different from a blockage of a fluid). We can understand and recognize the plausibility of these analogies (what it is like to experience pain through animal metaphors or humoral categories) even though they are not our own because our bodies are biocultural hybrids that are not totally disconnected from our Japanese conspecifics or our eighteenth-century forebears. We share some bodily experiences and associations (some “survive,” as Bourke herself notes), but others are variable, and that combination allows commonalities as well as differences in what experiences of pain or anger or pleasure are figured as by different communities.
Figurative constructions like metaphors and narratives draw on as-relations that are prefigured in experience—analogies like these are deeply embedded in embodied cognition in its different degrees and kinds of grounding—and the as-if relations on which they are based can then be reconfigured in poems or narratives or other fictional forms in simulations that may reinforce or disrupt the configurations of the like and not like they draw on. The circuit between figuration and refiguration in the transformation of experience into narratives (the movement from mimesis1 to mimesis2) is possible because the connections posited by a cognitive or fictional simulation both are and are-not originally given in experience. This in turn is why the simulations of the as in the as-if of fictive reenactments can reshape the experiences of the recipients (as mimesis2 is taken up in mimesis3), which is how these culturally variable analogies get established and disseminated, through the stories that members of a community tell each other about what their anger, pain, or pleasure is like. The circuit of figuration (mimesis1, mimesis2, and mimesis3) through which experience gets transformed and reshaped in fictional reenactments joins action and perception through the as of simulation. Narratives have the power to refigure our lives in these ways because we are biocultural hybrids who engage the world through configurations of action and perception that are constrained by our bodies but also open to wide variation.
Narratives are constituted by the intertwining of different modalities of action, most basically by the interaction between discourse and story—that is, the relation between the act of telling and the actions and events emplotted in the told that is fundamental to our capacity to exchange and understand stories. The ability of narratives to bring into relation different kinds of action is grounded in (and can consequently have a powerful effect on) the role of action in coordinating different cognitive processes in our embodied engagements with our worlds. As Berthoz and Petit observe in their authoritative study of The Physiology and Phenomenology of Action (2008, 42–43), the synthesis of “multiple systems . . . of sense data” in the brain is accomplished through action: “The unity of the perceived world depends upon the extraordinary capacity of the brain which, in the first instance, breaks up the world into multiple components. The world of our lived experience is then the result of a synthesis of the activity of all these stations. . . . The act is an indispensable feature of this unity.” For example, as Merleau-Ponty (2012 [1945], 151) points out, “What unites the ‘tactile sensations’ of the hand and links them to the visual perceptions of the same hand and to perceptions of other segments of the body is a certain style of hand gestures, which implies a certain style of finger movements and moreover contributes to a particular fashion in which my body moves.” Because the felt unity of embodied experience is generated by the underlying “style” of these interactions, Merleau-Ponty famously declares that “the body cannot be compared to the physical object, but rather to the work of art” (152). Narratives are powerful cognitive and existential instruments because of their ability to organize and reorganize action of many different kinds, on different levels, across many perceptual and experiential modalities.
A particular style characterizes different modes of cognition because they are what Alva Noë (2015, 10) calls “organized activities.” As he explains, “Our lives are one big complex nesting of organized activities at different levels and scales.” These styles of organization are patterns integrating the differences registered by our perceptual equipment, the epistemological habits that develop over our lifetimes of embodied cognitive experience through Hebbian wiring and firing. The style that organizes these perceptual and cognitive modalities characterizes in turn our style of being-in-the-world, the characteristic intentionality (the directedness toward people, states of affairs, and the future) that comes to define our sense of self. Cognitive narrative theorist Guillemette Bolens’s (2012, 22, 28) definition of gesture as “kinesic style” encapsulates these linkages: “Kinesic styles interconnect all degrees of expressivity. . . . A person’s kinesic style is perceptible in her idiosyncratic movements and the singular way she negotiates social codes and physical constraints, while the kinesic style of a literary work is conveyed through its narrative dynamics.” A person’s distinctive kinesic style has to do with how her different ways of acting in the world are organized and interrelate, and the “dynamics” of a narrative are the ways in which it organizes the interaction of the different kinds of action it sets in motion.
The ability of linguistic narratives to coordinate embodied action depends on the relation between language, speech, and gesture. As Merleau-Ponty (2012 [1945], 187) argues, “The body is a natural power of expression.” In his view, the expressive powers of language ultimately derive from an embodied capacity for communicative action that is evident in gesture: “Speech is a genuine gesture. . . . This is what makes communication possible” (189). According to this theory of the gestural basis of communication, the codes and conventions of language are secondary rather than primary. The rules of a language are “the depository and sedimentation of acts of speech” (202), abstract and socially encoded residues of expressive linguistic and bodily action.
Contemporary neuroscientific findings about the development of language confirm Merleau-Ponty’s phenomenological intuition about the interdependence of language, speech, and gesture—the many deep connections between our ability to communicate with one another and our embodied capacities for expressive action. These connections pass through the motor cortex. As Pulvermüller (2018) explains, the motor cortex plays a key role in language development because it must bind with the auditory cortex in order to link the perception of speech sounds with the physical action of articulating speech through movements of the tongue and the lips. According to Marc Jeannerod (2006, 154–55), these developmental connections explain why, in adult communication, “speech perception . . . activates motor structures involved in speech production.” When we listen to someone talk or even when we read a written text, the motor cortex is active even though our bodies may be totally still. As Michael Tomasello (2003) argues, these linkages are part of a broader network of connections between the motor cortex and various other areas of the brain (visual, auditory, haptic), cross-cortical connections that develop through the communicative exchanges that begin in infancy as parents and children imitate one another’s expressions and otherwise gesture toward one another. These exchanges create links between the production and observation of actions of the kind registered in the mirror-neuron experiments (also see Armstrong 2013, 131–74). According to Tomasello (2003, 35), “It is gestures that for many children seem to be the first carriers of their communicative intentions. And it is gestures that seem to pave the way to early language.” Kenneth Burke’s (1966) memorable description of language as “symbolic action” is evocative and appropriate because it concisely encapsulates the diverse and profound connections between linguistic action and the bodily capacity for expressive communication evident in gesture.
As symbolic action, however, language both is and is not like other kinds of bodily action, and these differences also matter. The relation between linguistic and bodily action is paradoxical, as Shaun Gallagher (2005, 122–23) observes: “The thing that makes gesture more than movement is that gesture is language.” At the same time, however, he cautions that “it would be wrong to lose track” of the fact that “gesture is nonetheless movement.” The gestures that accompany speech often have communicative power, but not all of the gestures people make when they speak are expressive—some are just movements. By the same token, the expressive and instrumental dimensions of an action are related but not reducible to each other. Some movements that may seem expressive are primarily instrumental actions aimed at making changes in the world. Because actions have styles, an instrumental action may seem expressive, but not all actions are communicative gestures, and it is often possible to distinguish the communicative and instrumental aspects of an action. For one soccer commentator it is the beauty and grace of a goal that stand out, for example, while for another it’s the tactical cunning and skill that matter.
Our bodily capacity to act makes it possible to communicate, but action and communication are different and (sometimes) separable. To recall an example I have already cited, people can lose the ability to act instrumentally through injuries to the motor cortex but still retain language and the capacity to produce and understand speech. Someone who is physically paralyzed may be able to understand and discuss actions which he or she cannot perform, and this would not be possible if action and communication were identical (see Hickok 2014). It is sometimes the case, however, that physical impairments result in deficits in language production and comprehension, and that is because the motor cortex is a hub that connects instrumental and expressive action. Physiologically as well as experientially, instrumental action and communicative action are inextricably linked, but they are not the same.
These distinctions are subtle and difficult to pin down, but they have important consequences. The relation between communicative and instrumental action is a paradoxical relation of equivalence and nonequivalence that is powerfully generative. For one thing, it helps to explain how we are able to organize actions into stories. The paradoxical relation of like and not like that holds between motoric and communicative action enables narratives to stage interactions between different levels and kinds of action. The fact that expressive movement and other modalities of cognitive and motoric action are at one and the same time like and not like one another—equivalent in many respects as styles and kinds of action, but nevertheless ultimately not reducible to each other—is what makes it possible for them to interact. Because of their many similarities and connections, their differences can resonate off each other and be shaped into meaningful configurations—as when dancers engage in expressive movements that can be read as steps in a choreographed performance or as when narratives emplot actions and communicate them in acts of narration. Similarly, the distinction between communicative action (gesture) and instrumental action (motoric and bodily capacities of various kinds, whether the movements of a dancer or the activities of characters in a story) makes possible the intertwining of discourse (the communicative act of telling) and story (emplotments of instrumental actions and events).
Keeping track of the differences between kinds of action is important for understanding how narratives perform various characteristic functions. For example, the ability of narratives to coordinate and integrate different modalities of action helps to explain how a story can create an illusion of presence and facilitate immersion in a fictional world. As Anežka Kuzmičová (2012, 25, 29) points out, narratives can produce “a higher degree of spatial vividness, arousing in the reader a sense of having physically entered a tangible environment (presence) . . . when certain forms of human bodily movement are rendered,” and that is because “the motor and pre-motor areas of their cortices become somatotopically activated—the hand area of the motor strip responding to hand-related action words, the feet area to feet-related action words.” Terence Cave (2016, 81) similarly explains that “when fictional mimesis works, . . . that’s not because it ‘paints pictures’ for the imagination. It stimulates the neural areas that are involved when we perform a given action or have a perceptual experience.”
Anatomically based neuronal resonance is necessary but not sufficient, however. The responses of particular neurons must also be organized into action assemblies. This is what Tolstoy does with such mastery, according to Elaine Auyoung (2018, 20–21). By directing the reader’s “attention to ordinary physical actions, such as pulling on a loose button, stepping on a piece of ice, or swallowing a piece of bread,” she explains, Tolstoy activates the reader’s kinesthetic, procedural knowledge of how to perform everyday actions, and this integration of motor resonances into patterns of familiar instrumental activity “enables the text to evoke perceptual immediacy in a surprisingly effective way” (and, as she shows by comparing various renderings of Russian phrases into English, in a manner that survives linguistic translation). When Levin swings a scythe in the memorable mowing scene, for example, the resonances activated by the descriptions of his actions in our motor cortex become the basis for a vivid representation because they are in turn organized into physical sequences of action (walking down the field, pausing for a drink and a chat). Only when the motor resonances set off by action words at the neuronal level are organized into recognizable bodily actions can they contribute to an illusion of presence—but a linguistic description alone without motor resonances would also lack the vividness Tolstoy achieves.
What matters is not just specific bodily references and neuronally based motor resonances but the narrative recreation of patterns of action, and these in turn must be readily synthesized in the act of reading. Here yet another level and modality of action comes into play. As Auyoung explains (32), what cognitive scientists call “fluency” in processing—“the subjective ease with which we perform mental acts”—can make something that is strange and even improbable seem familiar and acceptable. Her example is the wordless proposal scene in Anna Karenina that seems vivid and even moving although it is strictly speaking preposterous: “While the pleasure afforded by this miraculous exchange has many sources, one of them is how the scene, so utterly unique in the history of love, has been constructed from materials that are completely ordinary and ready to hand,” with Kitty and Levin drawing letters on a tablecloth with a piece of chalk (33). Communicative action (telling and reading) interacts productively here with instrumental action as the ease of following the word game the lovestruck couple play makes their bodily interaction come alive.
As we read or listen to stories, the ability to fluently construct consistent patterns fosters the building of illusions (see Iser 1974, 282–90). Fluency in comprehension aids and abets the illusion of presence because mimesis is not simply a correspondence of sign and thing or a matter of motor resonance alone but a process of connecting modalities of action. The ease with which connections can be made at the level of communicative action facilitates the rendering of instrumental action because, borrowing a metaphor from neuroscience, it “mylenates” the circuit joining the three types of mimesis. Just as mylenation speeds the passage of an action potential down an axon (see Bear, Connors, and Paradiso 2007, 96–97), so fluency in communicative consistency building facilitates the interaction between a narrative’s configuration of actions (mimesis2), a reader’s figurative reconstructions (mimesis3), and patterns of activity from past embodied experience (mimesis1). Fluent connections naturalize the strange and unfamiliar by facilitating the linkages through which narrative figuration passes.
The construction of patterns in reading or listening draws on our perceptual experiences with horizonal absences—the hidden and not yet in the situation of perception that Merleau-Ponty characterizes as the intertwining of the visible and the invisible. Recall how, as he explains, every act of perception has spatial, temporal, and intersubjective horizons across which it projects expectations about what is hidden from its perspective but available, we tacitly assume, to other perceivers or to us in the future. As Auyoung observes (2018, 21), “Literary artists who seek to evoke perceptual immediacy must contend with the fact that every sensory modality of the represented world—not just the visual modality but also the auditory, olfactory, haptic, spatial, and kinesthetic—is at all times absent from the text itself.” These absences are not fatal, however, because gaps and indeterminacies are a familiar feature of perceptual experience. “We are accustomed,” Auyoung explains, “to believing that more lies beyond the scope of our knowledge and perception,” and it is “our readiness to contend with partial representational cues in everyday, nonliterary experience” that makes the mind alive to “what minimal cues imply” (2013, 69, 60, 66).
The action-perception circuit in cognition interacts productively here with the communicative action of following a story. Again what matters is the interaction between modalities of action that are different but that resonate and intertwine. Just as in everyday perception we integrate the partial perspectives through which people, places, and things present themselves in our experience and fill out what is hidden from our view, so in reading we are predisposed to take necessarily incomplete perspectives and, through what Iser calls “gap filling” and “consistency building,” configure them into gestalts whose coherence encourages us to believe in them (see Iser 1978, 118–32, 165–79; also see Ingarden 1973 [1931], 217–84). As Iser argues, the eventfulness of the temporal act of reading contributes to and reinforces this illusion by re-creating the quality of happening that characterizes the represented events in the storyworld (see 1976, 118–51). Because we ourselves produce a narrative’s configurations, we may feel a particular intimacy with them that further occludes their immateriality and endows them with a sense of immediacy. The coherence between these related modes of activity implies a completeness that a text’s absences necessarily lack, and the ease with which readers can produce these patterns reinforces this illusion. The act of reading and the act of perception can interact in these ways because both involve processes of pattern completion that are facilitated by their fluency and coherence.
Reading is not the same as perception, however, and the concordant discordances of narrative rarely if ever leave these coherences unperturbed. Here again differences matter as much as the interactions they make possible. The illusion of presence in a storyworld is not the same as an hallucination but instead involves what Iser calls a “virtual dimension” (1974, 279–81). The virtuality of narrative immersion has an as-if quality that both is and is not like the perceptual experiences it draws on. If, as Iser (1974, 286) explains, “a consistent, configurative meaning is essential for . . . the process of illusion-building,” nevertheless at the same time, this need for consistency leads readers to exclude many other possibilities. As a result, he argues, “the formation of illusions is constantly accompanied by ‘alien associations,’ ” disturbances that make us “oscillate between involvement in and observation of those illusions” (286). These “alien associations” serve as reminders of the “quasi-ness” or virtuality of a fictional world.
The quasi quality of as-if states of affairs is often misunderstood. For example, Kendall Walton (1990, 244) misconstrues the paradoxical combination of equivalence and nonequivalence in as-if stagings of affect when he claims that “the emotions that audiences have in response to fictional works are not real emotions, but merely make-believe ‘quasi-feelings.’ ” He is right, of course, that “when audiences experience fear and anxiety in response to ominous images on screen, they do not flee from the theater because their emotional response is part of the pretense” (244). But this does not make their feelings any less “real” (they may cry real tears and feel real sadness as the heroine sings her final aria and “dies” in Carmen, for example, or they may actually jump in reflexive fright at the sight of a snake in a movie). The bodily resonances evoked by as-if illusions have a virtual dimension because they are only traces of the original experience they simulate, but the pretense of immersive aesthetic experiences can have real effects on the bodies and brains of the recipient that make possible the illusion of presence. Not simply a quibble about terminological distinctions, this misunderstanding matters because the combination of like and not like in virtual experiences of the as-if is what allows fictional stagings of action to have real effects on our embodied habits of action in the world—it is how the quasi experiences of mimesis3 in response to the as-if configurations of mimesis2 can circle back and refigure our real, lived, everyday embodied cognitive practices (mimesis1).
A similar failure to credit both the like and the not like undermines Roman Ingarden’s (1973 [1931], 160–73) assertion that declarative statements in a work of art are only quasi judgments that do not make the same claim to truth as real judgments. Again, the fictional statements in a narrative may not be admissible in a court of law (no judge is going to convict Raskolnikov of murder), but the character Dostoevsky creates in Crime and Punishment may acquire a lifelike presence in what Ingarden calls the “concretization” of a text’s potentialities that can affect a reader’s real attitudes toward (and judgments about) states of affairs in her or his world (recall the discussion in chapter 2 of Jenefer Robinson’s argument about the ways in which identifying with a character like Anna Karenina can affect moral and social appraisals of larger matters across the horizon of the fictional world). To dismiss fictional emotions or judgments as unreal because they are merely quasi is to misunderstand the real powers of as-if simulations. Sometimes quasi feelings and quasi judgments are just as real and true as the nonfictional states of affairs they simulate (and maybe even more so).
Narratives construct patterns of action that they then interrupt, reverse, and revise and that resonate off each other with different degrees of concord and discord, harmony and dissonance. As Iser argues, “The inherent nonachievement of balance” in the building and breaking of illusions “is a prerequisite for the very dynamism of the operation” (1974, 287). The imbalances of discordant concordance in narrative are inherently dynamic, and that makes them like other aspects of our cognitive lives. Shifting configurations of concord and discord create different possibilities of immersion in a storyworld and lead us as readers to alternate between participating in and critically reflecting about the illusions we create. This experience of back-and-forth movement between building and breaking illusions, between participating in and observing patterns of action that we produce as we read or follow a story, is why Cave (2016, 135) is right to declare that “it is a mistake to think that immersion and reflection are antithetical poles.” They are, rather, interdependent products of the dynamic interaction of narrative concord and discord.
Narrative disjunctions interrupt immersion and prompt reflection by playing with a fundamental characteristic of embodied action—its habitual, natural “unthinkingness.” As Jeannerod (2006, 59) notes, “We remain unaware of most of our actions, unless an unpredicted event interrupts their course and brings them to consciousness.” For example, the surest way to make someone fumble in the execution of a routine behavior is to call attention to it and introduce self-consciousness into actions that are ordinarily automatic and habitual. Similarly, as Heidegger observes in his analysis of the ready-to-hand (Zuhandenheit), the tool is invisible as long as we can use it effectively. Only when the equipment breaks down do we notice its equipmentality—only when hammering fails do the act of hitting the nail and the characteristics of the hammer become available for thought (see 1962 [1927], 95–107). Our ability to act fluently, based on procedural memory and habitual skill, depends on not thinking about what we are doing—not noticing, for example, the time lags between execution and awareness that Libet measures in his mind-time experiments. Stage fright and performance anxiety are well-known examples. By the same token, as long as things fit together seamlessly while we follow a story, whether in the transitions between the events emplotted in the told or in the connections between different aspects of the telling, our deeply rooted tendency not to pay attention to the ordinary actions we are engaged in supports and reinforces our unthinking involvement in the states of affairs we produce. Only when this flow is interrupted does the illusion get broken and critical observation take over. What Vanessa Ryan (2012) calls “thinking without thinking” is a pervasive feature of lived cognitive experience and ordinary action, and it is a phenomenon that stories take advantage of in order to draw us in and keep us engaged—that is, until reversals in the told or disjunctions in the telling interrupt these involvements and make us reflect on what we had hitherto been processing unreflectively.
The characteristic unthinkingness of action supports and sustains what phenomenology calls the “natural attitude,” our customary cognitive posture of unquestioned engagement with the world (see Armstrong 2012). As Merleau-Ponty points out, however, whenever we begin to reflect, we find a whole realm of unreflected meaning already there (see 2012 [1945], lxxiv–lxxxv). We may not have thought we were thinking, but we were. Reflection is consequently, in Merleau-Ponty’s view, a perpetually incomplete, never-finished activity of reflecting on unreflected experience that always outstrips it.
As a systematic philosophical method, phenomenology brackets the natural attitude and suspends its engagements in order to bring to the fore various otherwise unnoticed aspects of lived experience. This technique is merely a formalization, however, of ordinary experiences of cognitive disruption and reframing. In our everyday lives, surprising interruptions and anomalies have a similar power to bring unthought thought processes into view. The ability of cognitive disturbances and anomalies to expose otherwise invisible epistemological processes is what makes optical illusions like the Necker cube or ambiguous figures like the oscillating rabbit-duck such favorites of philosophers and neuroscientists (see Armstrong 2013, 67–72). Narrative discontinuities have a similar ability to disrupt and suspend the natural attitude that typically characterizes the reception of stories. There is already a quality of metafictional self-display in narrative disruptions and reversals because they have the power to expose and cause us to reflect about cognitive processes whose smooth functioning ordinarily depends on their invisibility.
The various disruptions to narrative continuity for which Joseph Conrad is well known illustrate how interruption can suspend the natural attitude. In an oft-quoted passage from Lord Jim, Marlow remarks that “it’s extraordinary how we go through life with eyes half shut, with dull ears, with dormant thoughts. Perhaps it’s just as well; and it may be that it is this very dulness that makes life to the incalculable majority so supportable and so welcome” (1996 [1900], 87–88). As Conrad’s fictions repeatedly show, however, the shock of baffling, unexpected events may intervene and produce “one of these rare moments of awakening when we see, hear, understand ever so much—everything—in a flash—before we fall back again into our agreeable somnolence” (88). Moments of bewilderment in his narratives again and again suspend unthinking involvement in the world and expose its tacit assumptions and operations to view (see Armstrong 1987)—for example: Jim’s surprise that he had abandoned ship (“I had jumped . . . it seems” [67]) and that the bulging rusty bulkhead had not burst, Marlow’s angry annoyance that a guilty young man should “look so sound” (“well, if this sort can go wrong like that” [29]), Gentleman Brown’s violent destruction of the community of trust on Patusan, and the disillusioning impact of Jim’s death on his supporters (“He passes away under a cloud, inscrutable at heart” [246]). These interruptions all disrupt patterns of consistency building that Jim, Marlow, and Patusan hadn’t previously been aware of. As Marlow comments at one point, “It is always the unexpected that happens” (60)—and the unexpected can give rise to thought by exposing what we hadn’t known we were thinking.
The revelatory power of narrative discontinuities—whether the twists and turns of traditional stories or the more radical disruptions of experimental, “unnatural” fictions—is parasitic on the habitual, automatic, unthinking everyday fluency of action and perception that they interrupt. In Lord Jim, at the level of the events of the story, discontinuities in the flow of the action in what would otherwise seem to be a tale of adventure or an imperial romance raise moral and existential questions about the contingency of our ethical norms and cultural codes that give Conrad’s works an existential and metaphysical dimension usually absent from these genres whose conventions he subverts. In the discourse, Conrad deploys various strategies of narration that disrupt the presentation of events—the reversals of effect and cause, for example, that Ian Watt (1979) famously calls “delayed decoding” (“What had happened?” the narration asks when the accident occurs—“The wheezy thump of the engines went on. Had the earth been checked in her course?” (20)—only many pages later to identify the cause: “A floating derelict probably” [97]), or the many gaps that refuse to cohere in the perspectives through which the tale is offered as Marlow tries to piece together the contradictory views of the many different informants he consults. If fluency of action in the telling and the told facilitates immersion by drawing on the natural attitude, disruptions like these in the act of narration break the illusion in order to expose and call for reflection about processes of pattern construction that narratives otherwise tacitly exploit.
The interaction of these modalities of narrative action calls to mind the well-known terminology of affordances introduced by Gibson’s (1979) ecological theory of perception. According to Gibson, what we perceive in a scene are not the properties of objects but the possibilities of action they implicitly afford—that is, in his words, what “the environment . . . offers the animal, what it provides or furnishes, either for good or ill” (127).15 According to this “ ‘do-ability’ theory,” as Berthoz and Petit (2008, 66) explain, “we perceive primarily what has a practical bearing on our action.” When we follow a story, we similarly respond in different ways and at different levels to the possibilities of action it offers. As Glenberg and Kaschak (2002, 559) explain, this entails a process of “meshing affordances”:
For example, one can judge that the sentence, “Hang the coat on the upright vacuum cleaner” is sensible, because one can derive from the perceptual symbol of the vacuum cleaner the affordances that allow it to be used as a coat rack. Similarly, one can judge that the sentence “Hang the coat on the upright cup” is not sensible in most contexts, because cups do not usually have the proper affordances to serve as coat racks. . . . If the meshed set of affordances corresponds to a doable action, the utterance is understood. If the affordances do not mesh in a way that can guide action (e.g., how could one hang a coat on a cup?), understanding is incomplete, or the sentence is judged nonsensical, even though all of the words and syntactic relations may be commonplace.
The consistency building that Iser describes is not only a matter of synthesizing the perspectives through which states of affairs unfold in a text but also a process of integrating (meshing) the affordances it offers—that is, of coordinating the possibilities of action held ready, for example, not only by the events emplotted in the story but also by the acts of interpretation and evaluation rendered by the discourse. As a structure of concordant discordance, however, the dynamics of a narrative will sometimes mesh these affordances and sometimes not, with various effects and for different purposes. Most competent readers of literature, for example, can readily imagine contexts in which the affordances of “Hang the coat on the cup” would be appropriate (think of Gertrude Stein’s Tender Buttons or Alice’s Adventures in Wonderland) precisely because the dissonances in their refusal to mesh in an ordinary, predictable way are resonant with new and interesting possibilities of response.
This is the kind of example Cave (2016, 46–62) has in mind when he proposes the concept of literary affordances. These refer not only to the possible actions associated with states of affairs represented in a novel or a poem but also to the responses held ready by its forms. The conventions of a genre, for example, afford more or less specific sets of action by writers and readers in their construction of and their responses to texts. These affordances are sometimes quite constrained, as with the sonnet, but they are never completely unbounded even in more open-ended genres like (to echo Henry James) “the big baggy monster” of the novel. There are affordances that have to do with the telling as well as with the told, and these are not the same, but the dynamics of the narrative are a matter of how these affordances intertwine—how the actions set in motion by the discourse mesh (or not) not only with each other but also with the actions afforded by the story. Discourse and story can have such interesting and unpredictable effects on each other because they are not inert forms with objective properties (as the reified taxonomies of structural narratology suggest); rather, they are modalities of action whose affordances we seek to mesh but that (thanks to the concordant discordances of narrative) settle into stable, coherent patterns only (if ever) at the end (and then only with texts where the possibilities of action afforded by both the telling and the told are finally closed).
As Cave points out, the affordances of literature typically invite improvisation: “What literature affords, indeed, is cognitive and cultural fluidity” (39). Affordances constrain but do not absolutely, completely determine our engagements with them. A chair affords sitting, for example, but it can also be used for a handstand or as a weapon in a bar brawl. With linguistic artifacts, according to Cave, this duality also reflects the fact that “all uses of language are highly underspecified” (25), manifesting the gaps and indeterminacies that Iser, Ingarden, and Auyoung emphasize. Literary texts take particular advantage of the underspecification of language and the capacity of these absences to make possible improvisatory, unpredictable responses. As Cave vividly explains, literature’s “bold and highly precise modes of underspecification act like a prompt or a trampoline, creating unlimited possibilities for imaginative leaps into the blue—or into the minds of others” (27).
The duality of affordances in constraining response while inviting (and in the case of literary affordances even requiring) improvisation gives them a particular kind of historicity. In Cave’s words, they are “Janus-faced: they point both back towards” the history of previous involvements that have established their current configurations and our inherited sense of their potentialities, even as they point “forward along time’s arrow towards the present and future outgrowths of human cultural improvisation” (62) that may take them in unforeseen directions. Noting, for example, that “the essay had an afterlife that Montaigne could not possibly have imagined,” Cave observes that “literary instruments are often defined and recognized retrospectively. Thinking of genres as affordance structures helpfully reorients attention forwards, towards the restless reworking of existing templates” (58) that writers and readers engage in by improvising in response to the possibilities of action that previous texts make possible.
In the case of narrative, the terminology of affordances is especially relevant to the much-discussed problem of probability.16 Responses to the probabilities afforded by a narrative are, as Karin Kukkonen (2014b, 372) argues, a particular instance of the general rule that “embodied cognition is profoundly informed by sensing how actions are going to develop.” Proposing a Bayesian model of cognition as predictive processing, Andy Clark (2016, 5) argues that “we see the world by . . . guessing the world.” This is true of perception in general as well as in what Iser (1976, 108–18) describes as the anticipatory and retrospective to-and-fro processes of reading and following a story. According to Kukkonen (2016, 157), narratives have what she calls a “probability design”: “As the plot arranges the sequence in which events are related in the narrative, it gives readers new observations about the fictional world, and these new observations can confirm or contradict their probabilistic, predictive model of what is likely to happen in the fictional world.” These predictions about probability are not confined to the plot, Kukkonen points out, but interact in “feedback loops” with other modalities of action in the narrative that have an effect on readers’ expectations—their predictions (typically nonthetic, unthinking, intuitive) about what will happen next in their experience of following the story, whether this is triggered by interactions among the characters and events in the plot or by a narrator’s storytelling acts (what she holds back or reveals, how she judges, and whether we trust or are suspicious of those evaluations).
Not restricted to the level of the plot, predictions of probability in the unfolding of narrative action bring into dynamic relation all the levels of and interactions between the many various modalities of action in a narrative. As the narrative world unfolds in our experience in ways that either accord with or defy our expectations, Kukkonen (2014a, 724) observes, “it is not unusual that readers find different probabilistic models for the fictional world” in conflict or competition with one another, as when narratorial discourse comments ironically on plot events and leads readers to expect different things from what the characters may anticipate, or when our sense as readers, listeners, or viewers of what a genre typically affords (the rewards and punishments characteristic of a Greek tragedy or a Hollywood movie, for example) leads us to project probabilities not yet predictable from the events themselves. The probability design of a narrative is thus not a fixed structure but a dynamic, fluid, complex interaction between different modalities of action that sets off various predictive processes in the reader, feedback loops and “cascades of cognition” (as Kukkonen calls them), usually occurring beneath awareness through intuitive, prereflective assessments but sometimes breaking into self-consciousness in abstract, reflective judgments. These loops and cascades in response to narrative probabilities once again draw on and simulate perceptual processes at work in our everyday cognitive lives, where we are constantly comparing, synthesizing, and negotiating predictions about our worlds. And this (again) is why our experience with narrative probabilities can inform and refigure our predictive behavior in other aspects of experience.
The ways in which actions intertwine in a narrative are paradoxically both evidence of its social and historical particularity and evocative of resonances that enable stories to produce effects in listeners across temporal and cultural distance. This paradox is a consequence of our biocultural hybridity as embodied creatures whose capacities for action and cognition are shaped both by species-wide evolutionary processes and by our local, situated experiences as members of specific communities. As Kukkonen and Marco Caracciolo (2014, 267) point out, “Bodily experience shapes cultural practices” even as “cultural practices help the mind make sense of bodily experience.” On the one hand, as they observe, “relatively stable bodily patterns” across history enable readers to respond emotionally and kinesthetically to texts from different cultural settings and historical periods (266). But on the other hand, as Bolens (2012, 42) notes, “the historical, cultural, and social context in which any kinesic signal is performed always bears on the gesture’s sense, both for the person who emotes and for the person who perceives the emotional signal.” The sedimentation of kinesic signals in linguistic signs, Bolens observes, can even result in perceptible cultural differences between the resonances of action words that embody “different ways of conceptualizing the dimensions of motion events” (38), differences evident in the English lexicon in the contrast between verbs with French or Latin as opposed to Anglo-Saxon roots.
Linguistic renderings of action may consequently at one and the same time demarcate and transcend historical distance. For example, quoting a passage from The Adventures of Ferdinand Count Fathom, a novel by Tobias Smollett published in 1753, Kukkonen (2014b, 368) claims that the “forceful kinesic language of the effects that fear has” on a character “give[s] the embodied reader a fairly precise idea of [his] state of mind, and this idea is replicable in the reader’s own embodied resonances to the kinesic shape and directedness of the words she reads.” The diction of the passage conveys these emotions, however, through phrases like “the brandishing of poignards,” “every fresh filip of his fear,” or “a new volley of imprecations” that are remote from the linguistic habits of most twenty-first century readers.17 These linguistically coded kinesic signals, in all their strangeness, mark the historical and cultural boundaries that are being crossed if the passage succeeds in conveying an embodied sense of the character’s anxieties—and it might not succeed in doing so if the unfamiliarity of the diction overwhelms the familiarity of the embodied emotion. As Bolens explains, “The kinesic style of the narrative, . . . its corporeal hermeneutics, . . . is shaped by the way the author stylizes language so as to create signifying gestures” (35). Because a style is an as-structure that conveys what an action is like through linguistic means that are also not like what it simulates, the kinesic style of a story will paradoxically signal its historical and cultural specificity in and through the very means (the linguistic action of the description) by which it transcends these horizons (in the embodied resonances the language produces in distant, future readers).
Stories can act as mediators between our different lives because they configure and refigure modalities of action that intertwine in our experience of the world. As biocultural hybrids, we share embodied motor capacities with other members of our species and engage the world through action-perception circuits that have evolved over a long shared history. As configurations of action, stories can put us in relation with other worlds that may seem familiar and strange because they draw on but also reorganize aspects of embodied experience we have in common with other conspecifics. Our biocultural hybridity makes it possible for us to resonate bodily to stories whose actions are far from our worlds, even as the way these resonances are styled may also make us feel how distant those worlds are from our own everyday experiences.
These differences in turn can be productive in all sorts of ways because of the further actions they can set in motion. Like other kinds of affordances, the stories we exchange with one another face in two directions. On the one hand, they bear traces of the experiential, cognitive worlds they draw from even as, on the other hand, they may reorganize these resonances into configurations of narrative action that offer possibilities of response to future generations of readers and listeners that are not entirely specified in advance. Human actions are intrinsically horizonal and future directed, characterized by both immanence and transcendence, as they project us beyond our present situation to the purposes and aims they are directed toward. The historical lives of stories as they get exchanged across cultural and generational horizons are made possible by the powers of transcendence embodied in the actions that narratives set in motion. The power of stories to act across cultural and historical distance may seem miraculous, but it is simply an example of the capacity of our species to engage in future-oriented, boundary-crossing actions. The act of telling stories can project us across the horizons of our worlds because actions transcend the situation from which they are launched as they aim toward ends not yet fully in sight.