10   Embodied Understanding

In this chapter, I look specifically at the role of exemplification in dance. By so doing, I hope both to reinforce my earlier argument that the arts are vehicles of understanding and to highlight the ways in which symbols that are not and do not purport to be true can be vehicles for understanding. This is crucial for what follows. In the next two chapters, I will argue that such symbols are ubiquitous in the sciences.

Dances as Symbols

Swan Lake is beautiful. It is delicate, graceful, enchanting. Martha Graham’s Night Journey is not. It is riveting, harrowing, horrifying, often ugly. Yvonne Rainer’s Trio A isn’t even that. Being utterly pedestrian, it does not play on the emotions at all. But it is intriguing. Taken together, these three dances raise questions: What is dance up to? What does it do and how does it do it? Night Journey discredits the thesis that the end of dance is beauty. Trio A discredits the thesis that the end is affective engagement. Possibly dance as such has no end. Different works and different genres pursue different ends. But whether or not dance as such has a telos, questions arise: What does this particular dance do? How does it do it? And why?

I will argue that beauty, ugliness, motion, and emotion are among the means dance uses to embody and convey understanding. This is not to say that dance has only epistemic functions. We may value dances for their own sake, or for the pleasure they give, or for other reasons as well. What I claim is that one important function of dance is epistemic. Dance enriches our lives at least in part because it enables us to understand things differently than we did before.

Like Goodman (1968), I believe that dances consist of symbols. This claim is not, on the face of it, problematic. Dance critics regularly speak of a genre’s vocabulary or idiom. If this way of speaking is accurate, choreographers draw on the resources of a symbol system to create their works. In that case, to glean an understanding from a dance, spectators must interpret those symbols correctly. They must recognize what the dance symbolizes. They must know how to read the dance. If they do, they understand what the dance conveys. But for the dance to convey understanding, rather than merely to be understood, more is required. The spectators must also have reason to accept what the dance conveys.

This is problematic. How can dances provide such reasons? Even if dances are comprised of symbols, they present no arguments. They do not make statements, express propositions, or assert that the world is this way or that. Some, like Swan Lake and The Firebird, have a narrative structure. But they are fictions. And fictions do not make literal assertions about the way the world is. Moreover, much of what seems significant in such works does not figure directly in the plot. So, evidently, something else is going on besides telling a story. Other works, like Trio A and Merce Cunningham’s Points in Space are nonnarrative. They do not stand in a representational or denotative relation to the world. They are not truth apt.

To make my case, I need to undermine the idea that only symbols that figure in arguments—that is, only symbols that have a propositional structure or are components of items with a propositional structure can advance understanding. If nonpropositional items can advance understanding, then the thesis that dance advances understanding has some chance of being correct. Let’s look at some cases.

Consider the counterexample. If Mike asserts, ‘All swans are white’, all it takes to refute him is one black swan. No words need be spoken. The bird alone is enough. The black swan contributes to Mike’s understanding of ornithology by demonstrating to him that his belief about swan plumage was false. It may, of course, do more. Perhaps it has a ripple effect, prompting him to revise a cluster of associated beliefs.

As van Fraassen (2008) argues, perspective also figures in our understanding of the world. By adopting a different perspective, we come to see familiar items in new ways. We thereby appreciate relationships between them that we previously had overlooked or underemphasized. For example, the shift from a third-person to a first-person perspective may be crucial to appreciating the close connection between belief and assertion. The assertion ‘It is raining and Kate does not believe it is raining’ is unproblematic. There are, after all, many facts that any given person does not believe. But ‘It is raining and I do not believe it is raining’ is an instance of Moore’s paradox. It is not something I can reasonably assert. The first-person perspective, but not the third-person perspective, thus affords reason to think that assertion is intimately connected with belief.

A third case is pattern detection. Even if all the evidence is in hand, understanding is enhanced when a pattern emerges. In such a case, although the facts were previously known, the relations between them were not perspicuous. Snow’s plot made manifest that virtually all the cholera victims of the 1854 epidemic got their drinking water from a single source (Tufte, 1997). It led to the obvious, but at the time radical, conclusion that contaminated drinking water was spreading the disease.

One might object that all of these can be captured in propositions. So, one might think, there is an implicit argument. It’s not the black swan, then; it is the proposition: ‘Here’s a case that shows that your hypothesis is false: [insert black swan]’. It’s not the perspective, then; it is the proposition ‘Looked at from this point of view, you will see that the following is unassertible [insert an instance of Moore’s paradox]’. It’s not the pattern; it is the proposition ‘This plot shows that the data cluster around a single point [insert map]’. Although one can frame such arguments, they do not paraphrase away or capture in propositions the items in question. Those items have been embedded into propositions. But they do their cognitive work independently of such embedding, and are only worth embedding because of their prior cognitive status. In Wittgenstein’s (1961) terms, they show rather than say. They exemplify.

Propositions are standardly held to be that which truth-apt sentences express. Two sentences express the same proposition just in case they are co-intensional. Moreover, it is held, the co-intensionality condition is sometimes satisfied; some pairs of sentences are co-intensional. If ‘vixen’ and ‘female fox’ are synonymous, then ‘Vixens are fierce’ and ‘Female foxes are fierce’ are co-intensional. They express the same proposition. Any world in which vixens are fierce is one where female foxes are fierce; any world in which female foxes are fierce is one where vixens are fierce. Recently, metaphysicians have argued that propositions are too coarse-grained to do the work philosophers want done. One problem stems from the human capacity for ignorance and error. Even if ‘vixen’ and ‘female fox’ are synonymous, it is entirely possible for Harry to believe that female foxes are fierce without believing that vixens are fierce. Even if ‘London is beautiful’ and ‘Londres est belle’ express the same proposition, Pierre can believe one but not the other (Kripke, 2011b). Propositions, as standardly characterized, are not sufficiently fine-grained to reflect belief contents. Such contents are evidently hyperintensional.

Some, such as Bealer (1998), reject the co-intensionality criterion and maintain that propositions are hyperintensional too. If so, perhaps we might say that the contents of dances and other works of art are propositional. This would require a rather drastic departure from tradition in at least two respects. First, it would recognize that propositions need not be truth apt. It is hard to see what the truth conditions on Points in Space or Trio A could be. Second, it would recognize that there are cases where a symbol is so fine-grained that no other symbol could have the same content. Such symbols are semantically dense (see Goodman, 1968). Although the notion of a proposition can be stretched, I am not confident that hyperintensionalists want to stretch it this far. In any case, our concern is with what dances exemplify, not with their claims—if any—to be propositional.

Dances highlight certain properties, rendering them salient, and thereby affording epistemic access to them. Classical ballet, for example, literally exemplifies properties such as grace, delicacy, and beauty; and metaphorically exemplifies properties such as love and longing, weightlessness and ethereality. Martha Graham’s modern dances metaphorically exemplify psychological properties such as grief, regret, horror, and hope. They literally exemplify that the body of the dancer has a certain weight—that it is subject to literal as well as metaphorical gravity. George Balanchine and Merce Cunningham, choreographers who bridge the modern—postmodern divide, created works that exemplify properties of dance itself, movements of dancers in time and space. They also exemplify properties like vitality, dexterity, and sinuousness. The works of the choreographers in the minimalist, postmodern Judson Dance Theater exemplify properties of ordinary pedestrian movement. Rather than exemplifying properties stereotypically associated with dance, they exemplify activities like walking, running, carrying a mattress, and climbing over a barrier.

I have characterized the properties dances exemplify using monadic predicates. This might suggest that they can be instantiated in an instant. But the important properties in dance are typically dynamic. They emerge and develop over time and across space. Jocasta’s convulsive grief, Odette’s ethereal grace, Cunningham’s jittery counterpoint, and Paxton’s prosaic walk are spatiotemporally extended. This, I suggest, is crucial to their epistemic functions.

A Look Back

The Judson Dance Theater consisted of a group of minimalist postmodern choreographers who sought to pare down dance to what they took to be its essence: human bodies moving in space. They sought to democratize dance, to eliminate its elitist, distancing, off-putting qualities. They had no interest in storytelling, in transcendence, in illusion of any kind. Their dances consist of mundane, pedestrian, nonstylized, uninflected movements of the sort you can see on the street. Yvonne Rainer expressed her choreographical ideals in her over-the-top “No Manifesto” of 1965:

No to spectacle no to virtuosity no to transformations and magic and make believe no to glamour and transcendency of the star image no to the heroic no to the anti-heroic no to trash imagery no to involvement of performer or spectator no to style no to camp no to seduction of spectator by the wiles of the performer no to eccentricity no to moving or being moved. (Rainer, 1995)

Once all these aspects of dance are excluded, what is left, the minimalists maintained, is the essence of dance—movement as such.

What might be the value of such a dance? We see people walking, running, climbing over barriers, carrying loads every day. Why should we go to a performance to watch them? Why should we buy a ticket? Sally Banes (2003, 3–5) suggests that the answer lies in defamiliarization, a process by which what is familiar is rendered strange. The idea is this: when something is familiar, we are so accustomed to it that we do not really look at it or attend to it. A passing glance enables us to recognize it as what it is and then move on. Defamilarization heightens awareness of things that are so obvious that we routinely ignore them. We walk, run, climb, and see others doing so without giving it much thought. When we carry a mattress, we do give it thought. We are painfully aware that carrying a mattress is hard. It requires continually readjusting our bodies to accommodate the awkwardly shifting center of gravity of the bulky, heavy, unwieldy burden. But we are intent on the task—we want to get the mattress moved. So we attend to the task and not to our doing of it.

The Judson Theater dancers put us in a context where we attend to the physical intelligence that goes into such mundane activities. We notice and attune ourselves to the minute, intricate, muscular adjustments involved in keeping one’s balance while schlepping a mattress. We notice the rise and fall, the small and large physical adjustments that it takes to walk or run across the floor. The dances thus exemplify features that mundane motion instantiates but that we, either makers or observers of that motion, routinely ignore. The exemplification is literal. The dancers exemplify features of walking by walking. They exemplify features of climbing by climbing. On the one hand, their message seems to be ‘What you see is what you get’. On the other hand, they put spectators in a context where they can ask, ‘Well, what do we get?’ and see, perhaps for the first time, what was before their eyes all along. By making us aware of the physical intelligence of ordinary, mundane movement, the Judson Theater’s dances increase our awareness and advance our understanding of ourselves as organisms capable of locomotion.

Arguably, they do something more. Although many of the discrete, component movements in Rainer’s Trio A are ordinary movements that pretty much anyone could do, it is not the case that the complex movements that they are part of are things that just anyone can do. As Jill Sigman says, “Moving a head one way and feet another is difficult enough, but switching quickly from head to feet to other body parts is even more challenging.… Furthermore, some of the movements are simply difficult to accomplish. One passage requires slowly rising into relevé on one leg and repeatedly alternating legs. Another involves squatting and extending the left leg fully to the back, then bringing it under the torso and through to the front without losing one’s balance” (1998, 164–165). This undercuts the idea that the uninflected, unvirtuositic dances are just ‘slices of life’ brought indoors and presented in such a way that we can attend to ordinary movement for its own sake. Nevertheless, the dances exemplify ordinary movements, and present the more complicated movements in the same uninflected way as they present the ordinary ones. Perhaps they thereby exemplify that the ordinary is continuous with the extraordinary, or that the ordinary is itself extraordinary.

According to the 1960s minimalists’ ideology, there is something phony or inauthentic about dances that purport to be something other than what they are. So it is perhaps not surprising that exemplification figures prominently in their works. Since exemplification requires instantiation, a symbol can exemplify only what it is—that is, only features it has. This suggests that what is characteristic of postmodern minimalist dance might not be characteristic of other forms of dance.

Ideologically, Merce Cunningham and George Balanchine are not far from the minimalists. They too eschew narrative and psychological expression. They too want to pare dance down to its essence. But rather than taking the essence of dance to be human bodies moving in space, they take it to be dancers moving in space. So their works exemplify features of dance itself. Dancerly forms, movements, and patterns are exemplified in their works. Points in Space, Cunningham (1986) maintained, is about dance; it is not about anything else. Unlike the Judson group, Balanchine and Cunningham do not purport to restrict their range to movements anyone at all could do. They recognize that dancers have extraordinary physical and expressive abilities, and are willing to exploit the full range of those abilities. Thus they take properties like grace, virtuosity, suppleness, and the illusion of weightlessness to be suitable candidates for exemplification. So are abstract geometrical and kinematic patterns that can be realized only by trained, talented dancers.

Again the question is: what is the cognitive value of such a dance? A dance about dance could embody and convey an understanding of the art form. It could show what dance (or perhaps ballet, or perhaps a certain style of modern dance) as such does and how it does it. It could make manifest that dance is not just, or not mainly, a sort of entertaining pantomime for telling fairy tales without words—for conveying something that could be better done in words. Balanchine seems to suggest that classical ballet starts, ‘Once upon a time, there was a prince who fell in love with an enchanted swan …’ Now take away the prince, take away the swan, take away the love and betrayal, take away the enchantment, and what do you have left? The answer to that question, arguably, is what Balanchine makes manifest in his ballets.

Still, there is something irritatingly self-indulgent about artists’ talk of exploring the limits of their medium. One wants to reply, ‘Yes, yes, I can see why artists working in a medium and art students studying a medium need to care about the limits of the medium. But why should the rest of us care? What sort of understanding does such an exploration yield for us?’ Later, I will suggest reasons to think that Balanchine, Cunningham, Rainer, and their colleagues provide acceptable answers to these questions. For now, however, let us look at their predecessors.

Modern dance tends to respect nature. Human bodies present themselves as human bodies, not snowflakes or enchanted swans or ethereal spirits. They are subject to the laws of physics and psychology, and sometimes ground down to the earth by the forces acting on them. Modern dances are more likely to present dancers writhing on the ground than leaping improbably through the air. Psychology is central. In Martha Graham’s works we see the outward manifestations of mental states, expressing fear, joy, elation, and revulsion. Rather than relying on established conventions, as classical ballet does, they create their own meanings—they constitute the symbols that convey their content.

One way is through narrative. Knowing, as we do, the story of Oedipus, we have resources for interpreting the gestures in Night Journey. But Night Journey is no mere pantomime of Oedipus Rex. It presents Jocasta’s mindset at the moment of her suicide, something Sophocles left out. It reviews the joys and sorrows and eventual horror of her life with her son/lover/husband Oedipus. It displays the guilt, revulsion, and self-loathing that make suicide the only option. It does so through tensions and releases, conventional gestures and newly invented ones—motions that express tenderness, repugnance, love, and profound regret. The dance intimates that the true tragedy of Oedipus Rex is Jocasta’s. To have borne a son and, loving him, to have (inadvertently and with the best intentions) brought it about that such a horrifying fate could befall him is to have utterly betrayed one’s obligation as a mother. To have loved a man and put him in a position where all he could feel for you or himself is loathing and disgust is to have utterly betrayed one’s obligation as a lover and a wife. Night Journey enhances our understanding of Oedipus Rex. It presents the story from a novel point of view, and reveals features that are not salient in Sophocles. It modulates and elaborates the understanding we glean from the play, convincing us that there is more to the story than the fate of one man with a tragic flaw. Whatever insight into the human condition we gain from the original play is extended and ramified through this new interpretation.

Modern dances, such as Night Journey, are not insular in the way the works of Rainer, Balanchine, and Cunningham are. They point beyond themselves and appeal to resources drawn from the outside. They refer to things that are apparently not integral to dance. They may tell stories to express feelings. They elicit and express emotions. I have argued elsewhere that emotions exemplified in the arts are vehicles for understanding (see Elgin, 1996, 2007b). Still, there is a worry. It seems that Martha Graham could give an utterly compelling performance of Night Journey even if she happened that day not to be feeling the amalgam of love, regret, revulsion, and self-loathing that the work expresses. The emotions expressed by the work evidently need not be felt by the dancer.

This, however, presents a difficulty for my position. Exemplification, as I keep insisting, requires instantiation as well as reference. If the dancer does not experience the emotions in question, how can the dance instantiate them? This is a tricky question that, I will argue, reveals something important about the philosophical significance of dance. But before I take up the problem, let us take a brief look at classical ballet.

Classical ballet is practically the antithesis of Judson Dance. It says ‘yes’ to many of the items listed in Rainer’s “No Manifesto.” Yes to spectacle, illusion, and make believe. Yes to magic, virtuosity, glamour, transcendancy of the star image. Emphatically yes to the seduction of the spectator. Classical ballets tell stories, magical stories about enchanted princesses and evil magicians. They rely on a fixed, regimented, vocabulary of techniques, positions, and steps that conspire to present the illusion of weightlessness, of ethereality. Classical ballet looks outward. It is about something other than dance. Still, it hardly seems to open a window on the world. It portrays fairy tales. It is populated by necromancers and their victims, by wizards and magical creatures. If there is a message here, it may be that ballet seeks to enchant, but we should be wary of being enchanted. Although ballets are about something in the sense of being comprised of representational symbols, there is evidently nothing in the world they are literally about. They are fictions. While telling stories about nutcrackers or firebirds or enchanted swans, they exemplify features like grace and delicacy, fluidity and transcendence. They make manifest how beautiful, light, and ethereal human beings can seem.

Dance as Philosophy

This cursory reverse history of dance in the West shows a paring down, a stripping away. Each genre I mentioned took its predecessors’ works to contain excesses that needed to be expunged, leaving only what was essential to dance. The move away from balletic spectacle was a progression toward the exemplification of only what is essential to dance. Philosophically, this is interesting. Who would have thought that dance is a hotbed of essentialism? But this trajectory raises an important question. Dance consists of human bodies in motion. So on this essentialist account, dance should be restricted to what human bodies in motion can exemplify. What is that? We know that an item can exemplify only properties it instantiates. So the question is, what sorts of properties can human bodies in motion instantiate?

The postmodern minimalists maintain that bodies cannot instantiate properties like being weightless, or being an enchanted swan. So dances that portray them as such are in some respect violating the essence of dance. They object to ballet’s pretense of defying gravity—of being lighter than one really is. But pretending is something ordinary people do, and pretending to be lighter than one is is something lots of us do. They object to the grand leaps as not the sort of thing that ordinary people do. But Michael Jordan, at the height of his career, probably made as impressively graceful leaps as Mikhail Baryshnikov. Granted, Michael Jordan is far from ordinary. But his extraordinary talent has nothing to do with dance. They object to ballet’s standardized, regimented symbol system, considering it an imposition of an authoritarian dance culture rather than deriving from the way ordinary people behave. But human beings are by nature acculturated. So to find that something is a product of culture is not to show that it is not natural for beings such as ourselves. Many of our cultural practices involve subjecting ourselves and each other to rigid rules. Ordinary language is a cultural construct that imposes grammatical rules on acceptable verbal behavior of human beings. The inside—outside distinction turns out to be hard to draw.

The postmodern minimalists object to the psychological properties exemplified in both ballet and modern dance, on the grounds that they are mental, not physical. This suggests that postmodern minimalists are closet dualists. A dualist would say that a figure bowed in grief exemplifies a posture from which one can infer that she is grieving, and that a trembling figure exemplifies motions from which one can infer that she is afraid. Still, grief and fear are mental, not physical properties. But according to materialism, mind and body are one. So the figure bowed in grief itself instantiates, and therefore is capable of exemplifying, grief. The trembling figure instantiates and is capable of exemplifying fear. No inference is needed.

This last point is critical. One way to accommodate the exemplification of emotional properties conveyed by dance is to say that they are metaphorically exemplified. One dancer droops, displaying a particular posture that metaphorically exemplifies grief. Another leaps, displaying a motion that metaphorically exemplifies joy. This may be the right thing to say. But the materialist position might be correct. In that case, postures and movements themselves can exemplify properties that we call mental. This posture just shows grief, that movement just shows joy. For this is what grief looks like and that is what joy looks like.

This brings us back to a question I left hanging earlier. If the performer is not experiencing the emotions, can the dance exemplify them? The first inclination is to say ‘No’, or anyway ‘Not literally’. Emotions need to be experienced, and if the dancer is not experiencing them, who is? But the assumption that emotions need to be experienced to be instantiated is not as obvious as it looks. Although emotions are associated with distinctive feelings, emotions are not feelings. It is possible to have an emotion that one does not feel. This, psychiatrists tell us, is what happens when a person is in denial. Other psychological factors override or short-circuit the connection between emotion and feeling, so that an emotion that is plainly being exhibited is not felt. It is also common in what Hume calls the calm passions—emotions like an enduring grudge or the standing affection for a life-long friend, which has no distinctive feeling associated with it, but which manifests itself in a complex disposition to behave in certain ways toward that friend (see Hume, 1928, Book II). Still, one might object, even in these cases someone has the emotion in question. Whether it is felt or not, it resides in someone.

Could we say, then, that the emotions exemplified in Night Journey reside in the dancer? Even if Martha Graham does not feel the regret and revulsion and self-loathing the dance exemplifies, those emotions are manifest in the motions her body goes through as she dances the part of Jocasta. She has them, whether or not she feels them. If they are literally characteristics of a human body, and are characteristics a human body can have whether it feels them or not, then they can be literally exemplified in the dance, by the dancer, regardless of what the dancer feels.

Similarly, of course, for other properties, like ethereality or weightlessness or enchantment. Real bodies cannot instantiate those properties. But they can seem to instantiate them. So they can appear to be ethereal or weightless or enchanted. And they can literally exemplify the appearance. I mention this possibility, not because I think I can demonstrate here that it is correct, but because it seems to me that dance raises interesting and important questions in the philosophy of mind. If the mind just is the body, then what dancers do with their bodies, they do with their minds. The full title of Rainer’s work is Trio A or The Mind Is a Muscle, Part 1. I suggest that the ‘is’ here might be close to an “is” of identity or an ‘is’ of constitution—The muscle is (at least part of) the mind.

I suggest, then, that the postmodern minimalist contention that earlier works are somehow phony or inauthentic because they go beyond what is distinctive of bodies in motion is more problematic than it looks. It involves a tacit commitment to a limited, apparently dualistic conception of the range of things that human bodies actually do. This is not to say that the postmodern minimalists were wrong to limit their range in the way that they did. Rather, it is to suggest that their understanding of the basis of the limits they set may be wrong. One of the ways dance advances understanding is by raising philosophical questions, like ‘What exactly is the relation of the body to the mind?’ and ‘How does the body in motion manifest intention, intelligence, emotion, and other putatively mental properties?’

Many of the features exemplified in human movement are exemplified elsewhere, as well. Patterns are abstract mathematical structures. So they might be exemplified in a dance, and instantiated not only in other human activities, but in events of different kinds. The dance of the snowflakes in The Nutcracker, for example, exemplifies the sort of pattern one sees when snow skitters across the ground. Patterns exemplified in dance are typically dynamic. They develop over time. So they often look like the patterns one sees in a kaleidoscope. These are complex kinematic regularities. By exemplifying such patterns, dances sensitize spectators to them, enabling those spectators to recognize them when they encounter them in other venues or to discern subtle or enigmatic aspects of them.

Dance frequently exemplifies political properties as well. Yvonne Rainer’s egalitarianism is manifest in, among other things, a democracy of the performance space. There is no mandatory center of attention. Since the actions are uninflected, all the actions of all the dancers are potentially equally significant. The dance does not impel or compel the spectator to look at one spot or another. The ‘costumes’ of the dancers (if they are not nude) are pedestrian, thereby depriving the spectator of the sort of social cues that clothing often provides. Virtuosity is neither displayed by nor required of the dancers. All, evidently, are created equal.

The collaboration of Cunningham and John Cage was, on the face of it, odd. In works like Points in Space, dance and music were created independently. The dancers typically did not even hear the music until the first performance. So rather than either art accommodating itself to the demands of the other, each stands alone. Each is worthy of attention. And in performance either may call attention to or distract attention from the other. This is not an accident; nor, from the point of view of the creators, is it a defect. For the works exemplify the autonomy of the different arts. They also exemplify the capacity of autonomy to provide an occasion for serendipitous juxtapositions.

In classical ballet, as in many other dances, music and choreography work together. Each enhances and draws on the other. The interdependence thus exemplified allows for the heightened effects that neither alone could achieve. Classical ballet also invokes the ‘star system’ that Rainer decried. It exemplifies a hierarchy, an inegalitarianism. Politically, as well as stylistically, postmodern dance and classical ballet are antithetical.

We need not endorse its political stance to understand and appreciate a dance. If my contention is correct, what we should do is understand what the dance symbolizes, and how it reflects on other aspects of our experience. Different dances display different values. They afford access and insight into the values of egalitarianism, autonomy, inegalitarianism, and interdependence. Ordinarily, we might not think of works like Trio A or The Firebird as particularly political works. But once we realize that among the properties they exemplify are political ones, we are in a position to recognize a political dimension to other actions and institutions that we might otherwise think of only apolitically. We emerge from a performance of such a work better equipped to recognize such features in other situations where we encounter them.

Interpretation and Understanding

I have been urging that dances are symbols that exemplify features and render them epistemically accessible. But how can we understand such symbols? At a very abstract level, the answer is clear. We understand dances the same way we understand other symbols—we know how to interpret the symbol systems they belong to. I can read the menu if I understand German, for the symbols on the menu are words in German. It is plausible that ballet has something like a ‘language’—a set of repeatable, conventional symbols whose interpretations are reasonably clear. But when Doris Humphrey is writhing on the floor, or Steve Paxton is carrying a mattress, or Merce Cunningham is twitching in the corner, or Jill Sigman is dancing on crutches, what are we to make of it? There does not seem anything like ballet’s regimented conventions to fall back on.

Here, I think, it is worth taking seriously what writers on dance say when they speak of a choreographer’s vocabulary or idiom. They take it that to understand a dance, we need to be able to interpret that vocabulary or idiom. There is no reason to think that this is easy or automatic. It may be that one needs to know quite a bit about what has been going on in contemporary dance in order to understand a new work. It may be that we need to know what has been done, what has been tried, and whether it succeeded or failed, to figure out what Sigman is up to. (But you also need to know German to read the menu.) If one has the requisite background, what is being symbolized may be perfectly clear. If not, one may wonder why, for example, the dancers are nude, or are wearing masks, or are evenly distributed across the stage, or whatever. One may wonder what the connection between the score and the dance is or why words rather than music constitute the score. There is no reason to think that the answers to these questions are obvious or are readily available to the novice spectator. We should no more expect to be able to interpret a dance in an unfamiliar idiom just by looking than we expect to be able, just by looking, to interpret a menu, much less a poem or a news report or a scientific paper, in a foreign language.

Confronted with an alien dance form, we may initially be bewildered. We venture hypotheses and test them to discover whether they make sense of what we are seeing. There are many modes of access, so we may find that we have relevant background resources to build on. We might, for example, appeal successfully to our knowledge of other arts or of popular culture. In some cases we will formulate plausible hypotheses that assign to a work an interpretation that accounts for the features we find salient. Then we have insights to export to other aspects of experience, and to bring to the interpretation of other works of art. The interpretation we venture may or may not stand up to further scrutiny. If, for example, it yields a reading that makes a new work anomalous, when it seems continuous with the choreographer’s previous works, we have reason to doubt its adequacy. If the insights it leads to seem banal when the work seems intriguing, then again we have reason to think our interpretation is inadequate. If, on the other hand, it makes sense of the factors we find salient and illuminates other aspects of experience, it is a prima facie plausible interpretation.

I claimed earlier that dance has the capacity to embody and convey an understanding of the wider world. I argued that this capacity is largely due to exemplification. We can now see what this involves. Dance, like other arts, exemplifies properties that are instantiated elsewhere but may fail to be noticed or properly attended to in the motley array of inputs that regularly confronts us. In exemplifying these properties, a dance draws our attention to them and stresses their significance. It thus equips us to recognize them when we see them again and intimates that we would do well to attend to them. In effect, then, dance may be a source of working hypotheses. Obviously, not all such hypotheses are sound. An insight we attempt to export from a dance may fail to illuminate anything significant about other aspects of experience. Many dances are banal. In this they do not differ from other symbols that purport to reveal things about the world. Many putatively informative conversations, and many scientific experiments, are banal, too. But when a dance is cognitively effective, it reveals something to us. We come to see the story of Oedipus in a new light, or come to appreciate the complex physical intelligence of ordinary movement, or come to recognize the precariousness of what we standardly take for granted.

In some cases we may remain bewildered. We have no idea why these people are doing those things, why anyone would consider what is going on art, or dance, or the sort of thing anyone would want to do in public. Then we can make no sense of the work. But even this may be an advance in understanding. This is Socrates’s point. Knowing that you do not know is the first step toward knowledge. Appreciating that you do not understand a work of art constitutes the first step (or at least an early step) toward figuring out how to make sense of it.

Conclusion

I have argued that dance highlights features and underscores their significance. These may be features of human behavior—mundane or extraordinary. They may be more abstract dynamic or kinematic patterns. They may embody views about weighty philosophical matters—the relation between mind and body, between autonomy and interdependence, between the ordinary and the extraordinary. Dance can deepen understanding by sensitizing us to things we tend to overlook, by undermining stereotypes, by problematizing assumptions we did not even know we were making. Since dance, like the other arts, lacks the capacity to adduce arguments, it might seem that at best, it can advance hypotheses whose merits need to be demonstrated by argument. Even if this were so, dance would have an important epistemic function in identifying hypotheses worth testing. But, I suggest, argument is not the only way to buttress epistemically tenable convictions. Examples can be equally tenable. An exemplar—even if it is a nonverbal symbol—can show that something is the case, even though it is powerless to say anything. A work of art, such as a dance or a symphony, being an extended, textured, multilayered symbol, can show something quite significant.