When a camera caught the Pittsburgh Penguins center Tyler Kennedy licking his hockey stick after a shift against the Florida Panthers, the video went viral on YouTube. The reaction of hockey players to the video was somewhat different from the simple disgust expressed by the general public: to be sure, licking your stick isn’t something you really want to be seen doing in public, but the impulse is perfectly understandable.
Among the bonds that athletes form with their equipment, that between an ice hockey player and his stick is perhaps the most intimate. He holds it in his hands roughly five times as many hours per week during the ice hockey season as the average major league baseball player holds his bat, including batting practice. It becomes very much an extension of the player’s body, as David Fleming explains in “To Each His Own,” his excellent account in ESPN Magazine that focuses on the NHL’s leading scorer in 2007–2008, Alex Ovechkin. Fleming notes that in addition to using his stick to pass and shoot, Ovechkin “uses it as a crutch to get up after big hits. He wields it high and with two hands, like a nightstick, in scrums in front of the net … He taps it against the boards to applaud a teammate after a fight or against the goaltender’s shin pads after a good save. He uses his stick to open and close the bench door. To calm his nerves late in games, Ovechkin will sit on the boards with his back to the ice and his stick in his lap, like a baby blanket, and lovingly retape the blade.” While moving at high speeds in a hostile field of overgrown bodies and sharp skates, an ice hockey player must handle a small object from the far end of a long stick—an object that is prone to both sliding and rolling. It is a game at once of violence and of finesse.
Like an amputee who modifies and fine-tunes his prosthetic, the Washington Capitals coach Bruce Boudreau recalls prepping his stick for games during his days as a player. He would “sit in his kitchen and customize the fiberglass curve of his weapon by carefully steaming it over a teakettle. Then he’d wedge it under a door hinge and bend it until it was perfect, race outside and plunge it into the snow to set the blade. With the kettle at a boil, he’d have a cup of tea while waiting for the snow to complete its work.” The knob of tape he would make on the end of his sticks had his own signature to it. Handed an old stick by a fan, Boudreau instantly recognized it as his own, from thirty-five years earlier. “Our sticks become part of our DNA,” he says. (The fan took the stick back and said he’d return it if the Capitals won the Stanley Cup.)
There is a very real sense in which a tool may be integrated into one’s body, for one who has become expert in using the tool. There is a growing number of studies that support this idea of “cognitive extension”; the new capacities added by tools and prosthetics become indistinguishable from those of the natural human body, in terms of how they are treated by the brain that organizes our actions and perceptions.1
What is it like to be Alex Ovechkin? Perhaps we can regard his case as a more highly developed version of the excellence we were trying to understand in the case of the short-order cook. In a game like hockey, it is the rules of the game that make up the jig. It is played with regulation-sized sticks, on a regulation-sized rink, but within these parameters the scope for individual finesse seems to be inexhaustible. There is pleasure in those moments when we feel a growing mastery in some small domain—mastery that is exercised through our bodies, with the use of tools. To understand this, let us consider the way our attention is structured in a skilled practice, such that a tool may be incorporated into one’s body.
Consider the experience of using a probe to explore an unseen space, or the way a blind person feels his way by tapping with a stick. At first you feel the varying pressure of the probe against your palm and fingers, and you have to interpret this pressure, mapping it in some as yet uncertain way onto a spatial representation that you are developing of the object. But as you learn to use the probe, your awareness of this pressure at the handle end is transformed into something quite different. What you have eventually is a direct, unmediated sense of the probe’s tip touching the objects you are exploring. You are no longer attending to the sensations in your hand.
The philosopher Michael Polanyi analyzed this moment when we achieve competence with a probe, and in doing so found that he had to use the word “attend” in a new formulation: you are now “attending from” the sensations in your hand to the objects at the probe’s tip; the sensations themselves you are only “subsidiarily aware of.” In this way “an interpretive effort transposes meaningless feelings into meaningful ones, and places these at some distance [i.e., the length of the probe] from the original ones. We become aware of the feelings in our hands in terms of their meaning … to which we are attending.” This meaning is located at the tip of the probe. The probe itself becomes transparent; it disappears. You are no longer engaged in an interpretive effort. The crucial fact that makes this integration of a prosthetic possible is that there is a closed loop between action and perception: what you perceive is determined by what you do, just as when we make use of our own hands. Let us go a little deeper into the mutual entanglement of action and perception.
EMBODIED PERCEPTION
Consider a blossoming crape myrtle tree that I see in my backyard. How does it present itself? What I actually see is one side, one profile, of what I take to be a three-dimensional tree. Though I have never paid it much mind until now, I suppose I have seen the back side of the tree many times in the past, and I can make a point of going around and looking at it later, but while I am looking at the tree now its back side is not perceptually given. Yet my apprehension of the tree as a three-dimensional whole has an immediate character to it. It doesn’t feel like a recollection, or a consideration of hypothetical possibilities in the future. I “see” (in some sense that we need to investigate) the whole tree at once. In everyday life (as opposed to a freshman philosophy class), I certainly don’t worry that the back side of the tree is not there when I can’t see it.2
These facts of experience present something of a puzzle if we understand perception simply as the brain’s response to stimulation. To preserve that view, one has to posit some mechanism by which the discrete perceptual profiles of the tree somehow get integrated into a whole by the brain; there must be some kind of “processing” that leads to a “representation” of the tree in the mind.3 The basic supposition in this, the standard view, is that vision may be understood by analogy with a still photograph. The more rounded-out sense we have of the world is then taken to be the product of a dynamic 3-D modeling that our brains do, just like the software used by animation studios.
An alternative to this approach begins from a different set of facts: our brains are connected to eyes that are free to move within their sockets, located in a head that swivels on a neck, attached to a body that moves around on the ground, in ways that are characteristic of the particular sort of bipedal animals that we are. According to a school of thought that has been gaining traction in the last fifteen years, these facts—our embodiment, and the possibility of movement that our bodies provide—are no mere accessory to perception, but rather constitutive of the way we perceive. As one researcher puts it, “Perceiving is a way of acting. Perception is not something that happens to us, or in us. It is something we do.”4
James J. Gibson spent his early career as a psychologist assessing the aptitude of candidates for pilot training during World War II. Over the course of several decades’ research on visual perception, he began to chafe against the fundamental assumption that “sensory inputs are converted into perceptions by operations of the mind.” His 1979 book The Ecological Approach to Visual Perception announced a subtle but important reconception of vision, not as the purely mental processing of sensory inputs, but rather as an activity in which we use our body to “extract invariants from the stimulus flux.” In other words, one has to be able to explore a scene from different perspectives to perceive what remains the same about it—its nature and structure, regardless of perspective—and locomotion is an indispensable part of this process.5
Further, there is evidence that only self-motion accomplishes this; the visual system cannot develop if one is merely transported around, passively. In one of the earliest experiments in what would come to be called embodied cognition, ten pairs of kittens were reared in the dark, except for three hours per day that they spent in a carousel apparatus that allowed one twin to move freely, while the other was carried passively by the movements of the first. The active kitten could move up, down, away from, or toward the center of the carousel, as well as rotate in epicycles at the periphery of the carousel’s radius. The kittens could not see each other, and the surrounding environment was contrived so that both kittens received identical visual stimulation as they moved around; the only difference was that one moved itself, the other was passively carried. The active kittens developed normally; the passive kittens failed to develop visually guided paw placement, avoidance of a visual cliff, a blink response to quickly approaching objects, or visual pursuit of a moving object.
The still photograph turns out to be a poor metaphor for understanding visual perception, for the simple reason that the world is not still, nor are we in relation to it.6 This has far-reaching consequences, because some foundational concepts of standard cognitive psychology are predicated on the assumption that we can understand the eye by analogy with a camera, in isolation from the rest of the body. Nor is this a mere intramural fight between quarreling academic camps; what is at issue is the question of how we make contact with the world beyond our heads.
In the domain of visual perception, cognitive psychology set out to solve a certain puzzle: an indefinite variety of three-dimensional objects can project identical two-dimensional shapes on the retina of an observer. If static optical information is all that is available to the subject, then because such information underspecifies the shapes of surfaces, it follows that it must be supplemented with something else; something going on inside the head of the subject—namely, assumptions about the structure of the world. This is the motivation for thinking that perception involves an inferential process in the brain. This inference is taken to be computational. That is, “cognition consists in the manipulation of symbols, where these manipulations often involve the application of rules for the purpose of deriving conclusions that go beyond the information [that is presented to the eye],” as Lawrence Shapiro writes in Embodied Cognition, his excellent overview of the embodied cognition literature.7
A single retinal image is certainly not adequate to the task of specifying the world, but the visual stimulus received over time by an observer in motion is adequate, Gibson argues, and so on his account the whole motivation for conceiving perception as involving inference and computation collapses. This is completely revolutionary.8 The brain does not have to construct a representation of the world. The world is known to us because we live and act in it, and accumulate experience.
Surprisingly, it is in the field of robotics that some of the most convincing evidence has emerged that inference, calculation, and representation are a grossly inefficient way to go about negotiating a physical environment. In his now-classic article “Intelligence Without Representation,” published in the journal Artificial Intelligence in 1991, Rodney Brooks wrote that “the world is its own best model.” Roboticists are learning a lesson that evolution learned long ago, namely, that the task of solving problems needn’t be accomplished solely by the brain, but can be distributed among the brain, the body, and the world.
Consider the problem of catching a fly ball. According to the standard view, we might suppose that the visual system provides inputs about the current position of the ball, and a separate processor (the brain) predicts its future trajectory. How we might do this is a bit mysterious, given that most of us wouldn’t be able to calculate such a trajectory consciously, with pencil and paper. The Gibsonian approach suggests we don’t need to do any such thing, whether consciously or subconsciously. And in fact what we do, it turns out, is run in such a way that the image of the ball appears to move in a straight line, at constant speed, against the visual background.9 It so happens that finding and exploiting this invariant, which is available in the optic flow if you run just right, puts you in the right spot to catch the ball. (The same strategy appears to be used by dogs who catch Frisbees, even on windy days.) You don’t need an inner model of the pseudo-parabolic trajectories that baseballs follow, with corrections for air resistance at different altitudes and so forth. It’s a good thing, too.
We think through the body. The fundamental contribution of this school of psychological research is that it puts the mind back in the world, where it belongs, after several centuries of being locked within our heads. The boundary of our cognitive processes cannot be cleanly drawn at the outer surface of our skulls, or indeed of our bodies more generally. They are, in a sense, distributed in the world that we act in.
A good way to capture the force of this conceptual shift is to compare a humanoid robot designed on the principles of “good old-fashioned artificial intelligence” with a robot that reflects the new ecological thinking. Andy Clark has done just this (and much more) in his hugely illuminating book Supersizing the Mind.
Walking may be understood as a kind of problem solving. Asimo, a biped robot built by Honda, relies on precise control of its joint angles by motors, servos, and other mechanical actuators, which enables it to negotiate such challenges as a flight of stairs. The amount of energy Asimo requires to shift a given unit of weight a given distance is about sixteen times greater than that which a human body requires. Commenting on this general approach to building robots, one roboticist remarked that a specimen like this isn’t very pleasing to watch; it suffers from “a kind of rigor mortis.”
The efficiency of a human body has been matched, however, by a very different kind of contraption, one that relies on its own “passive dynamics.” A passive-dynamic walker is powered only by gravity, somewhat like the old toy Slinky walking down the stairs. A walking robot with legs and knees and swinging arms developed in 2001 requires only a slight downward slope. It walks smoothly, with an uncannily human gait. There is no control system. Its movement is not the result of movement planning and calculation, based on some representation of the world, but is rather a function of its form: the lengths of its limbs, their weights, and the damping and spring rates of the joints connecting them, much as muscles and ligaments connect the limbs of a human body. A powered robot can exploit the same design principles.
This is an instance of “ecological control,” or “morphological computation,” in which “goals are not achieved by micromanaging every detail of the desired action or response but by making the most of robust, reliable sources of relevant order in the bodily or worldly environment of the controller,” as Clark writes. The “processing,” as it were, is partially taken over by the dynamics that are inherent in the interaction between the robot and its environment.
A toddler begins to walk by learning to exploit the passive dynamics of his own body. Initially his body (which of course is growing and changing) is experienced as a beginner experiences a hockey stick; it is obtrusive and frustrating. The infant learns through exploration “which neural commands bring about which bodily effects,” and with enough practice he becomes “skilled enough to issue those commands without conscious effort.”10 At that point the child’s body has become transparent in the same sense that a blind man’s probe becomes transparent; it disappears as an object of attention unless something goes wrong with it. In Polanyi’s terms, the child is now attending through his body to the world beyond. He feels a growing mastery.
Friedrich Nietzsche said that joy is the feeling of one’s power increasing. This needn’t be understood as the motto of an insatiable tyrant. It captures something important about the role that skill plays in a good life. When we become competent in some skilled action, the very elements of the world that were initially sources of frustration become elements of a self that has expanded, by analogy with the way a toddler expands into his own body and comes to inhabit it comfortably. And this feels good.
RIDING MOTORCYCLES: GYROSCOPIC MAN
We remain capable of learning new skills throughout life. Evolution has endowed us not with a fixed design and scripted behaviors, adapted to cope with a particular environment (securing food and mates in the Pleistocene savannahs, as the evolutionary psychologists keep insisting), but rather with highly plastic neural resources and “an ongoing regime of monitoring and recalibration,” as Clark says.11 When we begin to learn a new skill, new equipment often mediates between our bodies and the world, and the loop of perception through action may take a detour through physical phenomena that are quite alien to the natural human body. In that case, we find ourselves returned to the condition of being like a toddler, figuring out how to maneuver ourselves through the world.
The steering dynamics of a motorcycle are a subtle and astonishing thing. At higher speeds, to make a motorcycle initiate a turn to the left, you apply pressure as though you were trying to turn the handlebars to the right. Motorcyclists call this counter-steering, and it is indeed counterintuitive. Turning the handlebars briefly to the right makes the bike lean to the left because of gyroscopic precession, and it is the leaning that accomplishes the turning. (The reader may recall a classroom demonstration of gyroscopic precession involving a spinning bicycle wheel that one holds by its axle, while seated on a chair that is free to swivel. Tilting the axle—that is, trying to rotate it on a horizontal axis perpendicular to the axle itself—is hard to do, and makes the chair you are sitting in swivel on a vertical axis. It is a very weird experience; there seem to be demonic forces at work.)
In cornering a motorcycle there is a series of motions and exertions that get installed in muscle memory through practice, and these are integrated with the visual cues of cornering. Once the integration is fairly secure, it is the visual cues that the motorcyclist attends to, not the muscular exertions. The higher the speed, the more intensely they are attended to; the level of concentration involved in motorcycle road racing is truly impressive, as will be appreciated by anyone who has seen a picture of a rider at full lean through a corner. You will see his inside knee, sometimes even inside elbow, scraping the ground (racing leathers have plastic pucks on the knees so they don’t get torn up). The next time you see such a photo, look for the rider’s eyes. If they are visible through the helmet’s visor, you will see them directed nearly perpendicular to the bike’s direction of travel, as the rider looks all the way through the corner.
This brings up another uncanny fact about motorcycle steering: the bike goes wherever your gaze is focused. Most important, if your eyes lock on some hazard in the road, you will surely hit it. This is not a superstitious motorcyclist’s version of Murphy’s Law; it is a reliable fact, and it reveals something deep about the “intentionality” of our prereflective sensorimotor negotiation of the world. Inhabiting the kind of bodies that we do, our gaze and our locomotion are connected in ways that work for us, and we don’t have to think about it. But this accomplished integration becomes a liability when riding a motorcycle, and must be deliberately short-circuited. You have to learn to unlock your eyes as quickly as possible from every hazard, and instead look where you want to go.
This visual demand is absolutely counterintuitive. When walking, we move away from a hazard (for example a snarling dog) while keeping it in view. Our action programs, visual system, and “affect” (immediate, visceral judgments of good or bad such as happens when we see a spider) are integrated in a way that is adaptive for us, and have achieved a certain automaticity. But when the relation of your body to the world is mediated by a machine, one that requires a very different set of muscle responses to achieve the desired avoidance, then you aren’t well adapted until you have reintegrated muscle response, visual system, and affect into a very different collection of automated responses. At the heart of this learning process for the motorcyclist is a phenomenon utterly unknown to the natural human body, namely gyroscopic precession.
Gibson’s work sheds light on all this. He suggested that the concept of an “ecological niche” is necessary to properly understand perception. A niche is not quite the same as a habitat. A niche “refers more to how an animal lives than to where it lives.”12 It is not simply the physical surroundings, but the aspects of those surroundings that are meaningful for an animal given its way of life.13 When you live on two wheels, gyroscopic precession is as important a feature of your ecological niche as gravity.
Gibson’s most interesting and controversial point is that what we perceive, in everyday life, is not pure objects of the sort a disinterested observer would perceive, but rather “affordances.” The affordances of the environment are “what it offers the animal, what it provides or furnishes, either for good or ill.” Affordances elicit and guide action; Gibson suggests they also organize perception. Things in our environment show up in the vivid colors of good (for a motorcyclist: a median strip with a curb that is low enough that it could serve as an escape route if things get hairy in front of you) and bad (the oily patches that are usually present in the center of a lane at an intersection, where cars drip fluids while idling). As Alva Noë puts it, “When we perceive, we perceive in an idiom of possibilities for movement.”14
Our perception of these possibilities depends not only on the environmental situation, but also on a person’s skill set. A martial artist faced with a belligerent man at a bar sees the way the man is standing, and his distance, as affording certain strikes and foreclosing others, should it become necessary.15 Because of long practice and habituation, when he looks at the man’s stance, this is what he sees. He may also perceive the furniture nearby, and the objects lying within reach on the bar, in terms of their affordances for combat. He sees things that people like you and me don’t. That’s why we shouldn’t mess with him.
Affordances lie in the fit between an actor and his or her environment. When that relationship is mediated by a prosthetic, such as a motorcycle, it changes the field of objects that we perceive and how we perceive them. Gibson considered only the natural human body when he investigated the ways movement through our environment influences our perception. But his idea of affordances provides a useful foundation for thinking about culture and technology—that is, for thinking about the distinctly human ecological niches that we create for ourselves. This becomes important for my concept of the situated self, and we will develop it further in the chapter “Encountering Things with Other People.” But for now let us stay focused on motorcycling and try to explicate this one instance of a specialized human ecological niche in fuller detail.
In addition to gyroscopic precession, a further “unnatural” challenge in motorcycling is the categorically different rate at which you are moving toward the things in your visual field, compared with our usual bipedal locomotion. This makes it imperative to keep one’s eyes fluid. One has a certain amount of time, typical of the particular environment one inhabits, to judge a hazard and respond accordingly. A ship’s captain on the open sea has a lot more time to avoid another ship than an alligator wrestler has to read the signs of his opponent’s impending move. At high speeds these “judgments” (if that is the right word) must be very fast. They cannot be the result of a conscious inferential process (just as they are not when we are running to catch a fly ball), because inference is a slow and cognitively costly activity. The subconscious integration of sensorimotor data that one is performing while riding a motorcycle at high speeds requires a great deal of concentration, but not a lot of articulate thinking.
Those data are inextricably bound up with a host of mechanical contingencies. A motorcyclist feels the road through his tires, and racers are very particular about the air pressure, the cross-sectional shape, and the particular rubber compound used. A constant curvature in the tire’s cross section will make the bike feel linear in its response to the rider’s leanings. A flatter cross section or “profile” will produce a larger contact patch while the motorcyclist is riding upright in a straight line (thus increasing braking power at the end of a straightaway, which is important because then the rider can initiate braking later), while a peakier profile will have a larger contact patch while the cyclist is leaned over in the turns. Choice of tire profiles therefore depends somewhat on the track and a team’s race strategy. Different rubber compounds break loose with more or less abruptness, which influences the rate at which the rider can apply throttle coming out of a corner (in this situation traction, not horsepower, is the limiting factor; in general, traction is “conserved” and must be judiciously allocated among lean angle, acceleration, and braking). Some tires communicate better than others. The rider also feels the road through the suspension, which on race bikes is separately adjustable for compression and rebound damping, as well as effective spring rate. Some racers are particular even about the steering head bearings, claiming they can feel the difference between tapered roller bearings and ball bearings. Needless to say, a racer does not attend to any of these elements while leaned over at 130 miles an hour in a corner, knee on the ground, separated by a few feet from other riders in a pack. (Such is the mutual trust of skilled professionals.) The mechanical contingencies of traction and gyroscopic precession become second nature, and are given no more thought than the hockey player gives to his stick while he is playing, or the blind man gives to his cane.
The philosopher Adrian Cussins writes about two different ways of knowing about speed.16 He relates the experience of riding his motorcycle around London, adjusting his speed in response to various weather and traffic conditions, and contrasts this with the way one knows one’s speed when one reads it off a dial or digital readout. In that case, speed is rendered as a number, and to learn the significance of this number one has to compare it with another number: the posted speed limit. But this “significance” is of a much thinner kind. Cussins’s point is that knowing one’s speed in this second way is to render speed as a proposition: I know that I am going forty-five miles per hour. This is a fact—the sort that has objective validity, but is divorced from the particular driving situation in which the motorcyclist finds himself. “Forty-five miles per hour” is not speed, it is a representation of speed. It has the virtue of standardization; it is the kind of fact that is transportable “from one embodied and environmentally specific situation to another.”
These two ways of knowing about speed “are taken up in very different, sometimes competing, cognitive orientations to the world,” as Cussins puts it. When the objective representation of speed interposes itself between the motorcyclist and his perception of his situation, it can interfere with his direct world-inhabiting. Cussins writes that “the great advantage of experiential content is that its links to action are direct, and do not need to be mediated by time-consuming—and activity-distancing—inferential work.”
If Cussins is right, reliance on a speedometer tends to subtly bump us out of a skillful way of driving, and this is due to the interference of objective knowledge with experiential knowledge. But Cussins doesn’t elaborate how this interference might happen. Following some clues in the cognitive science literature, I’d like to suggest that the interference is due to a substitution that occurs, wherein the symbolic representation of speed becomes an object of attention, displacing somewhat the ecological, sensorimotor experience of speed. Crucially, unlike the ecological experience, the symbolic representation of speed is “affect-neutral” (it isn’t scary), so it doesn’t prime the action programs (here, evasive maneuvers) that for an experienced rider have become integrated with his threat perception and have achieved a certain level of automaticity.
The negative affordances that a motorcyclist sees aren’t limited to things like oily spots on the pavement. The road is, after all, a social place. You do well to notice the brunette in the short skirt standing at the intersection, because the guy driving the car in front of you may slam on the brakes. But the old lady following closely in the car behind you won’t. You have been watching the old lady with interest. As far as you can make out in your vibrating mirror, she has a look of sour disapproval on her face, and it is directed at your taillight. You also notice the driver of the delivery truck that has just appeared at a side street. He appears to be laughing: he is engaged in a conversation with someone else in the cab of the truck, seated on the side opposite to the side that you are approaching from. The driver looks like just the slovenly sort who would pull out without double-checking. (Motorcyclists become ethnographers of necessity, or rather rank stereotypers, for the same reason that cops do: they face risk. Stereotyping is efficient for snap judgments.) Having been scared many times in the past, you are attuned to the kind of information that is important when riding in urban areas, and this information is different in kind from your instrument-read speed.
I suspect overreliance on the speedometer may slacken the bonds between action and perception, and indeed there is evidence that when our attention is diverted to symbolic representations, it can have such a loosening effect. This is what appears to be going on with a chimpanzee named Sheba who has been taught numerals. Andy Clark tells her story:
Sheba sits with Sarah (another chimp), and two plates of treats are shown. What Sheba points to, Sarah gets. Sheba always points to the greater pile, thus getting less. She visibly hates this result but can’t seem to improve. However, when the treats arrive in containers with a cover bearing numerals on top, the spell is broken, and Sheba points to the smaller number, thus gaining more treats.
The interpretation is that the numerals, because they don’t look tasty, “allow the chimps to sidestep the capture of their own behavior by ecologically specific, fast-and-frugal subroutines.” They provide “a new target for selective attention and a new fulcrum for the control of action.”17
Abstracting from the concrete objects of their environment, the chimps become better at maximizing their utility. They become more Protestant, we might say: to get maximum treats in the future requires a bit of asceticism for the moment, and this becomes possible if you redirect your attention to something abstract, such as money (for the Protestant) or numerals (for the chimp). The abstract thing becomes “a new fulcrum for the control of action,” as Clark says. Thus is born civilization.
But when it comes to motorcycling, “ecologically specific, fast-and-frugal subroutines” are mostly a good thing, because everything happens very fast.18 Let us call them perception-action circuits. Just as reaching is triggered by the sight of something tasty for chimps and other non-Protestants, the solicitations of the motorcycling-specific environment trigger steering and other control inputs for the rider. These circuits are tied to affect: the kind of response you have to the sight of something tasty or something dangerous. In the case of a learned skill, these perception-action-affect circuits represent an achieved integration, and serve as the foundation for fluid, relatively effortless performance.
Is there a role for explicit thinking in this kind of performance?
THE ROLE OF LANGUAGE IN ACQUIRING SKILL UNDER CONDITIONS OF RISK
Among philosophers there is currently a quarrel about what role (if any) concepts—the kind you can state in language—play in skilled activities.19 On one side is Hubert Dreyfus, who says that when we are engaged in an activity that we are already competent in, anything so thoroughly mental as a “concept” doesn’t normally play a role. It can only get in the way and disrupt our “smooth coping.” By “smooth coping” he means a way of acting where our responses to the things we are dealing with are elicited from us by the situation, without articulate thought. This is how you tie your shoes in the morning, for example. It is a skill you learned long ago, and has become automatic. Dreyfus offers this idea as a corrective to the view that our actions are always caused by prior “mental” operations.
On the other side is John McDowell, who offers what I take to be an important countercorrective to the “smooth coping” notion by emphasizing the role of concepts in skilled activity. We don’t shut off our thinking, the way Dreyfus seems to suggest. Though McDowell doesn’t mention the fact, I believe his emphasis on the role of conceptual thinking is especially necessary when the activity in question is dangerous, and there are contingencies that remain beyond your control no matter how skilled you are, as in motorcycling. (Motorcycling is different from tying your shoes in this regard.) We needn’t enter into the minutiae of this academic debate, only know that it hovers in the background of what follows.
Bernt Spiegel is a German automotive psychologist who spent much of his career consulting for Porsche. He was also a motorcycle road race instructor. His book The Upper Half of the Motorcycle, an unobtrusive masterpiece, makes no reference to the recent literatures on embodied cognition or “nonconceptual mental content,” but is overflowing with observations that contribute to these lines of inquiry. He writes, “One simply has to know about some situations before behavior can be adapted on the basis of this knowledge.”20 He gives the example of seeing your own shadow directly in front of the bike on the road. This happens to occur at just those times when you have excellent visibility ahead, and therefore feel especially relaxed. But the sight of your shadow should trigger an alarm: oncoming traffic or motorists waiting to cross in front of you likely cannot see you. It doesn’t feel dangerous; you have to have a theory about the situation, as it were. That is, to ride safely you have to actively summon a bit of knowledge that can be stated as a proposition, using language. Initially, this knowledge is abstract and obtrusive; it interferes with your flow. But this theoretical knowledge becomes integrated with your perception-action-affect circuits after you have had a number of close calls. (This is my own assertion, not Spiegel’s.)
When you nearly get creamed on the road, there is an intense physiological reaction that occurs. I experience it as my stomach leaping up toward my solar plexus, and this happens immediately, even as the event is occurring. It is followed soon afterward by a shaking that is presumably caused by adrenaline, and extreme weakness in my throttle hand, which is sometimes debilitating. This has a powerful effect: an association between the pertinent features of the situation you just lived through and DANGER gets burned into the circuits of embodied cognition that you rely on while riding. Suppose one feature of this situation is the sight of your own shadow in front of you. After a close call, the association between that sight and bodily danger becomes not a mere proposition, but something you really know and feel. No inference is necessary.
But I suspect this integration occurs only if you have the proposition in hand ahead of time, so that you have identified the sight of your own shadow in front of you as an element of “the situation.” There are an indefinite number of true facts that could be stated about the circumstances leading up to your close call: the color of the car that almost hits you, the phase of the moon, and so on ad infinitum. The role of attention in the exercise of a skill is to pick out those features of the scene that are pragmatically significant and that therefore, taken together, define “the situation.” In the early stages of learning a skill, explicit propositional knowledge, stated in language for instructional purposes, plays a crucial role in directing attention toward its proper objects. Bathed with attention in this way, these objects then become available for integration with affect and action routines by the subconscious mind, following a close call.
Note that the role played by language implies that achieving competence—even in an activity as solitary as motorcycling—has an important social dimension. You learn things from others by reading books, having conversations with other practitioners, or watching tutorials on YouTube. Gary Klein has famously studied the decision making of firefighters, and discusses their ability to discern when a building is about to collapse, allowing them to get out in the nick of time. Klein emphasizes their ability to integrate subtle sensual data and recognize patterns. But as far as I know he does not address the role of affect, nor the role that language plays (on my account) in priming the integration of affect with perception and action. Presumably, the more experienced firefighters try, at least, to describe to a beginner the peculiar sequence and combination of sights, sounds, and feelings that precede a building collapse, so that the beginner knows what to be alert to in the chaos of a burning building. If my hypothesis is correct, such explicit instruction would be an important preparation for the establishment of those perception-action-affect circuits which, once integrated, become the basis for high-level performance.
The cohesiveness and ongoing association of a firefighting unit offer an advantage not enjoyed by most motorcyclists: they are under mutual surveillance and can criticize one another’s mistakes. They can also cover one another’s blind spots, offering up a third-person perspective such as “There was a large ember floating upward right behind you as you exited that room. You got lucky.” Such facts, conveyed by a colleague, can become material for a firefighter’s retrospective understanding of “the situation,” or indeed a collaborative reconstruction of it. His own experience is altered in conversation.
That is, the array of sensual data that count as pragmatically relevant for grasping the situation may be expanded or shifted through a kind of triangulation with others who were not merely there, but engaged in the same task, facing the same dangers. You debrief one another. The fruit of this conversation enters into your ongoing rehearsal of the experience. If this rehearsed version bears up, and jibes with further experience, it becomes internalized, available to the subconscious mind in coping with future situations. For experiences to become part of the secure, sedimented foundation of a skill, they must be criticized. Other people (and the resources of language) are indispensable. Without them, your experiences are partial, and may sediment as idiosyncratic bad habits.
The power of these conversations to clarify your experience, rather than introduce fresh confusions, depends in part on the dialectical abilities of your colleagues. They have to be able to interrogate their own experience of the fire critically, and bring their experience into the conversation in such a way that their initial interpretation of it is put at risk. They have to be capable of offering it up, without undue attachment, to the shared enterprise of trying to understand structure fires. In other words, they must have the art of philosophical conversation (which is a kind of moral accomplishment). I believe the most competent people in any field do have this art to some degree, though they probably wouldn’t name it as such.
Getting things right requires triangulating with other people. Psychologists therefore would do well to ask whether “metacognition” (thinking critically about your own thinking) is at bottom a social phenomenon. It typically happens in conversation—not idle chitchat, but the kind that aims to get to the bottom of things. I call this an “art” because it requires both tact and doggedness. And I call it a moral accomplishment because to be good at this kind of conversation you have to love the truth more than you love your own current state of understanding. This is, of course, an unusual priority to have, which may help to account for the rarity of real mastery in any pursuit.
Another reason to remain self-critical while motorcycling is that sometimes there are features of a situation that present no sensual cues at all, for example the patch of gravel on the road around the blind curve you are just now leaning into. When you are dealing with potential hazards that aren’t present to perception, you have to actively form hypotheses about bad contingencies, and project them out into the world. Doing so primes the appropriate action plans, and makes them more quickly accessible. Imagining what could happen is an important role for the conscious mind, so it must stay involved. Being in a state of “flow” without such worries sometimes makes you feel like Superman, but it is easy to flow yourself right into the truck that has drifted into your lane around the blind curve ahead.
Isn’t it true in general that life is shot through with hidden contingencies? Risk is present in any activity that is directed to some goal—the risk of failure. Unlike animals that live in the moment, guided by instinct, we are constantly monitoring our own performance, second-guessing it, tuning it up.
Spiegel points out that when riding with limited visibility (which is usually the case), we tend to ride according to a “risk composite.” We are vaguely aware of some really bad contingencies, but we also know that the odds of these actually occurring are small. Our response, as folk statisticians, is to slow down a little bit, as though we are weighting the bad contingency by multiplying it with its unlikelihood, and thereby arriving at the appropriate speed. But this is self-delusion. Our reduced speed is nowhere near as slow as it would need to be to avoid the hazard if it should occur. (Then again, without self-delusion we’d probably never enjoy ourselves.)
Riding in the controlled environment of a racetrack is fundamentally different from riding on the street, in terms of the mental practice it involves. I tagged along with the MotoGP tour through its first two races of the 2013 season (in Doha, Qatar, and Austin, Texas) and talked to some of the riders. On the track, to be competitive one has to have complete faith in the mental image one has of a corner. If there is a rise in the track, blocking your view through the turn (for example Turn One at Austin’s Circuit of the Americas), this mental image of the corner will extend beyond what is perceptually present when you initiate the turn and commit to a certain line. Your mental image is based on repetition (you go around the same track many times in practice sessions) and is assumed to be reliable. In the event of a disrupting hazard, there are corner workers who wave yellow flags and position themselves where you can see them early. These corner workers serve to relieve the conscious mind of its burden of actively positing hypotheses about bad contingencies. A masterful road racer, thus relieved, takes the art of motorcycling to its highest level. It is beautiful to watch, and forces one to recalibrate one’s sense of what human beings are capable of.
On the street, riding like Marc Márquez (the current MotoGP champion) is an ideal that is in fruitful tension with the demand for continued vigilance by the conscious mind. Managing this tension is itself an art. As Spiegel puts it, the role for the conscious mind is “alert watchfulness, without meddling.” It is “an unstable condition, which degrades all too easily into either a complete lack of watchfulness or too much involvement.” When this mental practice is lived, it doesn’t manifest as something beautiful for a bystander to behold, as road racing does. Rather, it becomes impressive only as a cumulative accomplishment in the life of the rider, measured in miles ridden without incident but including also some all-too-brief stretches of Márquez-esque transcendence.
This “alert watchfulness without meddling” by the conscious mind while one is riding on the street often takes the form of hunches: hypotheses about what might happen that are conscious but not fully articulate, because they don’t need to be. You recognize a familiar situation: there are strip malls on either side of a major thoroughfare, each with entries to the main road. The street numbers are posted only erratically, on haphazard buildings set far back from the main road. The car in front of you slows down, then speeds up, repeatedly. Hypothesis: this person is looking for a particular business, and when he spots it he may quickly veer across two lanes to get to it. Your motor responses are cocked and loaded, as it were, because you recognize the pattern.
Recall the old story about the prisoners who have been together so long they know all the jokes in circulation and have numbered them. Eventually, the telling of a particular joke consists of simply yelling out “seven,” at which point everybody cracks up.
Similarly, when an experienced motorcyclist rides through a zone of strip malls, the descriptive complexity of the scene has been reduced to a type for him. When I first started riding, I remember being tense in such situations, besieged by unpredictability and an overwhelming variety of data that needed to be monitored. At some point my vigilance became more relaxed, more enjoyable, and also more effective. I noticed that I was riding with my left thumb poised on the horn button. My right hand is indexed to the front brake lever by the two fingers that rest lightly on it. My upper body is soft, primed for quick steering inputs. This riding posture is a kind of hypothesis, a provisional understanding of what will occur, installed in my body and ready to be deployed. Once this hypothesis/posture is set, it seems to free me up to ride loosely, without being overly taxed cognitively.
I have a little verbal formula that I say out loud when I enter such riding situations: “They want to kill you.”21 (Wearing a full-face helmet, you can hear your own voice booming over the road noise.) Similarly, road racers sometimes put stickers with verbal phrases on their gas tanks, such as “Look deep into corners” and “Use a late apex.” Such mottos are taught to the beginner, and for him they have the status of precepts. Why would an expert continue to use them? Andy Clark writes that even for experts, “verbal rehearsal supports a kind of perceptual restructuring via the controlled disposition of attention…”22 We use these verbal prompts to maintain performance in challenging circumstances.
Sometimes a verbal expression will serve as a “tag” for a particular feeling that an expert has become attuned to, an elusive state that he wishes to achieve. In these cases, the phrase will be mysterious when used as an instructional device for the novice. Spiegel gives the example of feeling your consciousness “flowing down through the contact patches.” (Once again, these are the patches of rubber where the tires make contact with the road; they change shape with different lean angles.) The formula may sound like mysticism, but is meant to capture what it feels like when your riding is totally dialed in and the bike has been incorporated into your extended body. Similarly, a pianist reports initially being frustrated by his teacher’s instructions to “get the time into the fingers” or use “jazz hands.” At first this is unhelpful, but eventually, “what seemed like just vague words to the novice has now become very detailed practical talk, a shorthand compendium of ‘caretaking practices’ for toning and reshaping the grooved routines.”23
This drive to continually tone and shape up a skill is lost sight of if we take tying one’s shoes as the paradigm of skilled action. That is an activity for which we adopt a “sufficing” standard: Is the shoe tied or not? Being able to tie your shoe is a secure accomplishment, a state of stasis. But in activities that we take seriously, such as music and sports and going fast, we strive for excellence. Unlike animals that live in the moment and merely cope with their world (however smoothly), we are erotic: we are drawn out of our present selves toward some more skilled future self that we emulate. What it means to be erotic is that we are never fully at home in the world. We are always “on our way.” Or perhaps we should say that this state of being on our way to somewhere else is our peculiar human way of being here in the world.
* * *
We have considered how perception is intimately bound up with action, and what it is like to attend to the world through implements that get incorporated into our consciousness. The boundaries of the self seem to expand. As we push out into the world, first as toddlers and then as tool users, we perceive it differently because we are now inhabiting it in a more determinate way, conditioned by the particulars of the skill and the implements we use. Through a skilled practice, the self has been brought into a relation of fit to the world. And this can be quite absorbing.
To emphasize the role that our bodies play in determining how we inhabit and therefore perceive the world, and to entertain the notion of cognitive extension, is to put oneself on a collision course with the central tenets of the official anthropology of the West. As we have already noted, embodied perception poses a direct challenge to the idea that representation is the fundamental mental process by which we apprehend the world. But these developments in psychology also pose a challenge to contemporary ethics. And this is no mere coincidence. In ethics, as in epistemology, the idea of representation is central to our Enlightenment worldview. Immanuel Kant insisted that in order to avoid special pleading, in moral reasoning we should regard ourselves and others not as individuals but as representatives of the genus “rational being,” and approach one another through the filter of this abstraction. In Part II, “Other People,” we will consider the flavor that this abstraction imparts to our interactions with one another, and contrast it with the kind of interaction that is made possible by an “ethics of attention” in which we are alive to the concrete particularity of others.
We have more work to do to establish what it means to encounter things, and the next few chapters will remain focused on this. But I want to start building the bridge that will get us from things to people, and from epistemology to ethics. This bridge begins to take shape when we notice that the technological concept of virtual reality also expresses a moral ideal. More precisely, it expresses what philosophers call a metaethical position. That is, it carries with it a certain understanding of the underpinnings of ethics: a picture of the moral agent and how she stands in relation to the world beyond her head. Taken in this sense, virtual reality provides an especially clear point of contrast to the concept of the situated self that I am offering.