Friday, 10 February 2017
You Don’t Catch a Ball by Calculating Its Trajectory, You Catch It by Moving
Today I’d like to talk about a problem that is a classic both for baseball players and for cognitive scientists. And the way that baseball players solve it has helped cognitive scientists to better understand the important role that the body plays in cognition.
The problem is as follows: how does a baseball player go about catching a baseball that has been hit high into the air, especially when the player is in centre field and the ball is following a long, parabolic trajectory that would otherwise cause it to land several metres from where the player is standing? How does the player go about calculating this trajectory and moving, in just a few seconds, to the right place to catch the ball? This is what has long been known in English as “the outfielder problem.” (If you’re more of a soccer fan, imagine a backfielder successfully heading a long throw-in by the goalkeeper.)
As some of you may know from high-school physics, if you know the ball’s speed as it leaves the hitter’s bat, and the angle of the ball’s trajectory relative to the ground, then given the effects of gravity on this trajectory (and ignoring, for simplicity, the effects of wind and air friction), you can calculate the spot where the ball will fall. But the player out there in deep centre field has no way of knowing these values, much less having the time to plug them into the right mathematical formulas to solve the problem and move to where the ball should land.
So how does the centre fielder position himself at the right spot? By a very simple trick: he moves so that the ball always stays at the same position in the sky, from his point of view! If he sees the ball rising, he keeps dropping back until it stops doing so. Once he sees it begin to fall, he keeps moving toward just quickly enough to keep it in the centre of his field of vision. Similarly, if the ball moves a bit to the left, the player moves a bit to the left, until the ball appears to have stopped moving leftward, and vice versa if the ball moves to the right. And in the final fractions of a second, if the player is in the right place, all he has to do is stretch out his glove to that spot in the centre of his field of vision where the ball no longer appears to be moving up, down, left or right but does appear to be growing bigger and bigger as it approaches. This process is known as “optical acceleration cancellation”, because the fielder is constantly trying to use his own movements to cancel out the acceleration of the ball as perceived in his field of vision.
Solving the problem of catching a ball that has been hit very high in the air can be regarded as a fairly elaborate cognitive process, but that does not mean that the fielder’s brain arrives at the solution by manipulating abstract symbols. In fact, working on its own, the brain could never solve the problem. It needs help from the perception of the ball in the player’s visual field and, most importantly, from the movement of the player’s body through space. The two interact in real time in what is known as a perception’action cycle, in which. at any given time, the perception tells the individual what action to take so that certain stimuli will match an internal model whose efficacy has been validated through training over time (in this case, the model that says to keep the ball motionless in one’s field of vision).
Two final remarks. The first concerns the eminently embodied nature of the cognition in this example. The baseball player is a human being with frontal vision and legs that let him move quickly, and these bodily characteristics form an integral part of the cognitive process needed to solve the problem at hand. Working alone, the brain would fail miserably at this task. So we are talking not only about embodied cognition, but about embodied cognition in which a part of the environment (the ball’s position in the sky) is used in real time to complete a task successfully. (For more on this subject, take a look at this course, if you read French.) The concepts involved certainly bring to mind Kevin O’Reagan’s assertion that sensation is a way of interacting with the world.
My second remark concerns the predictive nature of this task, and in particular the fact that the person performing it is constantly correcting errors in relation to a model that he has developed internally. To catch the ball, the fielder must keep it immobile in his field of vision, and hence must move to cancel any movement of the ball whenever the visual input ceases to match the prediction of this model, in which the ball must not move. This behaviour, in which the individual is generally trying to correct an error perceived through the senses, is a form of what is known as predictive processing or predictive coding, This approach, or, if you prefer, paradigm, was developed relatively recently in cognitive science and offers great promise to help us better understand the connections among perception, cognition and action. But its roots lie in a simple, far older idea: that our brain is fundamentally a machine for making predictions, for the ultimate purpose of keeping our organism alive.