Crash . . . Beep, Beep

In the old “Road Runner” cartoons, the luckless Wile E. Coyote lived by the laws of cartoon physics, defying Newton as he hurled himself through the American southwest in pursuit of his nemesis. In one classic scenario, he would run off a cliff, where he would stand for a few seconds suspended in mid-air. Only when he looked down, and realized his mistake, would the force of gravity kick in, plunging him to the desert floor below.

Why did Wile E. Coyote’s misfortunes make us laugh time and again? Sure, the exploding Acme products were good for a chuckle, but it was really the cartoon physics that made the “Road Runner” one of Looney Tunes’ most popular features in the 1950s. As improbable as it seems, defying Newton’s laws is funny. And it’s funny because we all have an intuitive sense of what’s physically possible and what is not.

But just how reliable is the intuitive physics we practice every day? Psychologist Neal Roese of the University of Illinois decided to examine people’s perceptions of such things as motion and momentum and trajectory to see how they affect judgment. For the purpose of the experiments, he and his colleagues turned to a not-so-looney subject: serious highway collisions. Roese had volunteers watch two computer simulations of head-on collisions, each one prepared for use in an actual courtroom trial. In one case, a car attempts to pass a tractor-trailer on a two-lane highway, crashing head-on into a second tractor-trailer. In the other case, a tractor-trailer swerves to avoid a slow-moving car that has just pulled into traffic; the truck crashes into a bus coming in the opposite direction.

Both of these outcomes are bad. That's not a scientific conclusion, just a thought. But some of the volunteers were spared the worst of it, viewing simulations that stopped short of the actual crash. The others watched complete simulations, including the crash. In addition to watching the videos, Roese had the volunteers read descriptions of the accidents and view diagrams, much as if a lawyer were presenting traditional courtroom testimony. All were then asked to estimate the likelihood of an accident occurring, as a juror would do. The results are reported in the journal Psychological Science.

The most interesting finding was this: Those who saw the incomplete simulation, stopping just short of impact, were more likely to predict a crash than those who actually saw the simulated crash. Yes, you read that right. Apparently when people perceive a dynamic event—with motion, momentum, trajectory—their intuitive physics clicks in and they become hyperconfident about the outcome. Roese calls this new psychological phenomenon the “propensity effect.” This gut-level feeling is probably most familiar to sports fans: For example, baseball fans see a ball hit toward the outfield bleachers and “know” that you can kiss that one goodbye. Indeed, fans say the actual homerun is a letdown after the momentary excitement of “knowing.”

Since both of the courtroom simulations (and the real-life driving errors) actually ended in crashes, the experimental results are less about the accuracy of intuitive physics than people’s confidence in their intuition. The propensity effect, says Roese, is the reversal of a well-known psychological phenomenon known as the “hindsight bias.” Hindsight bias is people’s tendency, once they actually know an outcome, to believe that they “knew it all along.” This was also evident in Roese’s experiments: Those who actually “witnessed” the crash were much more apt to predict a serious accident than were those who just saw the lead up to the accident.

So when it comes to objects in motion, the mind plays tricks, distorting both what we think we knew and what we think we know now about where things are heading. These findings have important legal implications, since computer simulations are increasingly used in courtrooms as a form of persuasive evidence. In those cases, playing with Newton’s laws may not prove humorous.

For more insight into human nature, visit the Association for Psychological Science website at www.psychologicalscience.org.