Do Your Eyes Deceive You? Mounting Evidence Says Yes

The smoking gun. The incriminating tape. The writing on the wall.

We think we know what solid evidence is. We think we know what we see with our own two eyes.

Of course, we don’t. And there’s no better evidence of that than the recent spate of criminal convictions overturned on the basis of DNA analysis--convictions often based largely, sometimes solely, on eyewitness testimony. Even, in some cases, confessions.

What we see with our own eyes is often no evidence at all.

Consider a series of experiments conducted by Dan Simons at the Harvard Cognition Lab. In one, you watch several groups of students toss balls to each other and count the number of completed passes. Simple enough.

Except that when you view the video a second time--this time without bothering to count the ball tosses--you see a man in a gorilla suit stroll right through the ball tossers, turn to face you and pound his substantial hairy chest. Most people don’t see the gorilla when their eyes are on the balls because of what Simons calls “inattentional blindness”--a trick long known to magicians and crooks.

In other equally creepy experiments, Simons shows how you can be merrily giving directions to a stranger--looking that person right in the eye--and not even notice when a different stranger is suddenly substituted. A person of a different height, with different hair, wearing different clothes.

So much for eyewitness testimony.

We see what we expect to see, don’t see what we don’t expect. In the 1930s, during the height of anti-Semitism, the presence of large numbers of Jewish players in pro basketball was interpreted as evidence of their “scheming minds,” “trickiness” and “general smart-aleckness,” according to Michael Shermer in the “The Borderlands of Science: Where Sense Meets Nonsense.”

On the other hand, even firm evidence has been dismissed when it rubbed conventional wisdom the wrong way.

Although anyone can see with their own two eyes that Africa and South America fit together like pieces of a puzzle, it was a long time before even scientists believed that the two had once been part of a single landmass. Until plate tectonics came along, no one imagined a mechanism that could motivate a continent to move.

In science, evidence rules. So it’s not surprising that scientists have come up with a trick or two for sorting out the firm from the flimsy.

What qualifies as “firm” might surprise you. For example, recent results from the Fermi National Accelerator Laboratory suggest that barely-there particles called neutrinos behave just a smidgen differently than theories said they should. And yet, this smidgen (a 1% deviation, to be exact) was greeted by scientists as a possible crack in the foundations of physics.

Why should such a tiny mismatch mean so much? In this case, the reasons have to do with both the solidity of the theory and the precision of the new results. That isn’t always the way: Brookhaven National Laboratory recently had to retract a similar announcement of “new physics” based on a similar smidgen of difference between theory and experiment. While their measurement stands as solid as ever, the theoretical predictions, it turns out, were thrown off by a misplaced minus sign.

Of course, the theorists were motivated to check the calculation precisely because of the mismatch in the measurement. And that’s the crucial point:

The mismatch itself, whether it’s between theory and experiment or eyewitness testimony and DNA, is clear evidence that something is wrong with our understanding--of how the universe works or how the justice system does.

If science has taught us anything, it’s that no one line of evidence is ever enough. Confused and frightened people can confess to crimes they didn’t commit. DNA samples can be contaminated or even planted. Scientists make errors in calculations or let wishful thinking skew their reading of the “facts.”

At the same time, when all the evidence points in the same direction, you have to follow where it leads, no matter how unsettling the result.

An obvious example is the birth of the universe from nothing and nowhere in a primordial explosion--a patently absurd idea if there ever was one. Yet it’s widely accepted because so many different lines of evidence point clearly in that direction: the motions of galaxies, the shape of the universe at large, the composition of matter.

It’s strange but true: You can be more certain of what happened 13 billion years ago when no one was around to see it than you can of what happens today before your eyes.

When it comes to evidence, eyewitness testimony can’t hold a candle to the Big Bang.