Twelve angry men

An article in yesterday’s New York Times disturbed me:

State lawmakers across the country are adopting broad changes to criminal justice procedures as a response to the exoneration of more than 200 convicts through the use of DNA evidence.

All but eight states now give inmates varying degrees of access to DNA evidence that might not have been available at the time of their convictions. Many states are also overhauling the way witnesses identify suspects, crime labs handle evidence and informants are used.

What I quote above isn’t the disturbing part; that’s the good part. Time was, what passed for a justice system relied solely on the testimony of witnesses, along with whatever meagre items of evidence could be collected at the time — blood-stained clothing, a glove left at the scene (ahem...), that sort of thing. Only, we had no technology to analyze the blood on the clothing, which might just as well have been the accused’s own, or that of his pig.

And we have only to look at the Salem witch trials to see how far the testimony of “witnesses” can take us.

Then along came fingerprints. Somewhere around the end of the 19th century, we developed the idea of using fingerprints as evidence to place someone at the scene of a crime. Fingerprints soon became an important bit of identification, and fingerprint evidence was long considered conclusive. Still, it turns out to be more subjective than one might think, and prone to error. And it’s of no use at all if the owner of the prints had perfectly innocent reasons to have been at the scene, to have handled the murder weapon, and so forth.

Nevertheless, many people were convicted, over the years, based on witness testimony, fingerprint evidence, and lack of an alibi. Many of these remain in prison still.

But now we have another means of identification: DNA evidence. DNA evidence is more accurate than fingerprint evidence. What’s more, DNA can be retrieved in instances and from places where fingerprints can’t. A rapist, for example, leaves DNA where he does not leave fingerprints. And the interesting thing is that our archive of physical evidence often includes items from which DNA can now be extracted, new evidence that wasn’t available to us when the crime was committed, investigated, and tried. To quote from the article:

“Technology has made a big difference,” said Margaret Berger, a DNA legal expert who is on a National Academy of Sciences panel that is looking into the changing needs of forensic scientists. “We see that there are new techniques for ascertaining the truth.”

We’re rightly allowing some of those convicted to challenge their convictions anew, using that evidence. And we’re going further, taking steps to reduce future errors. This is all good.

Then we come to this part:

Maryland, North Carolina, Vermont and West Virginia passed legislation this year to create tougher standards for the identification of suspects by witnesses, one of the most trouble-ridden procedures.

Nationwide, misidentification by witnesses led to wrongful convictions in 75 percent of the 207 instances in which prisoners have been exonerated over the last decade, according to the Innocence Project, a group in New York that investigates wrongful convictions.

Legislatures considered 25 witness identification bills in 17 states this year, the National Association of Criminal Defense Lawyers reported. Five states approved bills, while five states defeated them. Bills are pending in seven states.

“It’s become clear that eyewitnesses are fallible,” said Lt. Kenneth A. Patenaude, a police commander in Northampton, Mass., who is an expert on witness identification techniques.

And that’s where things get disturbing. First, it’s the best argument I can imagine for getting rid of executions.

Second, it hasn’t just “become clear that eyewitnesses are fallible.” We’ve known that always. Witnesses lie. Witnesses forget. Witnesses remember the wrong things. Witnesses add their interpretations to what they see. Nearly 60 years ago, Kurosawa showed us how different people can see the same incident very differently. And 50 years ago, Henry Fonda convinced eleven other jurors that an eyewitness was mistaken.

Yet it’s still “eyewitness” evidence that sways us the most. Jurors today watch television shows like CSI, and expect, mostly in vain, to see the sorts of evidence produced there. But give them someone who says, “I saw him do it,” and that’s what stays with them. That’s what they give the most credibility to.

And that’s the evidence that often is most lacking in credibility. Look at the numbers above, again: of the cases that have been turned around by new DNA evidence, three quarters got convictions based on eyewitness identification. There are other, similar numbers, from another source, later in the article. How many of those witnesses were purposefully lying and how many were just mistaken, we have no way to tell.

It’s good that some states are looking at better ways to handle “lineup” identifications, and other aspects of identification by witnesses. But I suspect they can only make a nominal improvement, and it won’t fix our reliance on other parts of witness testimony.

I just know that if I were in the defendant’s chair, I should be very concerned that my liberty might depend upon one witness’s memory, honesty, and sharpness of observation.