On the evening of December 19, 1974, a short documentary film was shown on the local NBC newscast in New York. In it, a young woman walks in a hallway. A man lurks in a doorway, wearing a hat, leather jacket, and sneakers. The man bursts from the doorway, grabs the woman’s handbag, and runs straight toward the camera, full-faced. The entire incident lasts twelve seconds.

After the film was shown, the show presented a lineup of suspects. The viewers were provided with a phone number and asked to choose the culprit from among those six, or to say that he wasn’t in the lineup. “We were swamped with calls,” Robert Buckhout, a professor at Brooklyn College who organized the experiment, would later write, They unplugged the phone after receiving 2,145 calls.

The “thief” was seated in lineup position Number 2. He received a grand total of 302 votes from the callers, or 14.1 percent of the 2,145. “The results were the same as if the witnesses were merely guessing, since on the basis of chance (1 out of 7, including the ‘not in lineup’ choice), we would expect only 14.3 percent identifications of any lineup participants, including No. 2,” Buckhout wrote in an article with the charming headline NEARLY 2,000 WITNESSES CAN BE WRONG.

- From “Actual Innocence

If people viewing this show at home from their couch couldn’t recognize the “thief,” why can someone who is getting robbed be reliable enough to consistently convince juries. And then when it comes to situations of rape or murder…how much less reliable should an eye witness ID be considered? 

If you want to get in my head a little as to why I want to be a defense attorney…pick up this book. It’s less than $15. Do it. 

Page 1 of 1