One fall evening in 1987, Anthony Hanemaayer, 19 and newly married, was relaxing at his Toronto home when the police banged on his door and arrested him for sexual assault. The victim’s mother had spotted Hanemaayer at work on a local construction site and identified him as the knife-wielding attacker she had confronted in her daughter’s bedroom. She had seen the attacker for perhaps 10 seconds, but she was positive it was Hanemaayer.
The case went to trial in 1989. Convinced by his lawyer that he would be locked up for a long time based on the mother’s testimony, Hanemaayer pled guilty in exchange for a lesser sentence. Hanemaayer, though, was innocent. The crime was actually committed by Paul Bernardo, Canada’s most notorious serial killer. Hanemaayer served two years in prison, losing his wife and job in the process. It wasn’t until 2008 that his name was finally cleared. The thing is, Anthony Hanemaayer looks nothing like Paul Bernardo. Hanemaayer has straight, wire-thin blonde hair, while Bernardo has thick, curly, brown hair. But the victim’s mother was convinced the attacker was Hanemaayer. How could she have been so wrong?
The Hanemaayer case is not the only example of eyewitness testimony gone wrong. Of the 244 exonerations that occurred in the United States since courts began allowing DNA evidence (17 of which were prisoners on death row), faulty eyewitness testimony played a role in 74 per cent of the original convictions. Is it possible that our memories are that bad? Well, as it turns out, yes – and a recent breakthrough in our understanding of memories helps explain why.
Nine years ago, a post doc at NYU named Karim Nader ran an experiment that his supervisor at the time told him would surely fail. First, he conditioned a rat to associate the sound of a bell with an electric shock. That is, he rang a bell just before delivering a mild jolt to the rat’s foot. After several bell-shock pairings, the rat soon remembered that the bell was followed by a shock and it began to freeze in fear at the bell’s ring. Nader then injected anisomycin, a drug that stops the construction of new neural connections, into an area of the rat’s brain where he thought the memory might be stored. He found that if the drug was injected just after the bell rang, while the rat was in the process of remembering that the bell signaled a shock, the memory of the fearful association remarkably vanished – poof! Gone. At the sound of the bell, the rat no longer froze in fear. Nader’s experiment provided the first evidence that the neural connections that store memories have to be rebuilt every single time they are remembered. And during rebuilding, memories can be altered or even erased.
Nader is now a professor in the psychology department here at McGill. An interesting aspect of his work is its ability to explain why eyewitness testimony is, at best, unreliable. Going back to the case of Anthony Hanemaayer, one can imagine the victim’s mother being asked by the police to recount the image of her daughter’s attacker over and over again. Each time she recalled the assault, the neural connections in her brain that stored the memory of the attacker’s face had to be rebuilt and thus became susceptible to alteration. Soon, the unfamiliar face of Paul Bernardo morphed into the face of Anthony Hanemaayer, a man she’d seen before around her neighbourhood.
No one is asked to remember an event more often than an eyewitness, but Nader’s discovery tells us that the most accurate memory is the one least remembered.