Dr. David Eagleman

So if the other person has a knife or a gun, that's all they remember.

Describe the guy's face.

I don't remember the guy's face because I was staring at the gun.

So it turns out that what they pay attention to is sort of the wrong thing for forensics purposes.

That's number one.

But number two is this much deeper issue that even amygdala memories are not necessarily accurate.

So our colleague Elizabeth Phelps –

did this experiment right after 9-11 in 2001.

Shortly after the event happened, she went and interviewed lots of people in downtown and midtown New York about what they saw on September 11th.

And she was smart enough to interview them also about what they remembered from September 10th, you know, what they ate for breakfast that day and so on.

Okay.

She then found them three months later.

She followed up a year later.

She ended up doing that 10 years later as well.

What they found is that the traumatic memories of 9-11, even though those are amygdala memories, they drifted just as much as the memories of what they ate for lunch on September 10th.

And so an unfortunate –

Fact for the law is that memories are not accurate.

They drift.

Every time we check in on memories, we're changing them, and it becomes kind of like the operator game where one person says something in the other person's ear, and the next person repeats that, and the next person repeats that.

There's a sense in which we're always playing the operator game with ourselves.