From a seat on a train, you watch a fight break out on the platform. A police officer is involved. Later, someone wants to know what color the train seats were. You had a better view of the upholstery than of anything happening outside. How sure are you that you could describe it?
The train existed in virtual reality. The question about the seats was one of several that psychologist Lilian Kloft and her colleagues put to volunteers in an experiment on cannabis and eyewitness memory. The researchers knew exactly what each witness had seen. That gave them a luxury rarely available after a real incident. They could tell when a volunteer had forgotten a detail and when someone remembered a detail that had never been there.
Learning a word under the influence of THC
THC, the main intoxicating compound in cannabis, is particularly good at disrupting the learning of new verbal material. In a review of ten meta-analyses covering several kinds of thinking, verbal learning and memory showed the clearest evidence of impairment during intoxication. But a low score on a memory test leaves a puzzle. Was the trouble hearing and learning the words in the first place, or finding them again when somebody asked?
Mohini Ranganathan’s team used the timing of a word list to tease those possibilities apart. In one experiment, 38 volunteers learned the list before receiving THC or placebo. In another, 57 learned it while the substance was taking effect.
THC reduced recall when people learned the words under its influence. Words learned beforehand remained about as accessible on this test. Think of a name you can repeat at an introduction but have lost by dessert. The moment it goes missing in conversation may be quite a bit later than the moment the difficulty began.
Remembering the scene but missing a detail
A word either appears on a list or it doesn’t. Pictures let researchers ask how closely a memory resembles the original. In a visual memory experiment, 24 occasional cannabis users studied objects set against different backgrounds. A gray cat on a beach was one example. Two days later, sober, the volunteers had to distinguish objects they had seen from close lookalikes. Having studied the pictures under THC made that distinction harder.
Why the wrong cat looks familiar
The beach itself could be persuasive. Restoring the original background helped volunteers recognize the right object, but it also made them more likely to identify a lookalike with high confidence. That happened after placebo as well as THC. A witness could be quite right about where a cat had been and, partly because the setting looked so familiar, quite sure about the wrong cat.
What happens when THC comes after learning?
Suppose two people each recognize eight pictures they saw yesterday. One also claims to recognize four pictures added this morning. Counting only correct answers would give them identical scores.
This distinction helps explain a seemingly odd finding from a University of Chicago experiment. Its 23 participants studied pictures and words while sober. Forty-eight hours later, they received THC or placebo before the memory test. Under THC they still identified as many studied items correctly, yet they also reported more false recollections. Here the error arose even though everyone had been sober when the memories formed.
What the witnesses remembered a week later
After the virtual train ride, the interviewers mixed questions about genuine details with questions about things the witnesses had never seen. Across Kloft’s trial of 64 volunteers, people under the influence were more susceptible to false memories in the virtual scenarios.
When interviewed sober a week later, they showed much less evidence of that difference. On a separate word-list task, however, the extra false recognitions were still detectable a week later. Even within one trial, the answer depended on what people were trying to remember.
The train experiment leaves us with a better question to ask of an eyewitness account. “Do you remember the fight?” asks for the broad outline. “Do you remember the seats being blue?” supplies a detail and invites the witness to accept it. A person can give the same confident yes to both. Only the second answer leaves you wondering who first brought the color into the story.