Here’s a strange little nugget of psychology trivia: Credit for the phrase “involuntary memory” — when everyday experiences prompt flashbacks to past life events — belongs not to any academic or brain researcher, but to the late-19th and early-20th-century French writer Marcel Proust, who coined it in the first volume of his novel In Search of Lost Time. One of the characters, while eating a tea-soaked cookie, felt himself transported back to a time in his childhood when he’d had a similar snack, recalling with exacting clarity where he was, whom he was with, and what his surroundings looked like. Today, psychologists sometimes refer to this as the “Proust effect”: a sensory experience that calls up a strong autobiographical memory.
Most often, the sense in question is smell, which is intimately tied to autobiographical memory — in fact, some research suggests that memories are most emotional when they’re triggered by scent, as opposed to sight or sound or anything else. That applies to bad ones, too, of course — life’s not all cookies and tea, and neither are our olfactory triggers. And a new study in the journal Frontiers in Psychology suggests one possible application for this sensory quirk of ours, arguing that it may be possible to identify strangers by their body odor — and, by extension, to use smell as a form of testimony in criminal cases.
Generally speaking, bystanders to a crime can provide eyewitness and, to a lesser extent, “earwitness” testimony, matching voices to their owners. “Nosewitness” testimony, on the other hand, has generally been limited to info picked up by dogs, whose noses contain around 50 times as many smell receptors as our do (300 million compared to 5 million).
But highly emotional events are particularly suited to be encoded in our minds as smell memories. As NYU psychologist Wendy Suzuki has explained in LiveScience, this is is a result of the way our brains are structured:
The olfactory system has unique connections with two key regions in the brain’s temporal lobe: the hippocampus, which is critical for laying down new long-term memories, and the amygdala, critical for processing emotions. Unlike all the other senses (i.e., vision, touch and hearing), which require many connections — synapses — to reach the hippocampus and amygdala, olfactory information has immediate access to those systems. It therefore has the ability to lay down long-lasting memories linked to particular times and places (a specialty of the hippocampus) and to include deep emotional resonance associated with those memories (processed by the amygdala).
And what, the Frontiers authors reasoned, could be more emotional than seeing a crime unfold? For the first part of the study, 73 students viewed a video clip of a violent crime and sniffed a piece of fabric containing a stranger’s body odor (a separate group of volunteers had supplied the researchers with the B.O., wearing special pads in the armpits of their T-shirts to capture their scent). After 15 minutes, they were presented with a collection of three, five, or eight smell samples and asked to pick out the one they’d sniffed previously, a sort of olfactory stand-in for the typical police lineup. The people who had chosen from three different smells identified the correct one 96 percent of the time, a number that went down to 56 percent for a lineup of five and 46 for eight (though both of these lower numbers were still better than random guesses).
In the vast majority of real-life cases, though, 15 minutes is nothing — it’s usually much, much longer from when someone witnesses a crime to when they’re offering up their testimony. For the second part of the study, the authors ran a similar experiment to see if they could replicate the effect with a longer interval, asking some people to wait 15 minutes and others one week. While the people in the 15-minute group still did well in picking the right body odor, the longer-interval group didn’t do any better than chance.
Another issue: The smell donors, for lack of a better term, also followed a highly specific set of instructions. In the day leading up to donation, they avoided deodorant and other scented body products, stayed away from spicy food, and generally took care not to do anything that would alter their natural odor. In real life, though, a person’s smell is complicated by all kinds of factors: deodorant, soap, laundry detergent, the last time they showered or washed their clothes, where they’d been that day, and what they’d been doing. The authors referenced a past study showing that people could still identify a smell that had been compromised with “artificial fragrance,” but generally had a harder time than they did with pure B.O.
It’s probably best, in other words, to tread very, very carefully in applying this research to any real-life situations. Eyewitness accounts are often unreliable and easily malleable, and even DNA testing is far from foolproof; Smell may be one more way to help people recall, but it could just as well be a new way to help them confuse the truth.
Perhaps it’s better to strip away the legal applications of this study and appreciate the underlying, more Proustian message: the crazy, fascinating fact that we can recall people, like memories, just by smell. Our noses are powerful instruments for social information, if not yet for justice.