The Smell of Suspicion

Around the world, from Lithuania to Lesotho, when anyone wants to describe something as suspicious they use a metaphor linked to smell.  And not a pleasant smell.

In the English-speaking world we use the expression “something smells fishy” to convey the idea that something may not be quite what it seems. In other countries the expression used may refer simply to something that smells rotten, like the stench of decaying food.

“It seems to be a universal metaphor,” said Norbert Schwarz, Provost Professor of Psychology and Marketing and founding co-director of the USC Dornsife Mind and Society Center. “In all the more than 20 languages that have been studied by psycholinguists the experience of suspicion is associated with smell.

“Only in English is this association described as being fishy. In German, for example it is rotten. In other languages it’s just a bad smell. But it always seems to be the smell of some spoiled organic food material.”

Turning the expression around on itself

Schwarz decided to take the metaphor literally and find out what happens when people are exposed to a fishy smell and whether it would shift their attention to how things might differ from what meets the eye. His research, the latest of a series of related studies, will be published in July in the Journal of Experimental Psychology. The lead authors on the paper are two of Schwarz’s former students while he was teaching at the University of Michigan, David S. Lee and Eunjung Kim.

“We asked, when you are suspicious, how does that influence how you think?” Schwarz said. “We found that being suspicious makes you more critical and more likely to worry that something is wrong. So you will inspect a statement more carefully and think about it more critically.”

The study asked participants to respond to Erickson and Mattson’s Moses Illusion, which asks, “How many animals of each kind did Moses take on the ark?”

“About 80 percent of people answered ‘two,’ ” Schwarz said. “What they missed is that the question should refer to Noah, not Moses. And this happens even though we tell participants beforehand to flag up a misleading question rather than respond to it.”

However, when the researchers sprayed fish oil under the table where participants sat to answer the question, thereby creating a subtle fishy odor, only about 50 percent of participants failed to detect the misleading nature of the question.

“This study highlights that the metaphors and sayings we use in daily life often have real consequences,” Schwarz said. “If something smells fishy, we really do get more suspicious.”

Schwarz believes this response is an adaptive one. “If we bring something close to us and it has a strange smell, we have learned to check it out very carefully before we expose ourselves to it. That is a wise move that has no doubt served us well for thousands of years, so the response that we are observing in our research is probably a generalization of that.”

A deeper look at the phenomenon

In a second study published in the same paper, the researchers used a classic reasoning test, Wason’s Rule Discovery Task, to test whether olfactory suspicion cues improve the scrutiny of one’s own self-generated hypothesis.

In the task, participants were given a number series — in this case 2, 4, 6 — and asked to find the rule by which the sequence is generated. The task then asks participants to test their own hypothesis by generating six new series. The researcher then tells the participant whether each new series is compatible with the correct rule or not.

“People are terrible at that,” Schwarz said. “Everyone assumes the rule is ‘plus two,’ and so they generate additional series that all follow that rule.” This tendency is called a “confirmatory strategy;” test subjects only look for series that seem to confirm the ‘plus two’ rule they’ve settled on. “That prevents them finding out the actual rule, which is ‘any series of increasing numbers.’”

To see the error of their ways, the participants need to try another method. “In fact, they can only find out the actual rule by using a disconfirmatory strategy; that is, by generating examples that violate their own hypothesis,” Schwarz said.

Participants were asked to solve the task in a room that either had no smell, a neutral odor, or where fish oil had been sprayed into a trash bin under the desk on which participants completed the questionnaire.

While the addition of the fishy smell did not influence which initial hypothesis participants came up with, it did influence profoundly how they went about testing their hypothesis. Without a smell, only 28 percent of participants generated at least one series out of six that could show their hypothesis was wrong, Schwarz said, indicating a low level of critical thinking. “But when it smelled fishy, the number of participants who generated at least one disconfirmatory strategy rose from 28 percent to 48 percent. Thus, fishy smells not only make us suspicious of others, they also make us more critical of our own ideas.”

The essence of embodied metaphors

In an earlier study, published in the Journal of Personality and Social Psychology in 2012 by Schwarz and Spike Lee, now assistant professor of marketing at the University of Toronto in Canada, participants were tested as they played economic trust games. The researchers found that playing the game with a partner in a room that smells fishy decreased interpersonal trust levels and cooperative behavior.

“Other disagreeable odors had no perceptible effect on trust levels,” Schwarz said. “But if it smells fishy then it elicits suspicion, and players’ financial investment in the game dropped roughly by 25 percent as they became less likely to send money to their partner, presumably because they didn’t trust the partner to later share the profits.”

In the 2012 paper, Schwarz and Lee also researched the reverse direction of influence by making participants suspicious and then testing their sense of smell. The results showed that people who are suspicious identify fishy smells at a lower level of concentration.

“The basic mechanism underlying these findings presumably evolved to keep us from ingesting spoiled stuff: If it smells bad, take a moment to make sure it is the right thing,” Schwarz said. “As metaphors in many languages indicate, this link between smell and suspicion has generalized beyond the initial domain, giving incidental smells the power to influence our thinking on unrelated tasks and to undermine trust and cooperation.

“The studies add to a growing body of knowledge bearing on the role of embodied metaphors and sensory information in human judgment.”