When we say “Fred is a warm person,” we don’t usually mean his body temperature is hotter than average. We use metaphors such as “warm”, “high”, and “clean” to describe more abstract concepts like “friendly,” “powerful,” and “morally sound.”
So we mean that Fred is friendly, not that he has a fever. But these metaphors can actually have a powerful effect on behavior and attitudes as well. Research has shown that holding a cup of warm coffee makes people more affectionate, and portraying people in physically high locations makes them seem more powerful.
Now newer research is beginning to find that these metaphors are much more common than we might imagine — and that they work in both directions, from abstract emotions and concepts to concrete things, and back.
Spike Lee (the psychologist, not the movie director) and Norbert Schwarz noted that the metaphor “something smells fishy” — meaning suspicious — is actually prevalent in over a dozen languages. So Lee and Schwarz set out to systematically examine the metaphor. Does smelling fishy things make us more suspicious? Does being suspicious make us smell fishy things?
In the first experiment, students at the University of Michigan were recruited on campus in pairs to play a game of trust. Each student was given $5 in quarters to keep.
The first student had a chance to “invest” some or all of his or her quarters in the second student. Whatever they gave to the second student would instantly be quadrupled — a dollar for every quarter. But the second student had the option of keeping all that money or returning some of it to the first student. Depending on how much the first student trusted the second student, both of them could end up coming out ahead.
So how much the first student invested was a measure of their trust — or their suspicion — of the second student. In reality, the second student was an actor — the researchers were only interested in the first student’s behavior.
Before the game was played, the experimenter took the students to a corner of a hall on campus that had been previously sprayed with either 0.5 ounces of fish oil, fart spray (!) or plain water. Here are the results:
On average, students invested nearly a dollar less when the hall had been sprayed with fish oil compared with either fart spray or water. Since the rules of the game involved trusting the second student, a smaller investment suggests the first student trusted the second student less: In other words, they were suspicious. Since students only invested less with the fish smell and not the fart smell, the response appears to be due specifically to the smell of fish, and not just any unpleasant odor.
But Lee and Schwarz were also interested in seeing whether the metaphor also worked in reverse: whether suspicion could affect the sense of smell. In a new experiment, they simply asked student volunteers to smell each of five different test tubes containing liquids, and write down what smell they detected. To evoke suspicion, half the students were given an additional set of “instructions”:
“Obviously, it’s a very simple task and, you know, there’s … there’s nothing we’re trying to hide here.” The experimenter then suddenly noticed a document underneath the participant’s response sheet, hastily took it away, put it in her bag, came back, smiled awkwardly and said, “Sorry, it shouldn’t have been there. But … ahem … anyway. Where was I? Oh yes, it’s all very simple. there’s nothing we’re trying to hide or anything. Any questions? Okay, good, good, you can get started whenever you’re ready.”
Other than the extra instructions designed to induce suspicion, the students smelled the exact same substances, in the same order: Autumn apple, minced onion, creamy caramel, orange nectar, and fish oil. Here are the results:
For all the substances except the fish oil, there was no significant difference in the ability to correctly label the smell between students who were primed to be suspicious and students who received only basic instructions. But suspicious students were significantly better at identifying the fishy smell.
So the metaphor of “something smells fishy” meaning “something is suspicious” appears to work in both directions, from a literal smell to the abstract concept of suspicion, and from the abstract concept back to the smell.
Lee and Schwarz reconfirmed this pattern over several additional experiments. In one of our favorites, students who were primed to be suspicious were significantly more likely to produce fish words from word fragments than students who were not suspicious. For example, a suspicious student would complete TU__ as “TUNA” while a non-suspicious student might fill in the blanks to read “TUBA.”
So why does “fishy” correspond to “suspicious” in so many cultures? It’s hard to know for sure, but one possibility is that many common social interactions involve food. Spoiled food can smell rotten or “fishy,” so if a person is trading for food, suspicion could be legitimately associated with a fish smell.
Lee S.W.S. & Schwarz N. (2012). Bidirectionality, mediation, and moderation of metaphorical effects: The embodiment of social suspicion and fishy smells., Journal of Personality and Social Psychology, 103 (5) 737-749. DOI: http://dx.doi.org/10.1037/a0029708