When a fearmonger speaks, people listen
- Published: 07 October 2015
- Written by Joseph Allen
- Hits: 2012
“The end of the world is nigh!” For millennia that spiel has attracted people to gloom and doom prophets. After all, what could be more attention-grabbing than a perpetual threat on the horizon? Less ambitious religious leaders limit their dire warnings to witches, curses, and demons, but the multitudes are still gripped by panic, regardless of the presence or absence of actual danger. Could it be that when our amygdalas squirt fear all over the place, we naturally trust the doomsayer? Pascal Boyer and Nora Parren think so. Their recent research suggests that, all things being equal, people are likely to judge someone relaying threat-related information as more competent than the bearer of boring news.
WARNING: The following study may disrupt your sense of well-being. Proceed with caution!
Anthropologist Pascal Boyer is a Guggenheim Fellow best known for his ambitious theories on the naturalistic origins of religion, which draw on neuroscience, information systems, and evolutionary biology. Nora Parren is a research assistant at the Institute of Cognition and Culture whose international work focuses on the psychology of violence and harm. These two researchers jump off from the premise that, as social creatures, humans depend upon each other for critical information about the environment. If two heads are better than one, the eyes of an entire community are better than two – although some people's vision is more trusted than others. The authors cite previous research which shows certain individuals can easily sway the crowd through swaggering authority, dashing good looks, and in-group signaling. However, in addition to these biases, Boyer and Parren hypothesize that humans also tend to judge sources of “threat-related information” to be more credible. If true, this psychological bias may be the mechanism underlying the rapid spread of terrifying rumors and the frequent violence that follows them.
The researchers conducted five web-based experiments utilizing the M-Turk online survey program. The first three studies tested the primary hypothesis, while the next two eliminated confounding variables. Population samples were pulled from the Unites States, and were relatively small (an average of 127 for the first three, and 69 for the two followups). It is worth noting that gender and ethnicity were somewhat skewed toward white males in some instances, women and minorities in others, which may have affected the results.
In the primary studies, participants read explanations about three new products, ostensibly delivered by various company representatives. The first two studies featured a guided trek in the Amazon, a database program, and a recipe for wildebeest stew. In the third study, they presented a seaside vacation, a new line of diapers, and a washing machine.
Each product presentation was given by two different “company reps,” for a total of six reps per experiment. Both statements about each product contained identical information with different wording, but with an important variation: one description had a sentence containing “threat-related information,” while the other contained either a neutral or nonthreatening negative sentence as the control. Subjects were then asked to rate each representative's competence.
For instance, the participant read descriptions of an Amazonian trek written by “Emilia” and “Maria.” Both laid out the details of what the vacationer could expect, but where Emilia vaguely stated that “There are many species of colorful birds and flowers,” Maria warned “There are leeches that cling to your feet and can give you very serious deep burns.”
Over 70% of respondents rated the rep who included threat-related information as being more competent, with only 25% choosing the neutral delivery, and a fraction choosing “Don't know.” Participants showed the same overall preference for the threat-mentioning software rep and the wildebeest gourmet who warned against overcooking, although the latter showed less bias.
The second study basically replicated the first, only instead of a neutral sentence, the control sentence was a nonthreatening negative statement. For the trekking vacation, this read: “The Amazon is the poorest region of Brazil, with fewer schools, cities, and roads...” This control was used to eliminate any possibility that participants were biased merely toward negative information rather than threats specifically. As predicted, the results were nearly identical to the first study: the majority judged the source who conveyed threat-related information to be more competent.
Interestingly, the third study showed mixed results. This could be because of gender differences (the sample had a greater female-to-male ratio), the content difference, or differences in perceived level of threat. Despite this seeming inconsistency, the overall results of the three primary experiments show overwhelming confidence in individuals who warn of potential danger, with two follow up studies controlling for both perceived honesty and pleasantness—neither of which influenced judgments of competence.
The takeaway lesson: scare tactics may or may not be a crowd-pleaser, but at least they give the impression that you know what you're talking about.
Boyer and Parren situate this cognitive tendency in the ecology of human evolution. According to their interpretation, reflexive trust in people who convey information about danger is an artifact of a dangerous and unpredictable ancestral environment. In a world teeming with mastodons and poisonous mushrooms, individuals who freak out every time a boy cries wolf are likely to survive in greater numbers than those who ask what the wolf looks like. False positives are safer bets than false negatives. Repeat that algorithm for a thousand generations and you have a population distribution favoring those who instinctively react to warnings – whether the threat exists or not. This does not explain why the trait isn’t universal in the U.S. sample, and Boyer and Parren admit their evolutionary interpretation remains speculative until cross-cultural studies are conducted. Even so, the authors' extrapolations to human nature and wider social phenomena do provide a compelling scenario.
Boyer and Parren's paper will hopefully stimulate cross-cultural research into this topic. Whether the tendency to trust fearmongers is biologically universal or culturally specific, what's most important is that the answer to this question might kill us all. You heard it here first.
For more, read “Threat-Related Information Suggests Competence: A Possible Factor in the Spread of Rumors ” in PLOS ONE.