Following the ‘wisdom of crowds’ can stifle diversity

By: James Dean,  Cornell Chronicle
July 20, 2021

People who believe there is a single right answer to a question are better at coordinating, but that benefit may come at the expense of a diversity of opinions, according to new Cornell research about the influence of crowds on moral reasoning.

In a series of mathematical simulations, Shaun Nichols, professor and director of cognitive science in the Sage School of Philosophy in the College of Arts and Sciences, and co-authors compared the behavior of universalists, who believe in a single truth and are more likely to conform to a majority opinion, and relativists, who see multiple potential answers and are reluctant to change their views.

Those dynamics, they found, significantly affect a population’s ability to rally around a common purpose – and potentially promote inequality and eliminate diversity in the process.

“Being a universalist makes you better at coordinating on solutions to problems, and that’s often important,” Nichols said. “But being a universalist also makes you more likely to dismiss minority views as mistaken, and that’s often wrong.”

Nichols is a co-author of “The Meta-Wisdom of Crowds,” published July 3 in the journal Synthese, with philosophers Justin Sytsma, at Victoria University of Wellington, New Zealand, and Ryan Muldoon, at the University at Buffalo.

Coordination – defined as people deciding to do the same thing – has posed a challenge across human history, Nichols said, and is sometimes critical. Consider which side of the road to drive on: left or right, everyone must agree on one side. Language and greeting customs are additional contexts in which arriving at a single answer is beneficial.

That happens much more quickly among universalists, who rely more upon the wisdom of crowds, the scholars demonstrated in coordination games that varied the composition of crowds and thresholds for adopting consensus beliefs. Populations of die-hard relativists, on the other hand, sometimes never achieve coordination.

“There’s a big advantage to being a universalist in conditions where coordination is important,” Nichols said.

In some contexts, however, coordination is unnecessary and may even be undesirable, the researchers argue. For example, if a minority disliked a popular style of music, relativists would not regard that as an error but accept that people have different musical tastes. Universalist pressure to conform to one aesthetic truth could be problematic.

“It would be stultifying if everybody thought there’s just a single right way to make music, or to write literature,” Nichols said.

Crowds inclined to believe in only one right answer may also deepen inequalities, the research determined. That concern was illustrated through a simulation with an “unequal coordination game,” involving a scenario in which coordination is important for both parties, but no outcome is optimal for both parties: a pair of friends who can attend either a gallery or concert together, each having a different preference.

“If you’re a universalist in those games,” Nichols said, “you may rigidify these inequalities and institutionalize an unfair arrangement.”

Examples of harmful coordination include a dinner party in which everyone brings an entrée, or diners all converging on the same restaurant at the same time. In those cases, the authors suggest, people would be better off doing different things, employing a diversity of beliefs embraced by relativists.

Relativists, Nichols said, don’t assume others who make different judgments are wrong, “so the minority isn’t disadvantaged for having minority beliefs.”

The scholars said the research overall presents a more nuanced understanding of the value of universalism, challenging the “problematic assumption” that coordination is an unalloyed good. Beyond social norms, they write, morality may be seen as an area in which widespread agreement is important, “but moral progress is only possible when there is room for people to hold minority views.”

In future research, they plan to explore whether people can learn when to act like a universalist or relativist, when to coordinate or go their own way.

“The next step is figure out how people can determine when they’re in a situation where they should think there’s a single truth,” Nichols said, “and when they should think, ‘Let a thousand flowers bloom.’”


  People placing their hands together in a circle