Long before Reddit users discovered the firefighter Les McBurney, humans were fascinated by the idea that a person’s name influences his destiny.
The ancient Romans even left us a rhyme for this concept, nomen est omen, or “the name is an omen.” The proverb found real-world expression in 70 B.C., when Gaius Verres, a Roman official whose last name translates to “male swine,” was put on trial for myriad acts of plundering and extortion in Sicily. Unfortunately for Verres, the prosecutor in his trial was none other than the legendary orator Cicero, who argued that Verres’s conduct “confirmed his name” — an early example of what we might now call a sick burn.
In the millenniums since Cicero’s gibe, the relationship between names and destinies has increasingly become the subject of scientific inquiry — something not just to be wondered about or disseminated through epic stories but also to be quantified and tested empirically.
I’ve dug into the evidence for nominative determinism, or the theory that a person’s name influences his choice of occupation, interests or spouse, and I think there are good reasons to be skeptical of it. But the continued interest in the idea — across centuries and, arguably, against the evidence — is in itself revealing, highlighting humans’ deep-seated desire for order in a chaotic universe and the role science plays in satisfying that need.
Modern popular interest in nominative determinism can be traced to 1994, when the magazine New Scientist cited an article pointing out that scientists and writers often seemed pulled by their own names toward certain subjects. The best example cited in that column by far is an article about incontinence in the British Journal of Urology written by the team of A.J. Splatt and D. Weedon. It was a New Scientist reader who coined the phrase nominative determinism to describe the theory “that authors tend to gravitate toward the area of research that fits their surname.” (Since then, the term’s meaning has broadened.)
Eventually, the idea that names predict not just occupations but also other life outcomes received full-blown charts-and-graphs scientific treatment. In the early 2000s, a trio of articles published in the prestigious Journal of Personality and Social Psychology argued that people’s names influenced their decisions not only about which professions to go into but also about where to move (they were drawn to towns and streets with names similar or identical to their own) and whom to marry (ditto, but for spouses and last names). To be clear, this is different from research into how others’ responses to people’s names affect their life prospects, such as in audit studies of names presumed to be Black versus white.
Nominative determinism became the stuff of psychology textbooks. Researchers hypothesized that it was driven by implicit egotism, or the idea that we like and are drawn unconsciously to that which we associate with ourselves.
But there were skeptics. One of them was the psychologist Uri Simonsohn, then a professor at the University of Pennsylvania. His 2011 article in J.P.S.P. is a careful, methodical dismantling of those prior studies arguing that implicit egotism draws us to spouses, cities and jobs on the basis of their names. The existing literature, he showed, consistently failed to account for other, simpler possibilities.
Take the finding that Americans are disproportionately likely to marry individuals who already have the same last name they do. Dr. Simonsohn argued that there was no need to invoke implicit egotism to explain this because there was another likely explanation: patterns of ethnic names and intramarriage. That is, if you are an American of Korean descent with the last name Kim and you are looking to marry another person of Korean descent, your pool of marriageable partners includes a disproportionate number of other Kims.
Baby name trends, he argued, were another neglected explanation for apparent cases of nominative determinism. It wasn’t, as a 2002 study suggested, that Dennis was more likely to be a dentist than the similarly common Jerry or Walter; it was that Dennis was more likely to be working, since the average Jerry or Walter was older at the time of the study, and therefore more likely to be retired. (Dennis was also more likely to be a lawyer than the other two.)
The article is a wonderful example of just how complicated and strewed with land mines science can be — and the vital role that principled skeptics can play in driving it forward. When I asked Dr. Simonsohn whether his views on nominative determinism are, overall, the same all these years later, he said yes, he’s still skeptical. He added, “I also have wondered why people are so into this effect.”
To me, that’s an even more interesting question than whether nominative determinism exists.
In my work as a science writer, I’ve found that humans have a powerful drive toward theories that simplify the world and that explain outcomes that otherwise seem random. In the modern era, we’re particularly drawn to scientific theories that allow us to tame all that chaos and uncertainty, distilling it into peer-reviewed research.
Those theories generally don’t stand the test of time: What happened to implicit egotism is a common tale of a burst of excitement followed by a quieter debunking (and a continuing existence of believers and follow-up studies seeming to provide evidence for the phenomenon in question). But the popularity of such theories is revealing in its own right.
Think about how much stuff had to happen just to put you where you are today — from the existence of the universe itself to the rise of life on Earth, to your parents meeting and your early family and educational experiences. Think about how easily any of that could have turned out differently.
Dizzying, no? What would you rather believe: that you are where you are because of an incomprehensibly vast and complex field of particles smashing into one another for billions of years, or because right below the surface subtle forces have guided you, and we can figure out what those forces are? In a strange, mystical way, isn’t it comforting to think that you ended up in San Francisco not because of the vicissitudes of geography and employment but because you’re named Fran and something inside you pulled you there? Or that you can influence the fate of your child’s life by giving her a name that will help guide her toward happiness and success?
I consider myself a natural skeptic. Yet when I think about the countless fascinating ways the theme of names’ influencing destinies has come up throughout human history and literature, I’m intrigued. Part of me wants to be able to enjoy the concept and the mystery of phenomena like nominative determinism without worrying whether they’re true in the peer-reviewed sense.
Don’t get me wrong: Science is great. Science saves lives and got us to the moon. But scientism — the inexorable drive to systematically quantify and explain everything — can also rob the world of a bit of its wonder.
The post Are People Named Dennis Really More Likely to Become Dentists? appeared first on New York Times.