Does your Name Shape your Behavior?
According to a venerable meme, American families gather for Thanksgiving dinner and argue about politics.
However, in recent surveys, fewer than 10% of us report this experience.
What happens instead, at least in some families, is that we try really, really hard to avoid conversational minefields at holiday gatherings.
If that describes your family, no need to worry about Thanksgiving this year. The new study I'll be discussing here is a great conversation piece, because the results are so, ahem, unusual.
According to this study, peoples' first names influence where they choose to live as well as the kind of work they do.
Specifically, the study shows that people choose professions and cities whose names begin with the same letter as their first names. So, Grace becomes a geologist and moves to Gaithersburg. Albert becomes an artist and moves to Addison.
I'm not making this up.
A note on credibility
Before continuing, I want to stress that this study doesn't come from some pay-to-publish vanity magazine. It appears in the latest issue of Journal of Personality and Social Psychology (JPSP): Attitudes and Social Cognition. JPSP is among the most prominent and widely-cited journals in the field of psychology.
You shouldn't trust a study just because it appears in a prestigious journal. (JPSP happens to be well-known for debunked articles.) However, research published in top journals calls for close scrutiny, because these studies tend to have the strongest impact on science as well as practice.
So, let's have a look.
Nominative determinism
The new study, conducted by Promothesh Chatterjee and colleagues at the University of Utah, focused on what the researchers call "nominative determinism", or the ways our behavior is unconsciously shaped by our names. Here's how the researchers frame it:
"In this research, we examine the presence of nominative determinism as it indicates people’s preference for profession or city names that begin with the same letter as the first letter of their first name. For example, nominative determinism would suggest that a person named Dennis is more likely to choose to be a dentist than, say, a lawyer, or that Dennis is more likely to choose to live in Denver than Cleveland."
Boom. (That's the sound of my brain exploding with objections.)
You've probably already thought of reasons why the Dennises of the world choose professions and cities that don't start with "D". But before you dismiss this study as a joke, keep in mind three things:
(a) Scientific findings don't always align with common sense. In the mid-16th century, I would've objected pretty strenuously if someone told me that the earth revolves around the sun.
(b) Nominative determinism isn't determinism in the philosophical sense. In the quotation above, that squishy phrase "more likely to" implies that only some people choose professions and cities that align with their first names.
(c) Older research demonstrates nominative determinism effects (though methodological flaws and replication failures have been observed). This new study didn't appear out of the blue.
Some background
A number of studies show that people unconcsiously prefer the letters that appear in their names. This is called the name-letter effect, and although it may seem odd, it's not as implausible as what Chatterjee and colleagues predicted.
In the classic name-letter effect studies, people view pairs of letters on a computer screen and press a key indicating which letter in each pair they like more. Spontaneous, speedy responses are encouraged.
Given this setup, it seems only mildly surprising that people show a slight preference for letters in their own names.
Explanations for the name-letter effect include familiarity effects (we like what we're most familiar with), ownership bias (we like what belongs to us), and implicit egotism (we like what expresses our identity). Chatterjee and colleagues used implicit egotism to generate predictions for their study. I'll let them explain further:
"How do first names influence profession and city choices? Choices are a conduit through which people express themselves and validate their identity...
An implicit egotism account suggests that because people’s names capture their identity, they make decisions in line with their identity when possible and when their name is the strongest indicator of their identity...
[Thus] choices of profession and city to live in implicitly provide a source for identity expression..."
In case you still have doubts, please note that the lead author, Promothesh Chatterjee, is a professor.
Dennis becomes a dentist, Promothesh becomes a professor. Case closed.
By the end of this newsletter you'll understand why my tone is a little snarky. Here I just want to emphasize that the rationale for the study is suspiciously under-explained. Why would implicit egotism involve first names as opposed to surnames? Why not the meanings of the names (as when Carpenters become carpenters)? And, why not other expressions of identity? Why not predict that because I'm a Ken, I'm not only inclined to teach Kindergarten in Kenosha, but also to shop at K-mart and eat donuts at Krispy Kreme. (Krispy Kreme is amazing, but still....).
Chatterjee and colleagues don't explain clearly why they singled out the first letters of first names. I suspect that what they did was to reverse-engineer a prediction that fits their data. This isn't unethical per se, but it leads to bad science. I'll say more about that in some future newsletter.
Methods
Nominative determinism effects could be revealed via any number of relatively simple approaches. Frankly, it's suspicious that the researchers didn't use one them.
For example, you could just choose a few professions, sample members from a professional directory, statistically control for differences in name frequencies and demographics, then look at the data for high-frequency names.
If you did this, you might find that the percentages of Donnas and Diegos and other "D" names are higher among doctors than among lawyers, while the relative percentages of Latoyas and Larrys and other "L" names are higher among lawyers.
When I first heard about this study, that's what I assumed Chatterjee and colleagues had done. In fact, they took a more indirect approach, using data mining algorithms to examine text from literally billions of web pages, Twitter feeds, and books.
This isn't big data. It's BIG data. And it hints at one reason why the researchers didn't use simple methods and statistics: They had more complicated tools to work with. In science, people who own tanks don't bring water pistols to the battle. Sometimes this is a good thing – complex methodology can be more discerning – but sometimes you just see a methodological arms race, and the most complex approaches aren't the most suitable ones.
In any case, we can see here the confluence of statistics and extraordinary advances in both computing power and artificial intelligence. The researchers used complex AI-driven algorithms to snip name-profession and name-city associations out of unfathomably large swaths of text.
Main findings
The data supported both predictions:
–People were significantly more likely to be in professions whose names begin with the same letter as their first names do.
–People were significantly more likely to live in cities whose names begin with same letter as their first names do.
Startling, right?
(If you're a stats person, you may be wondering: Significantly more likely than what? Using the example of profession, what Chatterjee and colleagues did was to aggregate the data into two groups, compatible (first letters of name and profession match) and incompatible (first letters of name and profession don't match). Regression analyses then determined that cosine similarity was significantly higher for the compatible group. Among other things, cosine similarity allowed the researchers to compare groups without being hampered by the fact that there are fewer professions whose names begin with any one letter than professions whose names begin with all other letters combined.)
A fatal flaw
I've already mentioned some of the many reasons to doubt the findings, but all of that was speculative. What about the actual data?
Let's suppose for the moment that there's nothing wrong with the data. I'm not sure I feel that way, but just for argument's sake, let's assume no methodological or statistical missteps. There's still one enormous, honking problem with the study. It's a "fatal flaw" – in other words, a problem that completely undermines the findings.
Chatterjee and colleagues analyzed trillions of words to determine how people's professions and cities of residence were described in web pages, news reports, and books. Notice though that professions and residences can often be described in more than one way. For instance, someone could write that Michael "just moved to Memphis", or that he "just moved to Tennessee".
Word choice is more or less contextually determined. If Michael had lived in Chattanooga up until last week, you wouldn't write that Michael just moved to Tennessee, because he was already in the state. On the other hand, if Michael had previously lived in Detroit, you could write that he just moved to Tennessee, or you could write that he just moved to Memphis. In some contexts, either description might be fine.
The nominative determinism effect, which is quite small, could arise if there's a slight tendency for people to write that "Michael just moved to Memphis" instead of "Michael just moved to Tennessee" in cases where either phrasing would be fine.
Why would this slight tendency exist? In a word, alliteration. In other words, the choice of words that begin with the same sounds as their neighbors or near-neighbors.
Alliteration is not just something for poets. Writers of all sorts are sensitive to the sounds of words. For instance, reread my previous sentence out loud and you'll notice the alliteration of "sorts" and "sensitive" and "sounds". This works well with the near-pentameter form of the sentence: Wrìters | of all sòrts | are sènsitive | to the sòunds | of wòrds.
My point here is not that alliteration is a major influence on how people write non-literary prose. I'm just saying there are some instances where writers consciously or unconsciously choose terms that alliterate. They write that "Michael had been a musician" rather than "Michael had been an entertainer" or "Michael had been a guitarist", if the exact nature of Michael's profession isn't crucial to the passage.
All it would take to create a nominative determinism effect in this study is a slight tendency toward alliterative writing. The researchers controlled for alliteration at the level of large swaths of text, but not at the level of individual names, professions, and cities.
In short, even if you completely trusted Chatterjee and colleagues' findings, the data doesn't show that a person's first name has a slight influence on choice of professions and cities. Rather, it only shows that the sounds of words have a slight influence on how writers choose to describe people.
Boom. (That's the sound of a study collapsing.)
Further considerations
I've been very snarky about this study. Here are two reasons:
1. Elitism.
Most people don't have much choice about their professions or where they live. You need to be privileged, and/or lucky, to have a range of choices that includes a profession or city whose name has the same first letter as your first name.
2. Silliness.
When people do have choices of professions and cities, issues like money, family, personal preferences, etc. drive their decision-making. The notion that the first letter of their first names has an impact is plain silly. If that's what your data shows, don't publish it right away. Instead, question the data, and maybe gather more. Especially if you work at a public university, where you're being supported in part by other peoples' tax dollars.
(I reached out to the first author for clarification but haven't heard back. So far, when contacting study authors for these newsletters, the response rate has been just over 96%.)
An alternative view
For all their big data, AI-driven statistical modeling, in the end all these researchers offer us is a footnote to a children's picture book. Look at Alicia! She's an Astronaut! (Footnote: Because her name starts with an "A".) As I suggested, the footnote is wrong.
We may be influenced by our first names, but the influence is more subtle than what Chatterjee and colleagues described.
Perhaps the most powerful influence comes from stereotyping by others. There's racial discrimination, most famously described in the 2005 best-seller Freakonomics, as well as other surely unwarranted assumptions about personality and related characteristics. For instance, in a 2018 study of 400 names, "Elizabeth" and "Mark" were rated as higher in both warmth and competence than "Misty" and "Omar".
I call these assumptions "almost" surely unwarranted, because other research in social psychology suggests that people do have some tendency to conform to others' expectations. As they grow up, Misty and Omar may begin to exhibit less warmth and competence because people don't expect much of either from them. But this isn't a necessary outcome. Rather, the Mistys and Omars of the world grapple with discriminatory expectations that they may or may not succumb to.
Another alternative
You want to know what's more useful than this peer-reviewed journal article? A picture book by Nelleke Verhoeff called "Bruno Builder Bakes Bread." My daughter bought this book for her toddler and the three of us love it.
Verhoeff provides all the name-profession alliteration anyone could need. Along with Bruno Builder, there's Petra Pilot, Ann Architect, Diego Detective, etc. This might imply that the Brunos of the world all become builders, but it's better than claiming that they all become builders, bankers, bartenders, or baristas, and, anyway, this is a work of fiction for small children.
In addition, Verhoeff created the book in a mix-and-match style, so although Bruno starts out baking bread, by turning the lower halves of each page he can also serve food, care for nature, write a story, and so on.
Along with quietly dismantling traditional stereotypes ("Sulti Scientist Finds the Formula"), Verhoeff's book introduces children to the seemingly endless array of professions and activities that may be available to them.
As we grow up, we begin to choose among whatever opportunities we're fortunate enough to have. If those choices sometimes reflect irrational considerations, we shouldn't worry that our names exert a hidden influence. The only concern, if any, should be that names trigger stereotypes.
In China once, at a wedding, one of the more inebriated attendees misunderstood my first name to be "King" and subsequently treated me more deferentially than he treated the other foreign guests. That's not fair (although, to be honest, I enjoyed being king for an evening).
Thanks for reading!