Funny, he doesn't look like a Methuselah

Jun 8th 2007

Names carry social meanings. When you hear a name you automatically draw inferences -- the person's likely sex, age, race, and religion. You make subtler assumptions too, on measures like social class and even personality. But do you form an image of the person's appearance? Does one name have a round face, another long and thin?

A recent study suggests that we do make these connections between names and looks, and that there's broad agreement about what each first name looks like. Psychologist Melissa Lea and colleagues asked subjects to configure a face to fit their mental image of each of a group of names. The researchers then showed the consensus faces to another group of subjects and asked them to pick, for example, which one looked like a Bob and which like a Tim. There was statistically significant agreement: according to the study, a man with a roundish face and features just "looks like a Bob."

The interesting question to me is why. It could be the look and sound of the name itself. The letters of "Bob" carry multiple signals of roundness--think "blob." It could be style. Bob is a friendly, easygoing nickname--think "Bob's your uncle." But it could be something outside of the name itself. The results could be mediated by a broad social difference that carries smaller distinctions in its wake. (At this point I should warn the reader that I'm a recovering experimental psychologist myself. I'm on the wagon, haven't run a subject in 14 years, but still.)

Science Daily featured an image from the scholarly article, headshots of two men with the caption:

"Which person would be inclined to call Bob, and which Tim? An entire lecture hall of students chose the bearded man as Tim and the round-faced man as Bob."

The "Tim" photo has a slim, wiry build, a goatee, and a full head of hair in a longish, casual style. He wears a t-shirt, a choker, and the faint bemused smirk of someone trying hard to look serious. The man called Bob, meanwhile, appears more heavyset with fuller features. He is mostly bald. He wears a striped button-down shirt and a blank, tired expression that reminds me of my driver's license photo.

Yes, one's face is rounder than the other. But the cultural cues in those photos are far stronger than the facial cues. The cues in the Tim picture point much more toward youth...and a Tim is younger than a Bob. Take a look at the names' usage over time:

Granted, the actual experimental stimuli surely eliminated obvious cues like goatees and jewelry. Yet expected age could still be a huge confound in interpreting the results. Did the researchers control for the perceived age of names? Here's a description from the Science Daily article:

"Mixed gender groups of college students participated in the study, which used men's names that appear with equal frequency among that age group."

At this point, my Name Wizard sense is tingling. Something tells me this control is backwards. They asked how likely it is that a guy that age would be named Bob...but they didn't ask how likely it is that a guy named Bob would be that age! (OK, it's not exactly "they didn't move the graves," but we're talking baby name research here.)

The typical Robert is a solid 30 years older than the typical Timothy. If you factor in nicknames, the age difference between them might even be greater. The study used nicknames like Bob and Bill in place of given names, but those forms are out of fashion among young Roberts and Williams...while young Timothys are stil Tim. It's only logical that subjects would assign signs of youth like thinner, more delicate features to the younger name.

Here's a nice testable hypothesis for your next name-image study: our mental models of names include something like NameVoyager graphs, representing the ages of all the Tims or Bobs we've encountered in our lives. Perceived age is based on typicality and availability from that curve. Tease out the age curves from your test stimuli and then we can start finding out for sure whether a Bob is "rounder" than a Tim, and why.

Meet Nymbler

May 31st 2007

How do you start your search for a baby name? Wait, I've got it! You think "I'm looking for a name starting with H that's derived from ancient Phoenician." If so, you're in luck. Lots of baby name sites let you search that way.

But suppose that description doesn't happen to fit. Suppose that your thoughts run more like this: "I've always wanted to name a baby Sophie or Max. But now I know three little Sophies, and we named our beagle Max! Now what?"

That's the kind of baby-naming situation that inspired the Baby Name Wizard book. I've spent years building a storehouse of name information, coding thousands of names by their cultural attributes to find matches in the subtle realm of style. Wouldn't it be nice if a name-search tool thought the same way?

That idea had been in my mind for a long time. Conveniently, some other folks were thinking about it too and had the technical prowess to make it happen. So one day I got a call from a company called Icosystem. They had developed an interesting kind of search technology: "a way of doing search when you don't really know what you are looking for, but you'll know it when you find it." Their Hunch Engine was designed to help people sift through a huge range of options, guiding them in promising directions until they found the right fit. Could my data collection help the Hunch Engine to understand baby names? Could we teach the system about style, make it smart enough to respond to each user's unique tastes? Could it become a kind of virtual baby naming expert?

The result is Nymbler. (Pronounce it "nimbler," it likes that.) Nymbler may not be a full virtual Laura, but it is certainly like no baby naming tool that's ever been. You just choose some names that intrigue you and Nymbler points you down other trails worth following. It may take a little getting used to, but I'm pretty excited about it...and if I were naming a baby right now, I suspect I'd be addicted. Take it for a spin:

2006 Baby Name Pool: Meet the Champions

May 24th 2007

More than 600 name-loving people tried their hands at divining America's hottest and nottest names for the 2006 Baby Name Pool. The two top scorers turn out to be young women who have yet to name babies themselves. What they have done is kept their ears tuned to the frequencies of pop culture (and their noses buried in baby-name books). Allow me to present America's Top Namies.

Grand Champion: Angela "Aiea" S. of Chicago, Illinois. Angela is a 22-year-old actor and name afficionado who is a regular on baby name websites. Her winning lineup of predictions:

Rising - Addison, Shiloh, Angelina
Falling - Katrina, Ty, Jessica

The strength of Angela's entry is in her falling predictions which blew all competitors away. Incredibly, all three of her choices ranked among the 10 sharpest declines of the year. Angela was the only contestant to tab Ty, in a canny headline-watching move: "I based that on the decline in attention that Extreme Home Makeover has been getting." She also resisted giving in to sentiment: "I definitely didn't pick names based on my personal taste, or else I would've predicted that Zelda and Friedrich were going to rise this year. ;)"

A round of applause for Angela!

The trophy for best rising predictions goes to Melissa C., a 21-year-old student from Oshawa, Ontario. Melissa describes herself in terms plenty of us here can relate to: "I am a baby name addict and have been since I was a little girl." Her winning lineup of Addison, Sawyer and Cash were pop-culture inspired: "Addison came from Grey's Anatomy, Sawyer came from the show Lost, and Cash came from the movie Walk the Line after Johnny Cash."

Lest you think that such Hollywood predictions come easily, keep in mind one name that didn't score high this year. 23 entrants picked Shiloh as a rising name. In fact, Angelina Jolie and Brad Pitt's baby name didn't crack the top 1000. It's a crooked path from fame to names...congratulations to the name detectives who followed the trail.