My wife Grace and I welcomed our daughter home from the hospital this week after a normal 3-day stay. We’ve been anticipating this moment for months (pregnancy lasts a LONG time), and one of the ways we dealt with that anticipation was deciding on her name.
Names are honestly a bit strange. We are identified by them, and both we ourselves and the people around us interact with them in some way nearly every day of our lives, but unless we want to deal with the bureaucratic mess and potential for resulting confusion, we usually have zero choice in choosing them.
So when picking our daughter’s name, we wanted, to the best of our ability, to take her perspective: What would she prefer in her name?
We recognized that there are both objective and subjective components to this preference, so we eventually settled on a four-step process to incorporate both:
Use objective criteria to generate scores for a broad range of names.
Draft a “long list” of 20 names from the top scorers according to the objective criteria.
Get feedback from our immediate family on these 20 names, filtering to a “short list” of 5 names.
Get feedback from a broader community of friends and colleagues on the short list to narrow down to a single name.
To be clear, we only went through this rigamarole for our daughter’s English first name, as it matters so much more than the rest. For her middle name and her Chinese name, we just went with the traditional approach of “pick a name that sounds good,” with the additional intentional decision to use my wife’s Chinese family name of 鄞 (yín), to balance using my English family name. Grace’s parents also helped with the Chinese name, eventually settling on 心愉 (xīnyú), meaning “a joyful heart.”
After picking her first name of “Esther”, we chose one of the runners-up, “Iris”, as her middle name, although we weren’t officially committed to doing that ahead of time. As we walk through the process, I’ll trace how those two names stacked up. I’ll also give examples of other names that we’re unlikely to pick in the future to illustrate the general concepts, but I’ll keep the other runners-up out of the discussion in case we want to use those names for a little sister one day.
Objective Criteria
We started with the list of the 1000 most common names of girls born in the United States in 2021. While these collectively only accounted for about 2/3 of the baby girl names chosen that year, we decided that we were okay effectively eliminating less common names. And while we would also have appreciated more Singapore-based data, the reality is that the US Social Security Administration’s data on name frequency is both more extensive and, being a much larger country, is probably more stable.
In the end, we identified 44 binary criteria and rated all 1000 names on each of them, although as you’ll see, that overstates the complexity a bit. Each criterion got a weight between 1 and 10, with a total of 100, so you can think of them, and I’ll write them, as percentages. Here’s how we broke them down:
Practical criteria (36%):
Not in family (10%) or close friends (10%): We didn’t want to repeat a name, because that would be confusing. No “Grace Jr”, thank you very much.
Short (6%): Given how many times she’ll have to write her name in her life, shorter names would be preferable, which we scored at 1% per letter between 4 and 10 letters. (This therefore technically counts as six separate binary criteria, like “less than 7 letters long”.) Sorry, “Evangeline” and “Clementine”.
Single spelling (5%): Having to explain how to spell your name is annoying. “Adaline” / “Adalyn” / “Adalynn” / “Addilyn” / “Addilynn” / “Adeline” / “Adelyn” / “Adelynn” was the most egregious example — all eight of those variations appeared in the top 1000 girl names of 2021.
No nicknames (5%): It’s been annoying to me to have an official name of “Samuel” and a nickname of “Sam”, and to have to keep track in my mind which places I’m listed as Sam and which places I’m listed as Samuel. Fortunately, this is much more common with boy names than girl names, like “Maggie” / “Margaret” and “Luciana” / “Lucy”.
Frequency-based criteria (31%):
Common (9%) but not too common (9%): We wanted her name to be a name people are familiar with, but not one that she often has to distinguish from her classmates. Our ideal frequency was between 1 in 300 and 1 in 2000, which corresponded to the 29th most common girl name to the 347th most common girl name in 2021. Names in that ideal range could still get docked 1-2% if they were trending in the wrong direction. (Of course, every name scored perfectly on one of these two parts; you can’t err in both directions.)
Not too common in Singapore (3%): From the limited data we could find of most frequent baby girl names in Singapore from different sources, we had a slight preference to avoid names on those lists.
Not rapidly changing in popularity (4%): We didn’t want our daughter to have a name that “dates” her as above or below a certain age range, like how “Elsa” probably means that you were born in 2014:
Not “old” (6%): We looked at historical data to find the peak of each name, and identified which generation it was in (Alpha, Z, Y, or X). In calculating the peak, we ignored before frequencies before 1960, both because that’s roughly the age range that will matter to her, and because it was a convenient cutoff in the web interface. It was sobering to realize that most of the most recognizable names which peaked in the top 10 in our generation — names like Sarah, Ashley, Rachel, Lauren, Nicole, Stephanie, Amanda, Jessica, Jennifer, and Megan — are all currently fading so quickly that unless they experience a revival, they’ll sound old when our daughter is an adult!
Meaning-based criteria (19%):
“Culturally appropriate” (5%): This was a tad bit subjective, but let’s just say that US data includes a lot of names of Hispanic origin, and we don’t want people to get the wrong impression of our daughter’s ethnicity from her name. Sorry, “Lola”, “Amari” and others.
Clearly female (5%): As a “Sam” that could be short for either “Samuel” or “Samantha” (but usually “Samuel”), I’ve seen personally how it’s slightly annoying to have to clarify. Sorry, “Carter”, “Palmer” and others.
Christian/biblical origin (3%): As Christians who plan to raise our daughter that way, we saw names with a Christian meaning or biblical origin as a plus. This was the hardest criterion of all to pass, with only 12% of girl names making it by our estimation. (Second hardest was “4 or fewer letters” — only 14.2% scored full marks for being that short.)
Not a word (2%), location (2%), or common last name (2%): The more alternative meanings the name has, the more potential sources of confusion, mean jokes, etc.
Sound-based criteria, particularly paired with my last name of Elder (14%):
Pronounceable (5%): We tried to guess if the name would be easily and immediately pronounceable from its spelling, particularly to Singaporeans unfamiliar with the name who would probably try to sound it out from the syllables. This was bad news for the likes of “Aitana” and “Eloise”.
No last syllable emphasis (3%): After lots of practice saying names with my last name of “Elder”, I realized that emphasis on the last syllable of the name sounds a bit cacophonous, given that “Elder” has emphasis on its first syllable, which means you have to pause between the two emphasized syllables. For example, try to say “Celeste Elder” or “Colette Elder”. Note that this includes monosyllabic names, which is something my mom had already remarked on (it’s why she named my brother “Lucas” rather than “Luke”).
Two syllables (2%, half credit if 3 syllables): Similarly, with the meter of the two-syllable “Elder”, two or maybe three syllables are preferable for the first name. This previously had more weight before I realized that most of the problem was the last syllable emphasis.
Doesn’t end in L or a vowel (4%, half credit if ending in one vowel): And finally, with the “El” of Elder, a first name ending in an L or another vowel would be slightly cacophonous. For a couple examples, consider “Opal Elder” or “Gia Elder”. In both cases, I feel a need for an additional pause between the names to not get tongue-twisted.
As for those percentages, we didn’t decide them all ahead of time; we adjusted them as the data came in and we added more criteria. Sometimes it didn’t particularly matter how much weight we put on something; for instance, only 67 names out of the 1000 got less than the full 9% for being too common in the US.
The overall distribution of scores was fairly normal with a mean of 80.5%, a minimum of 48% to a maximum of 97%. “Esther” tied for 3rd best score overall with a score of 96%. It only lost points for being six letters long (2%), rising in popularity recently (1%, see the graph below), and for being common in Singapore according to one source (1%).
Meanwhile, “Iris” ranked 40th overall with a score of 91%, getting less than full marks for not being Christian in origin (3%), being an English word (2%), rising quickly in popularity recently (2%, see the graph below), which could take it into “too popular” territory soon (1%), and being common in Singapore according to one source (1%).
Drafting a Long List
With those scores in hand, we ranked the names. Four stood out with scores of 96% or higher, and we included those in our long list automatically. Grace and I then drafted names to the longlist in four additional rounds with a lower threshold each round, of 92%, 90%, 87%, and finally, no threshold in the last round, each drafting two names per round in an ABBA fashion, picking first in alternating rounds.
I felt like this approach helped concentrate our attention on the best names by our criteria while also allowing us to include names that we had a personal affinity to. By the end of the draft we were both out of names we were actually excited to keep in the mix, and felt like we were mainly picking red herrings to throw our families off the scent and get them to give their honest opinions.
“Esther” was automatically included in the first round of the draft with the 3rd best overall score. I drafted “Iris” with my first pick in the third round, the first round for which it was eligible.
Feedback from Family
With those 20 names, we sat down and discussed them in random order with Grace’s parents, thinking about four criteria with each name: Is it pronounceable? Does it have the potential for a good Chinese name? And do they each like/dislike it? At the same time, I put the names into a Google form (with the order shuffled randomly each time) and sent it to my brother, sister and parents to get their preferences.
Each family member’s opinion, as well as the other criteria Grace’s parents helped us decide on, was worth up to +2% or -2%. The overall cumulative opinion of our families ranged from -9% to +11%, with an average of +3.4%. Four names tied with that best score of +11%, including both “Esther” and “Iris”. “Iris” had gotten a perfect score from Grace’s parents, while “Esther” was one of only two names of the twenty which received no dislike votes from any of our family members.
Coming in last with -9% was the name that I drafted last, “Winter”, which had felt a bit like a red herring (at least it’s the season she was born in!). Its inclusion mainly led to some laughs as we thought about appropriate Chinese names — 冬天 (“dōngtiān”) is the direct translation, and definitely not a normal name, although 冬冬 (“dōngdong”) would be a funny nickname.
We then narrowed the search to the five names with the highest overall scores after these adjustments, which given the score compression of our drafting method, unsurprisingly included all four that scored +11% with our families. There was a tie for fifth, and we decided to give it to the other name (besides “Esther”) that hadn’t garnered any negative votes from our family members.
Feedback from Community
I then made a new Google Form with the five finalists and sent it to some close friends here in Singapore and our colleagues (mine in the US and Grace’s in Singapore). We got 32 responses in all with a good distribution, and gave each person the chance to influence each name’s score by -1% to +1%, half as much as each of our family members.
Even with 32 responses, those friends’ opinions tended to cancel each other out, with each of the five names getting overall adjustments only ranging from +3% to +9%, with “Iris” getting that top score but “Esther” just 1% behind. The result made “Esther” the clear winner by 4% over the rest. “Iris” was tied for runner-up, and we felt that it sounded good enough together with “Esther” to make it the middle name. By this point, we’d actually gotten quite used to saying “Esther” ourselves and were very comfortable with choosing it.
One reason we wanted to ask our community for input was to help decide whether a name like “Esther Elder” is too sing-songy — it’s both alliterative and rhyming. We found an interesting trend with the friends who responded: With my US colleagues, that aspect of the name was indeed a concern, but the Singaporeans we asked actually liked it. After we had made the decision, a Singaporean teenager I was discussing it with brought up the same properties of “Peter Parker”. Let’s hope that Esther also gets superpowers.
I don’t plan to share very much about Esther publicly like this — we want her to be able to control her own internet presence. When we posted our birth announcement to Facebook, we deliberately covered her face with a baby emoji. I mainly feel comfortable publicly sharing our thought process behind choosing her name because ultimately it reflects more on us as parents (and, frankly, on me in particular) than it does on her.
When she’s old enough, the main thing I hope she takes away from this process is that we tried our best to pick a name that she would be happy to have. In the process, we also got help from some of the people we know will be relevant in her life, like her grandparents and my siblings, and from others who if nothing else helped serve as representatives of many of the people she will meet and interact with in the future.
One of our friends from our church community group here in Singapore who voted strongly in favor of “Esther” talked over the names with her two daughters, 3 and 5 years old, before voting, and told us that “Esther” was their favorite, too. As long as our two families remain close, I hope it’ll be meaningful to be able to tell Esther that those girls, who she could grow to look up to as big sisters, helped to pick her name.
Love to know this. Your Uncle Craig and I each made separate name lists without discussing it. We compared the lists and if we both picked it it made the short list. That way we both felt like we picked it.
I wish I’d researched how common the name Aidan was because there are so many kids named Aidan now.
And we thought our process was complicated!