Thursday, January 31, 2008


What's in a name?

A while back I got to wondering about the surnames that no one hears anymore. About names that just went extinct, for lack of a better term, when there are no male children in a family to carry the name forwards. A check of the US Census list of names in use in the US shows that it is pretty lengthy, but some of those down towards the bottom are pretty rare. I wondered how quickly names disappear, and what percentage are left.

Then, today, looking for something else entirely, I stumble upon the Galton-Watson process, which is the statistical model that explains just that. Turns out that the longer surnames are used in a culture, the fewer surnames there are. Here's a few examples:
* Korean names are the most striking example, with 250 family names, and 45% of the population sharing 3 family names
* Chinese names are similar, with 22% of the population sharing 3 family names (numbering close to 300 million people), and the top 200 names covering 96% of the population.
* Japanese names have only included a family name since the Meiji restoration in the late 19th century, and there are currently estimated to be over 100,000 family names in Japan.
* Dutch names have only included a family name since the Napoleonic Wars in the early 19th century, and there are over 68,000 Dutch family names.
(For reasons that escape understanding, I am immediately reminded of a line from Hudson Hawk: "There are six hundred and seventeen Wongs in the phonebook... That's a whole lot of Wong numbers.")

Just a few months ago I discovered that all Sikhs share the same two last names, Singh ("lion") for males and Kaur ("princess") for females. (It's a little more complicated than that, I've learned, with a third name being used in some cases to denote clan or subsect, but that's the basic idea.) I very quickly went to check my notes, and sure enough found a Sikh character in a still notional story with the surname Kaur, which I'm sure I borrowed from some source or other. Problem being, the character is a male, and probably not liable to have been named that.

Anyway, just goes to show that for every question you may have, someone somewhere has worked up a statistical model to answer it...

Very interesting, Chris. Thinking about my background (Eastern European Jewish), I found this:

"Upon arriving at this JewishGen Internet site, the researcher is offered an opportunity to search the "JewishGen Family Finder" and the "JewishGen ShtetlSeeker." (Shtetl is the Yiddish word for "small town," but to genealogists it has come to mean "town of ancestry.") ...This database contains more than 45,000 entries of 17,000 unique ancestral surnames, as well as 6,500 town names."

Jewish surnames in the former territory of the Russian Empire only goes back to the 19th century, when required by the Czar's decree. That's when my name Shiffman (an occupation name rather than a location name or patronymic) was taken by numerous unrelated families.
Interesting. Sounds something like the example of Thai names mentioned in the Wikipedia article, only required by government decree since 1920.

I suppose the same thing happened in western European countries a few centuries before. I can't imagine that everyone named "Smith" or "Carter" or "Baker" actually started out related, but were just different individuals who happened to have the same occupation.
Hm. According to the US Census, my wife's maiden name doesn't exist, and therefore neither does my father-in-law or my sister-in-law. ("Ringier" is a pretty rare name, but I don't think it's quite extinct yet. Yahoo's people search brings up seventeen listings in the US for Ringiers.)

So I'm not sure whether to trust that list completely.
That US Census list has "Schiffman" (it is the variant) but not "Shiffman" so I guess I don't quite exist either.
"SCHIFFMAN 0.001 70.072 908"
Post a Comment

<< Home

This page is powered by 

Blogger. Isn't yours?