s Thoughts from the Physics Chick: Nine — I love you

Sunday, April 05, 2009

Nine — I love you

The other day, a 100 Hour Board question prompted me to do a bit of research on a type of language change called "analogical leveling."

While I was looking for information on this phenomenon, I came across a page which listed several different types of sporadic language changes, including one I'd never heard of: immediate models.

(For the rest of this post, I'm going to discuss immediate models and not analogical leveling. If you're dying to know what analogical leveling is, it's when a word which was formerly irregular in its conjugation or declension becomes more regular.)

And immediate model is when one word changes to become like another word because those two words are frequently grouped together, often in a series. For example, the English word "femelle" turned into "female" because it was often paired with "male," and the word "February" is often pronounced without its first /r/, modeled after "January."

When I read this, something I'd learned many years ago in a Russian class finally made sense. In Indo-European languages, the word for "nine" starts with "n" in almost every language. (Numbers tend to be pretty stable over time.) In French, it's "neuf," in German "neun," in Welsh, "naw," in Greek "ennea" (the "n" gets a vowel in front, but it's still there), in Albanian, it's "nëntë," etc. In Russian, however, the word is девять ("devyat'"), which doesn't seem to make much sense.

One of my professors told me that the "n" changed into a "d" because the Russian word for "ten" (десять/"desyat") starts with "d," but at the time I didn't really understand why. Now that I know about immediate models, though, it all makes sense. "Nine" and "ten" come together in a series, so that proximity gives ten a stronger influence over nine.

Of course, this isn't a very strong effect, since this "ten-ization" only shows up in the Balto-Slavic language family, and leaves the rest of the I-E subfamilies alone.

This makes me wonder, though, if the English "four" is a product of the same change (by analogy with "five"), since most non-Germanic language sub-families have words for "four" that start with "k" or "c" or "q".


At April 05, 2009 8:37 PM, Blogger Th. said...


I vote yes.

At April 05, 2009 11:27 PM, Blogger alea said...

You mean "the Germanic four"? (German/Dutch vier, Swedish fyra, etc.)

I think four is slightly less consistent than nine, though, since Greek has an initial /t/.

At April 06, 2009 8:12 AM, Blogger Katya said...

alea - I meant English to be representative of the Germanic language family, yes. I've clarified that.

And you're right that Greek has an initial "t" for 4 — I wonder if that comes from 3? — but some of the Indo-Iranian languages have words for 9 that start with "f" and at least one Sinhalese language has words that start with "p" or "h."

At April 06, 2009 12:34 PM, Blogger Petra said...

Interesting, I hadn't heard that name for it before, and we just talked about this in my historical ling. class. I'll have to go back to my notes to see what name we used for it. (Since I need to study for my MA exams anyway...)

Oh, and the examples we got were names of the month--in Tok Pisin October is Octember. I'll have to see if four/five is mentioned too.

At April 06, 2009 7:06 PM, Blogger Katya said...

Petra - I'd be interested to know what you called it, because I don't remember learning about it by this name at all.

At April 07, 2009 2:32 PM, Blogger Petra said...

Okay, I checked back through my notes during class today. We called it "contamination," and the examples given in class were Octember, as I said, and then also 'hokto,' a form for 'eight' in some Greek dialects, which was modeled on 'hepta,' seven, rather than the original form 'okto,' eight. (Well, and also the female/male one, because that's the parade example.)

Interesting stuff. I'm going to try to think of some more examples from the languages I know. Part of the interesting bit is why those two numbers, ie why "nine and ten" in Russian but "seven and eight" in Greek? Does there have to be some perceptual threshhold for original similarity between the two? And, of course, like you've asked, given that the numbers are a series in every language, why did that happen in Russian and not in the other languages? How could we predict such a change?

(Hmmm...someone's thinking about a paper topic...)

At April 07, 2009 3:13 PM, Blogger Katya said...

"Contamination." Hmm. OK, that might be a bit familiar, but I don't remember talking about it in terms of series.

I'd be interested to know if there's a general direction for the change. 9 influenced 8 in Russian, 5 influenced 4 in English, but 7 influences 8 in Greek for some dialects, so it's going both directions in numbers. And, of course, January influences February, but it's much more likely that February will drop the /r/ than that January will add an extra one. Maybe there's some optimality theory going on?

Also, if you're just interested in numbers, you'll find this helpful: http://www.zompist.com/numbers.shtml


Post a Comment

<< Home