Monday, December 17, 2007
Numerology of Many Lands
This is a bad picture (sorry) of the part of our whiteboard at work where we post the numerals in all the languages we're working in. Numerals are really useful when you're working with newspapers: dates and pages can be ascertained and strange collocations corrected. When we work off of microfilm we come across problems caused by microfilming newspapers Western-style (left-to-right) when they were produced in another sequence (right-to-left, top-to-bottom).
Right under the familiar numbers there are the Chinese numbers up to twelve. As it happens, there is a second kind of number used for fancy. Holidays and other occasions requiring gussying up get a more elaborate number. Those are the second line. My student helper is having trouble remembering them, so as we come across them he writes them up. We're missing three to nine.
Beneath that I wrote the Hebrew numbers just so I could feel useful. Of course I had to write them left to right, to avoid confusion, although if the microfilm issue teaches us anything it's that there can always be unforeseen results of compromises that attempt to simplify language diversity.
Then we have the Arabic numerals. You know, I always thought our numbers were Arabic numerals. I mentioned this to the Arabic-reading staff person who I asked to help me, and she seemed bewildered by this idea. Since they bear no resemblance to each other, she can't imagine why we would call our numbers "Arabic." (But I did find this explanation for how they evolved from actual Arabic).
What makes our work a little more confusing is that the language of the newspapers we're digitizing is not Arabic but Urdu, a South Asian language written in the Arabic alphabet. There are apparently some numbers which are not written quite the same in Urdu, or not always, or perhaps they used to be different but now are standardized, or something of that nature. Googling up "Urdu numbers" did not give us the characters we were finding on the newspapers. Consultation with the unicode character set was more helpful: there are two Arabic number sequences, one called "Arabic" and the other called "Arabic-Indic". Even then, that number two is pretty strange. The only thing I can imagine, after checking my sources, is that it's a holdover from using the letter ba' (the second letter of the alphabet) for the number two. This was apparently an earlier numbering system, much like the Hebrew one. In Arabic the ba' points the opposite way the bet does in Hebrew, and could conceivably have led to that c-shaped number two. The stray characters below the line of Arabic numerals are the variations we've found for those numbers. Could it be that Urdu, too, has different numbers for special occasions?
As I contemplated this plethora of numbering systems, I found myself thinking about Passover. I could not figure this out as we are barely out of Hannukah and it is still a good while to go before the season of our liberation. Then I caught myself humming "Ekhad Mi Yodea," no doubt some sort of subconscious result of staring at the Hebrew numbers. Then I started thinking about "Green Grow the Rushes, O," which might itself be related to a Medieval Hebrew song. This led, naturally enough, to trying to remember the verses of "Red Fly the Banners, O" which I learned at the knees of my Communist grandparents. Of the web versions I could find, this one seems closest to what I recall (up to twelve: we were too polite to sing the anti-Trotskyist verses). However, for two we sang "two, two, the opposites, interpentrating, o" as they have here.
And that brought me back to the ba' and the bet, contraposed, both meaning "two." Opposites, interpenetrating.