Victoroscartangoindianovembergolf...

Hey it's a news story on politics, language, and Boston. I just can't resist.

Ballot translations could mean too much
By Frank Phillips, Globe Staff June 26, 2007 Mitt Romney could be read as Sticky or Uncooked Rice, Fred Thompson as Virtue Soup, and Tom Menino? Rainbow farmer -- or worse.

That's one translation of their names into Chinese, according to Secretary of State William F. Galvin, and if the US Justice Department's voting rights division has its way, that is how they could appear on many Boston ballots in 2008.

Under a 2005 agreement, which Galvin is now challenging in court, the federal government required Boston to translate election ballots -- including the candidates' names -- into Chinese characters in precincts with prominent Chinese-speaking populations.

Galvin said that he has supported, even pushed for the ballots to be printed in Chinese, as long as the surnames remain in Roman letters. Translating them into Chinese, he said, would create chaos and imbalance in an electoral system that needs to be as precise as possible.

The problem, he said, is that there is no actual translation of the names. Instead, the Chinese translate English names phonetically, by finding characters that most closely match the sound of each syllable in the name. There are many different characters that could be used to capture that sound and many different meanings for each character, creating the possibility that the Chinese voters could read something quite other than "Romney" or "Thompson" when they read the ballot.

You know how sometimes when you need to spell a name or word over the phone and you have to give phonetic context words for the letters to clarify? ("That's 'S' like Ball..."). Using this algorithm with the English phonetic context words, we get a spelling for the word VOTING as "victoroscartangoindianovembergolf"

That's sort of how the candidates' names are being proposed to be spelled, except in the case of the proposed character forms, the phonetic context words would represent a syllable of the sound of the target word - like "MEH" from "Menino" - rather than representing a sound segment of the word - like "M" from "Menino".

I'm taking liberties, but overall I think the analogy is more or less sound.

My limited understanding of Chinese characters is that while they can encode sound elements of a word, they did not develop to do so and consequently they do a very bad job at it.

One might be tempted to make a similar point about the English writing system too. But the writing system we use for English is at least meant to encode sounds - it just doesn't do a great job 100% of the time when it's applied to English. It's not the writing system's fault. The mismatch between our current English writing system & spelling rules (a.k.a. "orthography") and spoken English words is due factors like the apparently very high appeal posed by England to would be invaders. We can also put some blame on the monks. Can't forget the monks and their damned roman alphabet and assumption of "when in doubt, go Latin" - which brought us such grammatical fuckery as "To whom do you wish to speak?". But anyone who's got some literacy experience with Italian and Spanish knows that the roman orthography certainly can do a rather excellent job when applied in a less haphazard way.

Not so with Chinese languages and the traditional Chinese writing system. At best, you can squeeze some syllable sounds out of Chinese characters, but you're still working with a writing system that developed to favor encoding meaning over sound. Needless to say, encoding meaning in the writing presents a problem for widespread literacy. Think about it for a second and you'll probably start to see why (I'm not going to summarize a topic which has inspired dissertations, massive research grants, and many years of study to even elucidate). To remedy the literacy problem posed by a non-sound based writing system, several types of phonetic scripts have come into use for Chinese language(s). The most popular and current is Pinyin, which uses roman script to encode the sounds of Chinese words. If you are literate in Chinese, you almost certainly are literate using the Pinyin orthography.

So Chinese can be written using the symbols of the roman alphabet. So why not just put the candidates' names on the ballot the way they are always spelled? Skipping over the fact that the spellings of people's names hardly ever match up 100% to the way they are said (e.g. "It's spelled L-I-P-S-H-I-T-S but it's pronounced 'leep-shech'"), for a writing system to ever have a chance of systematically representing the sound(s) of a spoken word, there need to be accepted rules or conventions for which letters are going to stand for which sounds. These are the familiar spelling conventions like " I before E, except after C... (hey you know who to blame). Pinyin, the writing system which uses roman script for Chinese words, has its own spelling conventions. Thus, putting candidates' names on ballots in a language which would be readable for someone who is Chinese language literate could be as simple as writing them in roman script but using the Pinyin spelling conventions for the sounds of the candidates' names.
Here's my somewhat underinformed first attempt

Romney - Romni
Thompson - Tomsen
Menino - Meninou

I'm sure these are not 100% accurate. For starters, Chinese languages use pitch in ways English does not. These spellings have no tone/pitch indicators because I have NO idea what tones a native speaker of Mandarin would assign to the names. I also don't know for sure that the syllabification native speakers of English give these names would be stable when the names are spoken by a native speaker of Mandarin. But presumably a Mandarin-English bilingual, biliterate individual could be consulted to produce a decent, nonambiguous transliteration (in Pinyin). Although all of this does raise an interesting point - should the written name more closely encode the way the name is said by the person whose name it is or should it represent the way a native speaker of Mandarin would say it?

Possible slight imprecisions in the mapping between the spoken word and the printed word seem so very minor when compared to the large scale lexical ambiguities which are raised by using a non-sound based writing system to encode a (non-native) word's sounds.

The silliness of using traditional chinese orthography to print the sound equivalent of names - most of which aren't even "English" but which appear exclusively in English - is striking. It makes me wonder why one of the underemployed linguists out there couldn't have been consulted before the federal government made this language stipulation.