i’m in the middle of writing up a quick script to convert between ULY and UEY, two methods of writing the uyghur language. i had originally experienced a few problems when putting together the xiaoerjing widget and so before i bothered to get too far into the ULY stuff i figured i’d throw together another quick widget that would convert from arabic written in the arabic script to a more of less common transliteration system (specifically DIN 31635).

the problem with transliterating arabic into latin letters is pretty much the same problem as transliterating anything into anything else: it’s not standard. i had the problem when trying to come up with a clear system of converting chinese into xiaojing that would keep the original chinese sounds somewhat visibly obvious. i’m having an even bigger problem in trying to sort out the local wu dialect. there’s a pretty skilled linguist from this city who wrote one of the few texts i’ve seen in english on the topic of changzhou wu named Yuen Ren Chao, or alternatively Zhao Yuanren in pinyin or 赵元任. most of the examples of chinese in his words are transcribed in gwoyeu romatzyh, a system he co-created that uses variant spellings to mark tones. 锅 is guo, 国 is gwo, 果 is guoo and 过 is guoh. if his name sounds familiar as it did to me, it’s probably because you’ve come across one of his other somewhat known works, the lion eating poet in the stone den. unfortunately for me, if i really want to make the best of his work on chinese linguistics, i’m forced to become familiar with gwoyeu romatzyh. but, back to the arabic script.

ULY is the official, xinjiang sanctioned form of transliterating uyghur which was approved at the beginning of this millennium. of course, since there are a huge number of uyghurs living outside china, it’s not really the de facto form used online. more common is a system based on turkish. in the case of arabic, the problem is even greater due to the length of time during which westerners have been studying the language. personally, if i need to transliterate it at all, i prefer a version that would most closely represent the original spelling. the problem there is that then it may not make the pronunciation clear to those not familiar with the system. of course it’s not realistic to expect everyone to just learn the arabic alphabet when their only use is a few phrases in their bookclub book.

the reason i opted DIN 31635 (based on a system from deutsche morgenländische gesellschaft, one of many connections to leipzig in my own arabic education) is that it’s what’s used in the hans wehr, which is in my view the most important book for anyone studying arabic to own. however i have made some modifications in my own version. i’ve made sure to include changes based on the sun and moon letters (al-salaam being actually pronounced as-salaam) that are often ignored in transliteration systems (including the surnames of many of my friends. yes i’m talking about you.) and have added transliteration for tanween which, in order to preserve some sense of the original writing, appears superscript. thus كِتابٌ would render as kitāb^un. السَلامُ عَلَيكُم would be something like as-salāmu ʿalaīkum, however in this case you may not see what looks like a superscript c representing ع. in html it’s ʿ however it may not show on most systems unless the user, in this case you, has gone to the trouble of installing a proper IPA font and it’s certainly not part of georgia, the font widely used on this blog¹. this all points to the problem of not knowing what configuration users may be set up with and at least relating to colour presentation it’s a big reason i got out of web design.

there are a few areas where i think the DIM/DMG system could be improved, for example what is typically rendered as kh or x is in DIM ḫ (h followed by ̮), which, again based on possible deficiencies of the user’s system, may become indistinguishable from ح or ه. in ULY it is marked as x which i think for most people would seem appropriate. in fact i really dislike the idea of spelling reform in english or most languages where it’s an issue. that’s partly why i think it’s important to maintain some sense of the original spelling with transliteration systems. in arabic all the difficult concepts tell volumes on the proper grammar of a phrase. more than that, it gives valuable information on the etymology and related vocabulary. in english, the spelling hints to the origins of the word and thus the proper way to conjugate or decline if you’re so inclined. were we all still using traditional characters, i honestly think learning to read chinese would take less time, not more, simply based on more obvious connections between words instead of making noodle 面² seem closely related to face 面.

slightly off topic, i’d like to quietly voice my dislike of the whole dashboard system on apple computers. i say quietly since i’m still developing applications for it that i truly do hope people find useful. it has been useful for giving me a quick interface to get tonal pinyin from characters (see popupchinese widget) and implementation of the transliteration scripts i’m writing without needing to worry about setting up a web server etc, but mostly i think the whole thing’s a bit of a waste. i must admit i really only have my own widgets in addition to a calendar and the weather set up on my main computer.³

at any rate, we can be thankful for pinyin, as poor a system as it is. regardless of my opinions on cross-straight relations, i wouldn’t be all that sad to see it widely adopted in taiwan as well as overseas mandarin-speaking communities.

—
1. i like it for it’s lowercase numbers and the fact that everyone has it
2. 麵 before simplification
3. i will at some point release the various transliteration widgets, but only after i’ve fixed a few small errors in how things are displayed.