At least there should exist conventions in all languages to transliterate in their own script an IPA representation (used as a "central" phonetic transcription, where the source languafge would be noted using its subset of IPA for representing its initial phonology rather than one particular phonetic realization). Then these phonologic IPA representations should find a good approximation in the target (script/language) pair, in order to produce consistant phonologic transcriptions that are readable orrectly in the target language.
Pure translierations are most often unreadable, or read very incorrectly (even of the target language has a good support for representing the most frequent realizations of a phonologic phoneme of the source language). This scheme could also help transcriptions from one language to another that share the same script (e.g. English "cheese" transcripted in French as "tchise", ignoring the representation of long vowels that are not heard in target French, or "tchiise", but not "tchīse" as the macron is not read distinctly in French). You may argue that you don't need this because we already have IPA, but IPA is unreadable by most people, and there's still the ned to use more conventional symbols (and IPA is completely unreadable for readers of other scripts than Latin, Greek or Cyrillic). The application would be to transliterate people names or toponyms in postal addresses or contact lists or on administrative forms to be used in foreign countries where people can't decipher other scripts (such as Arabic or sinograms), or in airports for travelling, or to avoid that people really "invent" their own choice of name in another script, in suc a way that the chosen name is not registered and verifiable anywhere (unless these people have officially registered in their own coutnry their alternate "usage names", but very few countries permit such registration of such "usage names" by individual people). For those countries that allow registration of people names in other scripts than the national script, most often they will only allow the usage of the Latin script (and frequently in a very restricted subset of it), but not in Arabic, or Greek, or Cyrillic, or Japanese kanas. To help this process, those countries are using their own national standard of transliterators to the Latin script (i.e. "romanizations"), simply because it is the most widely known and used internationally (and in all computer applications) and have no other support for registering additional usage names in other scripts, or for registering additional usage names that will be dependant of the target language (so these single romanizations supported will also be read incorrectly in many target languages, or could be offensive in those target languages and travellers may want ro use another usage name in those target countries). 2012/11/10 Vinodh Rajan <[email protected]>: > Hi, > > These are several standards for transliterating Indic script to Roman > characters such as IAST, ISO 15919 etc. > > I would like to know if any similar standards exist for expressing the Indic > set in Greek & Cyrillic with special diacritics. > > If they do exist, any pointers to their Unicode representations. > > Thanks > > V > > -- > http://www.virtualvinodh.com
