User:Wikipeditor/Soporifics

From Wikipedia, the free encyclopedia

[edit] Possibility of source reconstruction utilising anticipated wrong pronunciation

Romanisation systems that anticipate use by uninitiated users (foreign news presenters, for example) can be geared towards approximation of a native-like pronunciation by speakers ignorant of the source language (and of the system) by indicating all sound changes and approximating the target language's orthography – American English in the case of Korean romanisations. Whenever naïve speakers use such romanisations but miss the intended pronunciation, those listeners who know the language may be confused (“Was that supposed to be ㅗ or ㅓ?”, German news presenters mispronouncing ㅈ as y etc.) while those who don't know it will not notice or care much anyway. Such confusion would not occur if we kept sounds that are disctinct in the source language easily distinguishable by leading the unprepared reader towards uttering a sound that may not resemble a native speaker's pronunciation at all, but whose disambiguity prevents all confusion, hence allowing those who are familiar with the system to easily deduce the original word – probably with less ambiguity than when using systems of the first type: “The presenter's eh-aw must be based on the romanisation of ㅓ as eo.” Romanisation readers who know the language could still take care of sound changes themselves, just as Koreans do when they read hangeul texts.

In this tradeoff between decent approximation of perfect pronunciation (despite pronunciation by uninitiated readers) on one hand vs. decent semantic disambiguity (again, despite pronunciation by uninitiated readers) on the other, transliterations or – perhaps more so – morphophonemic systems such as Yale lean towards the second. I believe that the first approach should be given up since it doesn't serve its purpose too well (as the case of Korean vowels and plosives illustrates) and that designers of romanisation systems should strive for the latter, so I consider always romanising 리 as -ri a small step in the right direction.

Of course, people are reluctant to stray from romanising as if writing English. As long as the success of romanisation systems is measured by John Doe's performance in pronouncing romanised text compared with a native speaker's pronunciation, romanisations of the second type will not stand a chance.