2012-09-07 21:16, Richard Wordingham wrote:
Some reasons for romanizing:
<snip>
3. Make the language accessible to those who are not familiar with the
script
The rest of the post is irrelevant. Transliterations from Semitic
languages have been established for this reason, and possibly because
of costs of making and setting type.
Those are the key reasons, but there can be other reasons as well. Even
if you can fluently read, say, Arabic letters, an individual word or
phrase in those letters inside a text in Latin letters can look just odd
and distracting. Part of this is that you need to change reading
direction. And it can be difficult to match different writing systems
typographically.
One issue at hand is that there is
not a *single* transliteration to hand, and certainly not a single
pan-Semitic one. Therefore I strongly doubt that an 8-bit code would
encompass everything that was needed.
Indeed. Especially the scientific schemes of “transliteration” make
heavy use of diacritics. The schemes used more commonly in newspapers
and general texts tend to use less accurate systems and deploy letter
combinations, so if people think of such schemes only, it may look like
8 bits, or even 7 bits, could be enough.
(“Transliteration” of Semitic languages is mostly transcription, i.e. it
does not simply map letters of original text to letters or letter
combinations in another writing system—it also introduces letters for
short vowel sounds that are normally not written in original texts. The
difference is essential: pure transliteration is a simple,
character-level process, whereas transcription tends to require
information about pronunciation, meaning, and context.)
Yucca
