Lao Collation and Preposed Vowels

From: Richard Wordingham (
Date: Mon Oct 30 2006 - 16:31:17 CST

  • Next message: Philippe Verdy: "Re: extracting code values from PDF?"

    One of the differences between Thai and Lao collation is in the treatment of
    preposed vowels. For Thai, collation is done by treating the preposed
    vowels as though they immediately followed the following consonant
    regardless of the consonant, and this is captured by DUCET. For Lao, at
    least in the modern spelling, they are treated as though they immediately
    followed the initial consonant cluster, which may be two consonants long.
    However, DUCET applies the Thai rule to Lao. What is the rationale for
    this? Is there one?

    One possible explanation is that it is not always clear whether the
    consonants following the preposed vowel form a cluster. I've looked though
    Becker's Lao-English dictionary (a pocket dictionary) and found just two
    words where a possible cluster is not a cluster. The current DUCET
    treatment of preposed vowels seems to have been proposed in 2003, before CGJ
    became a disjunctor. Nowadays, the solution would be simple - one would
    encode the word for 'cliff, abyss, chasm' as <E, HO SUNG, CGJ, VO> and the
    derogatory word for 'Vietnamese' as <EI, KO, CGJ, VO>.

    A second possibility is that it was considered that the ordering would have
    greater non-Lao acceptance if it accorded with the Thai ordering.

    I appreciate that Lao collation needs much more tailoring than just handling
    the preposed vowels - it's not even clear that it can be done by tailoring.


    This archive was generated by hypermail 2.1.5 : Tue Oct 31 2006 - 01:28:15 CST