list James E. Hopkins wrote:

>Kristian wrote that a script without glitches and irregularities=20
>seems too "constructed".  Aren't all scripts constructed?  But=20
>irregularities do have a way of creeping in.

Of course all scripts _were_ constructed. But, the thing about=20
'natural' scripts is that populations use it, not just a single=20
person. This tends to give scripts a 'life' of their own. There are=20
several factors in play here when whole populations are involved such=20
as language change, cultural influences, or adaptation to a foreign=20

The Indic scripts are good examples of how scripts naturally develop.=20
None of them look at all like the original Brahmi script. Although=20
typologically equivalent, many have developed features that are=20
non-existent in the original or dismissed other features that were in=20
the original. And _all_ have developed certain peculiarities. Just to=20
name a few:

Buginese/Makassarese: Fails to mark gemination, glottalization, and=20
     syllable-final nasals.
Burmese: Has an etymological orthography rather than phonemic.
Dehong: Fails to mark tone (but Western diacritics or letters being=20
     experimented to represent tone). Syllable final /m/ represented=20
     by a diacritic, while all other syllable final consonants=20
     represented by independent graphs. Certain vowel and diphthong=20
     distinctions underrepresented.
Javanese: Some letters only used for honorific purposes to be used=20
     throughout a word of any honorific significance. Special=20
     grpahemes also developed for poetic purposes marking the=20
     beginning and end of a poem.
Khmer: Developed two parallel series of C graphemes where each series=20
     have consequences for the representaion of Vs.
Lepcha: Once written vertically, but turned 90 degrees while the=20
     orientaion of the graphs in a line remained unchanged. Has=20
     independent signs for syllable final consonants used as=20
     diacritics. Like Tibetan, independent V graphs have a common=20
     element, and very etymological spelling rather than phonemic.
Philippines: Failure to mark syllable final consonants.
Thai: Developed different classes of Cs for representation of tones.
Tibetan: Unlike most other Indic scripts, independent V graphemes=20
     have a common element. Length not distinguished (except in loan=20
     words where the graph for /ha/ is used as a subscript). Very=20
     conservative spelling - etymological spelling represents sounds=20
     no longer present (eg. /gye/ spelled <ba-rgya-da>).

Maybe some of these peculiarities can serve as inspiration for=20
conlangers designing a script?

-kristian- 8)