Point 1. Far in the past. Probably ergative case
system. Words created at this time don't have tone
deleted by tone in following or previous morpheme.
Final toneless a probably deleted if resulting C
cluster allowed (s and x extrasyllabic) (1).
bha'r+a'+s
bha'ra's
bho'ros "bearer"
pa'd(a)+s
pa'd(a)s
pa'ds 1
po:'s "foot"
Point 2. Middle past. High vowels become glides
before V (2). Now new affixes that are added delete
tone of touching syllable (3). Nom/acc system.
Genitive case created. Two low tones in a row cause
first to be deleted (4).
da`ya`xju+s
da`ya`xjus
daya`xjus 4
dye'xjus "god"
da`ya`xju+a's
da`ya`xjw+a's 2
da`yaxjw+a's 3
da`yaxjwa's
de'ixjwos "of god, divine"
gjha`ya`mb+s
gjha`ya`mbs
gjhaya`mbs 4
gjhye'ms "winter"
gjha`ya`mb+a`n
gjha`yamba`n 3
gjhe'imen(i) "in winter (now)"
Word da`yaxjwa's was probably originally genitive,
meaning "of god", reanalyzed as a'-stem (-a's>-a'+s)
and turned into separate word. Other forms of
daya`xus take tone of nom.
Common phrase "in winter" turned into one morpheme,
n-stem created from it in dialects.
Point 3. Near past. Word formation is as shown in my
synchronic rules.
diachronic changes...
Point 4. PIE as in traditional comparative
reconstructions.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com