Richard Wordingham <
richard.wordingham@...> wrote:
> And add the requirement that nominative and accusative be
> distinct.
> Personal pronouns with a rising tone have changed tone (to high)
> through lack of stress in Siamese,
In enclitic/unstressed (non-nominative) function, I suppose?
> Could this difference help explain the s/t- pattern in the
> inflection of *to?
A difference in stress?
I don't think so. The pronouin *so/*to- is a demonstrative, which means that it can be used attributively ("this man") as well as independently ("this is important"). In the second function, the pronoun is necessarily stressed, but in the first function (which is by far the most common one), stressed and unstressed variants may freely occur (the unstressed variant may further develop into an enclitic or proclitic definite article), including unstressed nominative forms, which makes a big difference with the personal pronouns.
In any case, PIE shows no distinction between the stressed and unstressed variants in terms of the initial consonant, which is *s- in the (animate) nominative, and *t- in the inanimate and animate oblique, independently of stress.
I think the diference is ultimately due to a difference in case vowel: nominative *tu- (> *sW- > *s-) versus accusative/oblique *ta- (> *t-).