Greek question (night?)

[X99Lynx at aol.com]

In a message dated 3/13/99 7:41:51 AM, you wrote:

<<[ Moderator's response:
>  In those languages in which it does not appear, other processes are at work,
>  usually word structure constraints.  Greek and Latin, for example, restrict
>  the inventory of phonemes which can appear word-finally (Greek much more so
>  than Latin); Latin reduces final *-ss to -s (including the results of the
>  more development of *ts > ss); Sanskrit will allow nearly any voiceless stop
>  to appear in final position, but radically reduces final clusters to their
>  first member; and so on.
>  --rma ]>>

So, if I understand correctly, the nominative sing form in Latin and Greek
reflect the dropping of the -t due to the internal rules of those languages
regarding "the inventory of phonemes which appear word-finally."  Does this
also account for the loss of the -t in Welsh and Polish for example?  Does
Celtic prohibit the ending of a word in /t/?  Can any of these t-less
nominatives be explained as a borrowing?  Does that make any sense?

Regards,
Steve Long

[ Moderator's response:
  For Polish and Welsh, presumably so; I am not versed in the histories or the
  synchronic phonologies of those languages.  Word-final *-t > Proto-Celtic *-d
  if I remember correctly, but the *t in the word for "night" is not word-final
  in PIE.  And why should we bother with borrowing when there are perfectly
  good explanations for the forms encountered that do not require an outside
  influence?
  --rma ]