On Thu, 24 Sep 2009 13:25:56 -0400, Alex Fink wrote:

> Well, except that the change in question doesn't change the fact of 

> having two consecutive consonants at the same place of articulation.

It depends on how far the "lenition" goes :)

I supposed that Njenfalgar's doubts were motivated by some considerations
like the following: the change could be triggered by the relative difficulty
of sequences of two similar-but-not-identical articulations; but then it is
not clear why the change did not eliminate the identicity of PoA in the
first place. 

So I thought it might be topical to mention the option of a more radical
"lenition" (e. g. to a glide or a laryngeal) :)   

> It would lead

> to funny cooccurrence relations, something like *C-V-(strong C at the 

> same position), where "strong" means 'not the product of any of the 

> possible lenitions'.  

> Overall it's a strange change; I'd be fairly surprised if there was natlang

> precedent.  

IMHO it depends. Imagine, for example, that the "strong" grade is a stop,
and the "lenited" grade is a tap or an approximant. And the latter don't
need to be restricted to just those environments where they result from the
change in question.

But I think I understand what you mean; indeed, a change like -tVt- to -TVt-
or -dVt- does not look particularly natural (especially if this turns out to
be the only environment where [T]/[d] occurs).