Re: Simafiran "r"

Jesse Bangs wrote:
 (I read this in the archive, and don't recall whose language Simafiran is):

>> At the end of words, or at the end of syllables when the next
>> syllable
>> begins with a consonant, the "r" hardens up.  It also acquires
>> something
>> of a trill, though I cannot in fact roll my r's at all.  I've seen
>> descriptions of a "tap" or "flap", though in the recordings I have,
>> it
>> simply sounds like an "r" sound read quickly.
>
>I'm not sure what you're describing, but it sounds like allophonic
>variation to me.  Such things are universal in all languages, and I
>guarantee you that you've already included other allophones in your
>language without even realizing it.  So I wouldn't consider this a flaw
>at all--rather, it's a very naturalistic, admirable feature.

Agreed.  My thought:  since your /r/ is velar or uvular, why not have it
become the corresponding stop in the "hardening" environment?