Don't touch my phonemes (was: minimal pairs ex: PIE e/o Ablaut)

[Rich Alderson]

On 16 Nov 2000, Stanley Friesen (sarima at friesen.net) wrote:

>> So it is not surprising that it is very difficult to contrast final
>> aspirated and unaspirated stops.

> Difficult, but I am not sure it is impossible.  Did Classical Greek lack
> final tau's? If it had them, then it constitutes an example of a language
> that had this contrast.

Classical Greek had a very limited set of permitted finals:  All the vowels
(5 short, 7 long), /Vn/, /Vs/, /Vps/ and /Vks/, where the two clusters involve
neutralization of voicing and aspiration, as for example in /thriks/ "hair"
where the genitive /trikhos/ undergoes Grassmann.  In those cases where a
dental would cluster with /s/, an old rule reduces the cluster simply to /s/.

> Or how about Sanskrit?  Did it have final unaspirated voiceless stops?

Indeed.  All absolutely final stops are voiceless and unaspirated, though the
sandhi rules conspire not to allow many absolutely final stops.  The citation
forms of nouns, though, is a good example, being the nominative singular--lots
of cluster simplification and the like going on.

BTW, the usual notation for the theta and edh sounds in 7-bit environments like
mailing lists and Usenet newsgroups is /T/ and /D/ respectively, leaving /th/
and /dh/ to represent aspirates, all according to the 1992 ASCII IPA standard,
described by the author at

	http://www.hpl.hp.com/personal/Evan_Kirshenbaum/IPA/

								Rich Alderson