Re: [tied] Voiceless Aspirates

No, there is no universal law of conservation to guarantee that. Change can affect syllabification and syllable weight just like it affects everything else. We do have compensatory lengthening of the type *ksenwos > kseinos [-e.:-], *olwos > oulos [-o.:-], or *tinwo: > ti:no: in Greek dialects, but also <ksenos, olos, tino:>. Compensatory lengthening is more likely if the lost segment belongs to the coda and if there is no resyllabification, e.g. RP harm > [ha:m].

Piotr

----- Original Message -----
From: wtsdv

To: cybalist@yahoogroups.com

Sent: Saturday, April 20, 2002 7:16 AM

Subject: [tied] Voiceless Aspirates

The voiceless aspirates in Indo-Aryan derive from a sequence of stop plus laryngeal that later became a unit phoneme. Before doing so, the first vowel in such words as the following should have been long by position, should it not? prthus 'broad, wide' < *prt-Xus vyathate 'trembles' < *vyat-Xa-tay rikhati 'scratches' < *rik-Xa-ti So then after the change of the cluster to a unit phoneme, should there not have been compensatory lengthening in the first vowel to preserve the metric structure? David