On 2008-02-12 20:17, alexandru_mg3 wrote:
> Not true: In Skt. vájra- < *weh2g^-ro- 'thunderbolt' => we have the
> loss of laryngeals before mediae in Indo-Iranian (see Lubotsky 81)
> => see also Skt. pajrá
Lubotsky's explanation is ad hoc. It doesn't account for cases like
*gHla:dH-/*gHladH-; it doesn't explain the absence of *wo:g^- <
*woh2g^-; it proposes paradigms that simply don't work. E.g., how is the
weak stem of the 'nose' word, *nh2s-, supposed to produce Slavic *nos-
or Ved. nas-, especially if (pace Kortlandt) the word has an initial
laryngeal, and so the nasal is _not_ initial? But even if it were, the
vocalisation of *h2, according to Lubotsky's own rules, took place in
Germanic, Italic, Celtic and Greek, not in Slavic or Indic. See also
http://tech.groups.yahoo.com/group/cybalist/message/47994
Cases like *k^aso- have been discussed here as well:
http://tech.groups.yahoo.com/group/cybalist/message/40898
In general, few Indo-Europeists would subscribe to Lubotsky's dogmatic
rejection of *a(:).
Piotr