Re: Words versus Roots

From: Richard Wordingham Message: 17139
Date: 2002-12-12

--- In, "P&G" <petegray@...> wrote:
> >An example is the word for 'red'. The stem is *h1reudH- (unless
> >misleads us about the laryngeal), but what was _the_ word?
> In which dialect of PIE? Or are you hoping to discover a dialect-
> version? There may never have been a single original proto-dialect
> which the others differentiate. PIE should I believe, be seen as a
> of close dialects, influencing each other mutually and unequally.

Your bottom line is the conclusion I am trying to derive from my
observation that PIE is rich in roots but poor in identifiable words
(I should have said 'stems', not 'words'), which would argue for a
long history of dialect interaction and provide extra time depth.
The issue is whether this view can be sustained. Jens Rasmussen
seems, under 'Does Koenraad Elst...', to have described this view
as 'heretical'; perhaps I misunderstood what was 'heretical' about
his suggestion.

A word that came up yesterday, in the context of Sanskrit
putra 'son', is the whole family of forms related to English 'son';
looking at Pokorny, PIE seems to have had a collection of stems,
namely *su:nu-, *sunu-, *suto-, *suyu- 'son' and possibly even *suka:
> Welsh hogen 'maiden'. I suppose one solution is to suggest that
these 'son' stems are independently derived (or moved from poetic to
normal usage) from the verb *sew 'to bear (a child)' and are not PIE
at all.

(As to Sanskrit 'putra', the question was whether it was a good PIE
word. I took one look at Pokorny and concluded that it was, though
the meaning has clearly been specialised.)