On Fri, 2 Aug 2002, Miguel Carrasquer wrote:
> Nobody is denying that PIE was a predominantly suffixing (SOV)
> language. But most if not all suffixing SOV languages do retain a
> small number of prefixes. Basque is SOV and suffixing, but the
> causative, for instance, is a prefix.
>
> PIE too does have a limited number of prefixes. The augment *h1e- is
> directly or indirectly attested in the majority of IE language
> families. Reduplication can also be seen as a form of prefixing. The
> "mobile s" may have once been a prefix, the meaning of which has been
> lost. And there is the "Rasmussen *R-".
Thanks for the support, weak and theoretical as it may be. By way if
parallel, I may add that Eskimo, perhaps the record-holder in terms of
suffixes, has innumerable suffixes and *one* prefix. The prefix is ta-
changing pronouns and pronominal adverbs into anaphorica, as Greenlandic
samani 'down by the sea, in the West* => tasamani 'at the place in the
West I told you about'. I *have* of course already tried to equate IE *O-
(my thanks to anyone sympathetic with its being properly *R-) with Eskimo
ta-, but I am left with a mere postulate. The rest is not based on
postulates, or wishful thinking, quite the contrary - like Glen I hated
the idea of a consonantal origin of some of the o's, not least for the
reason that this was a solution I could not expect anybody to believe even
if it should turn out to be correct -, but the facts have proved so
stubborn as to leave no other way.
Jens