Ellipsis (for the sake of economy and
convenience) is justifiable in such cases. An asterisked form _also_ represents
a set of systematic correspondences. We routinely say, e.g., that the PIE root
*bHendH- is attested in Germanic, meaning actually not even PGmc. *bind- but the
whole set of its ultimate descendants like Eng. bind, etc. in the individual
languages. The actual forms can easily be enumerated if need be. The presence of
the Alani (Roxolani, Iazyges, ...) among the Germani is known well enough, and a
few "Sarmatian" etymologies have been proposed for Germanic words (<path>
is a very convincing one); however, *xarja-mann- (don't ignore the double *-nn-; it's important in this
case!) stands "without competition" simply because the etymology is
straightforward, self-explanatory and without any formal problems. The second
element is declined just like the root noun *mann- (e.g. OE heremenn glossing
'milites'), so why should it be anything else? There is no need to introduce an
alternative Iranian etymology which is in all respects inferior to a native one,
being neither self-explanatory, nor elegant on the formal side; as far as I
can see the only motivation for it is someone's wishful thinking.
Piotr
----- Original Message -----
Sent: Friday, April 19, 2002 1:52 PM
Subject: [tied] Re: IRMIN
I do, actually. One of the things I remember that they taught
me is that non-attested forms are marked with an asterisk, wherefore your
elliptic expression (I assume you mean that presumed _descendants_ of
non-attested *xarja-man- are attested, which is not quite what you wrote)
confused me. But *xarja-man- as ancestor of <heremann-> etc stands without
competition, exactly because no linguist would consider Iranic as a source of
loanwords since most don't seem to be aware (I certainly wasn't and apparently
neither were you?) of the presence of the Alani among the Germani in the
migration period.