--- In cybalist@..., "Piotr Gasiorowski" <gpiotr@...> wrote:
> Of course a native speaker's opinion must be valued above anything
else, but what puzzles me is that I've seen this particular plural
form cited many times in various books. It may be a figment, as you
suggest. Anyway, mârios will do just as well as long as it does
mean 'gulf, bay, big lake' or the like.
>
> Piotr
>
As we all perfectly know, some forms, be they real or figment, tend
to migrate from one book to another, so arguments like "cited many
times in various books" are kind of multiplication by zero. Another
fact is that authors tend to use the forms that help to better
illustrate their ideas, sometimes not paying enough attention to the
status of the forms - their reliability, distribution, history etc.
I've consulted some lexicographical sources, so:
1. Common Lithuanian knows only ma~rios (plur. tantum) 'gulf' in the
compound <Kurs^iu, ma~rios>. Other gulfs are called <i,'lanka>.
2. In dialects, the following forms are registered:
2.1 additional semantics for <ma~rios>: 'a big water body'
2.2 <maria`> (sing.) 'the same'
2.3 (much more rare) <ma~re.s> (pl.), <ma~re.> (sing.) 'the same'.
3. Considering the Prussian (<mary>) and Latvian forms, Maz^iulis
postulates Proto-Baltic *<mari:>(sing.):<marja:-> (SING. tantum,
nomen collectivum), and treats <-e.>-stem forms as secondary (*-e:-<*-
ja:-). I don't know which author proposed that *mar-ej-es > ma~re.s
developement (in that case a secondary, analogical origin of ma~re.
[sg.] must be posited), but Maz^iulis would be very unhappy to hear
that.
Sergei