Re: Feminization of plurals?

2009/2/11 Benct Philip Jonsson <bpj@...>

> On 2009-02-11 Edgard Bikelis wrote:
>
>> there are some plural neuters from latin that
>> entered in portuguese as feminines
>>
>
> There are many collective neuters the
> plural of which ended up as feminine
> singulars in all or most Romance languages.
> Grandgent cited FRIGORA
> MARMORA RURA. This collective -A was
> also extended to some masculines and even
> to some non-first declesion feminines:
> It. _dita, frutta, fusa, grada_, Old
> French _crigne_ < *CRINEA < CRINES,
> Sard, Apulian Rumanian _frunza_ <
> *FRONDIA < FRONDES.
>
> Hasn't it been suggested that *-h2
> originally was a collective and/or
> abstract ending in the first place,
> and that neuters in pre-PIE didn't
> have a plural?
>
>
What I've read is that feminines were actually a conflation of inanimate
collective nouns in -h2 (indeed, whether it was an actual plural suffix or a
collective suffix is in question, although for the purpose of the rise of
the feminine gender it might not be that important, given that Romance
languages didn't have that much difficulty reanalysing neuter plurals into
feminine singulars) and animate nouns with a stem ending in -h2, which
seemingly referred mostly to semantically female concepts (woman, goddess,
that sort of words. Whether that was coincidence or a specific meaning of an
animate -h2 ending, I don't know, and it's not important here).

In other words, the phenomenon that gave rise to the feminine gender in PIE
ends up looking very similar to the phenomenon in Romance languages that
resulted in some plural neuter nouns being reanalysed as feminine singular:
surface forms looking similar end up being reanalysed as a single category,
and take on the characteristics of only one of the original groups. In this
case, inanimate collective nouns ending in -h2 got confused with
semantically female animate nouns ending in -h2 (probably because the
animate/inanimate distinction was losing its semantic meaning. Even in PIE
the distinction is only partially semantic), and they all ended up being
reanalysed as a single group. Since prominent members of that group
(including the word for "woman") were semantically female, the whole group
became a "feminine" gender.

Gender distinctions can change around quite quickly, especially when they
are mostly arbitrary, and the surface forms are not distinctive. Witness
Dutch, which is in the process of completely losing its
masculine/feminine/neuter tripartite gender system for a common/neuter
system, due to the fact that except for personal pronouns (and their
possessive forms) in the singular, masculine and feminine forms have become
identical (same article "de", same adjectival agreement, nouns that don't
have distinctive forms for the masculine and the feminine gender). Other
evidence of this phenomenon is the use of gendered personal pronouns on a
purely semantic basis (so for instance masculine and feminine pronouns are
only used with people or animals whose gender is known. syntactic agreement
doesn't play a role anymore, so "het meisje": "the girl" will be referred to
as "zij": "she", despite the noun "meisje" being neuter. Contrast that with
Modern Greek, where the noun "το κορίτσι": "the girl" is also neuter, as in
Dutch, but governs the use of the neuter personal pronoun "αυτό, το"), and
the rise of "die": "that" as a 3rd person singular pronoun for non-neuter
words whose actual gender is unknown (so we don't have a full loss of
syntactic gender as in English, otherwise people would just use "het":
"it").

Interesting phenomenon. Is anything like that also happening in class
systems like the Bantu system? (Swahili and related languages) I.e. have we
got evidence of reanalysis of classes and conflating due to similar or
identical surface forms?
-- 
Christophe Grandsire-Koevoets.

http://christophoronomicon.blogspot.com/
http://www.christophoronomicon.nl/

Replies