Just a couple of points
about Danny's references to gender and number in Arabic.
Very briefly, Arabic feminines
in all but a few cases, are derived from masculines, and this usually has a
semantic function, very often to particularise or individualise a general
concept expressed by the masculine, e.g. /shajar/ "trees (coll.)", /shajarah/ "a
tree"; /Darb/ "the act of striking", /Darbah/ "a blow".
Danny's comment :
"BTW, A four-number system of
singular-dual-plural-collective has already
appeared in some local dialects
of Arabic."
is not quite accurate. These
distinctions go back to Classical Arabic, and the situation is in fact more
complex. Leaving aside the dual, which is rarely used, except for
things occurring naturally in pairs - eyes, ears etc., and taking the
example of /shajar/ "collective tree(kind)", one can derive plurals from
the collective and the particular, thus
/shajarah/ (fsg) "a tree",
/shajaraat/ (fpl) "some (a few) trees", /shajar/ (msg) "trees (in
general)", /?ashjaar/ (mpl) "all (lots of) (different types of) trees". One
could presumably form duals from both, thus /shajarataani/ (fdu) "two specific
trees" and /shajaraani/ (mdu) "two lots of trees".
In Modern Arabic, the plural of
the feminine form is generally described as a "counted" or
"little" plural.
As for
"I'm not sure this is true for
Afro-Asiatic, but very often
Semitic singular nouns literally change gender
from feminine to masculine
and vice versa when they become plural, so the
"collective" notion may apply
here as well. (You also have singulars
with a vowel suffix that lose the
suffix in the plural: Arabic _g^inni_
"djinn, genie" > _g^inn_ "djinns".)"
The plural of _inanimate_ nouns
are treated for agreement purposes (verbal and adjectival agreement) as if they
were feminine singular. In my opinion, this not so much gender changing or
notions of collectivity, as euphony. Arabic avoids plural endings wherever
possible.In a normal verb-initial sentence, the 3rd person of the verb is always
singular, masc. for singular masculine nouns and plural masculine animates, and
fem. for singular feminine nouns, plural feminine animates and plural
inanimates. Likewise, "broken" plurals are preferred to plural endings, and the
feminine singular of the adjective is used in agreement with inanimate plurals.
This avoids sequences of /-uuna/ (masc.pl. ending) so :
/yajlis ul-kuttaab
ul-kibaar.../, instead of */al-kaatibuuna 'l-kabiiruuna
yajlisuuna.../, ( the chief clerks were sitting...), or
/tuujad ul-buyuut
ul-kabiirah.../, instead of */al-baituuna 'l-kabiiruuna yuujaduuna/ (the
big houses are located...)
Also, /jinn/ is not strictly the
plural of /jinni/. /jinn/ is a singular collective noun. /jinni/ is a specific
demon, and is not a vowel ending, but the "nisba" which is in fact
/-iyy(un)/.
All this, of course, is only
applicable to Arabic. I don't think it applies to Akkadian, or Egyptian for that
matter. So is it an Afro-Asiatic or Semitic thing?
Cheers
Dennis