Is there a language where, given that number and case are affixed seperately not fusionally, a noun can have the structure of , e.g. ithawen = itha-w-en [woman]+GEN+PL ("of the women, the women's")? I can only think of (1) languages that have indiscernible affixes combining case and number (fusional affixes in respect of number and case, e.g. Ancient Greek -ōn or Latin -orum, -ibus, that ‒ to my knowledge ‒ cannot (or no longer) be further reduced to a "sub-affix" component representing either number or case) or (2) languages where case is more proximal to the stem than number (e.g. women's). Examples where case might precede number (as a suffix), the combination may be treated as a new lexeme (is rather a derived than inflected form of the old lexeme), e.g. the Smithses meaning "the Smith's person + plural affix" ([[Smith]s]-es). I beg to be corrected.
-
What about the languages where pluralisation can happen with a modifier + singular form instead of an affix? If we count adpositions as equivalent to case then it seems fair.– Adam BittlingmayerCommented Mar 27, 2018 at 9:01
2 Answers
That's Greenberg's Universal #39:
Where morphemes of both number and case are present and both follow or both precede the noun base, the expression of number almost always comes between the noun base and the expression of case.
Though Greenberg's sample was small, I can't seem to find later counterexamples. The closest thing I could find is the standard reconstruction for Proto-Indo-European; accusative -m and accusative plural -ns could lead one to see the latter as n-s = ACC-PL (see #7 here), though when I look at the whole PIE paradigm, I think that's kind of a stretch.
-
It seems that Kaytete's "pair" suffix comes after the case marker; but the pair suffix is distinct from normal number, even from dual (it marks two items that fit together as a semantic pair; Singer, The Inclusory Construction in Australian Languages). Commented Mar 27, 2018 at 8:55
-
-
1In the Universals Archive homepage #2011 (typo.uni-konstanz.de/archive/ref/source_ref.php#Aikhenvald_1998) elaborates on Greenberg's Universal #39.– AbasCommented Apr 1, 2018 at 10:02
If you allow another order of suffixes like Stem+PluralSfx+CaseSfx instead of your example of Stem+CaseSfx+PluralSfx, then Armenian language has such declension, except for few very old words like շուն[shun] (dog), տուն[tun] (home), etc., which are declined with fusion.
Here PluralSfx is suffix for plurals, and CaseSfx is suffix for one of cases, including "zero" suffix for nominal case.
Example:
- սեղան [seg^ ՛an] (table) -i type declension
- սեղաններ [seg^ann՛er] (tables=pl.nominative-case)
- սեղանների [seg^anner՛i] (of tables, plural, genitive)
- ...
- սեղաններով [seg^anner՛ov] (with tables. plural, instrumental case)
Totally 6 lexical cases in Armenian.
-
3I think the order of the suffixes was the whole point of the original question. Commented Mar 26, 2018 at 21:37
-
I read example of "women's" in p.2 like thing not passing the "separate suffix" criteria. But there is chance that OP meant just that order, CaseSfx+PluralSfx; in that case question could be formulated more specifically.– MnaCommented Mar 26, 2018 at 21:48
-
Sorry, women's was meant to be the negative example. I am looking for an example of Noun+CaseSfx+NumberSfx. Another fictional example èpecen = ep("man")+ec(GEN)+en(PL) where genitive suffix is proximal to the stem than plural suffix.– AbasCommented Mar 26, 2018 at 22:07