classifiers and plural marking

[Wolfgang Schulze]

Let me briefly comment upon to Edit's Type III (no classifiers and
obligatory plural marking), "subdivided by Jan Rijkhoff into
singular object nouns and collective nouns depending on
whether the meaning of the noun is individual or collective". Though
criteria such as "individual" and "collective" seem to quite helpful in
many instances, a more raffined categorization should be considered at
least for some languages. E.g., South East Caucasian languages normally
conform to the general TYPE III. But in Udi (I should better say in the
language of my Udi informants), a plural must not be used with nouns
[-human], cp.:

(1)	xib ek'
	three horse

	xin z/e	 (_z/- = voiced palato-laminal fricative)
	three stone

Nouns[+human] are not plural marked, if they are indefinite, cp:

(2)	me'tay banek-e-y xib g/ar (_g/_ = voiced uvuluar fricative)
	DEM.GEN.SG be.3sg-PERF-PAST three boy
	"(s)he had three boys"

In this case, agreement is singular, too (_banekey_ shows an endoclitic
_-ne-_ "3Sg(subject)"), see below.
	But if the NP is formed by "MUM+NOUN[+definite]"
then the nouns takes a plural morpheme and agreement becomes plural:

(3)	me xib g/ar-ux gölö-q'un uk'sa
	DEM three boy-PL.ABS much-3pl eat.PRES
	"These three boys use to eat MUCH"

("much" is focused). A utterance like

(4)	xib g/ar gölö-q'un uk'sa
	DEM three boy much-3pl eat.PRES

would infer a context like "I don't want to have many children: three
boys would eat to much (for me)" [or the like]. Note that agreement
still is plural is this case. In (2) agreement is singular, because it
does not cross-reference the "boys", but the general "scene" ("it").
	Conesequently, it seems important to observe both the way of how nouns
are subcategorized in TYPE III, and (in more general terms) to add the
criterion of agreement to Edit's typology.
	In this context I would laos like to ask, what is meant by "plural". If
you simply refer to a morphological "marker", then some instances of
deviation from type IV (no calssifiers and no obligatory plural markign)
may escape your eyes. E.g., as I said in my last mail, Malgasy
definitely belongs to TYPE IV, if morphological plurality is THE
criterion. But what about reduplication? In Malgasy, there is a (now)
prefix (formally probably a noun) _tsi-_ that forms distributive
numerals ("in threes" etc.), cp.:

(5)	tsi-telo-telo  (*tsi-telo)
	PREF-three-three
	"in threes"

Now, if we assume that _tsi-_ originally was a noun, than we would get
another TYPE of NUM-NOUN relationship, that it: Plurality marked on the
numeral, but not on the noun. As far as I remember, this type is also
documented with adjetives, e.g. in Sumerian:

(6)	din/gir-gal-gal  (_n/_ = palatal nasal)
	god-great-great
	"Great (mighty) gods"

This also is valid for some numerals, such as _dili_ "one"

(7)	lú-dili-dili
	man-one-one
	"men (in a group)"; "each man" (distributive).

	s/a(g)dub-dili-dili
	tablet-one-one
	"tablets" ("each of the tablets" (but all of them)

Wolfgang
	
		
--
Prof. Dr. Wolfgang SCHULZE
Institut für Allgemeine und Indogermanische
Sprachwissenschaft * Universität München
Geschwister-Scholl-Platz 1 * D-80539 München
Tel.: +89-2180 2486
http://www.lrz-muenchen.de/~wschulze/