[Lingtyp] genifiers (gender markers/classifiers)

[Alan Rumsey]

I agree with Guillaume’s proposal and the logic behind it. I haven’t been in on all of this discussion, so I don’t know if someone has made this point already, but in northern Australia there are many languages with 2-10 such classes, two of which always include male vs female humans as prototypical kinds of referents. Those of us who have worked on languages with 2-5 such classes (in my case Ungarinyin) have sometimes called them ‘genders’, while those who have worked on languages with more have called them ‘noun classes’. But I think that most or all of us would agree that they are basically the same kind of system, or at least that there is no important difference in kind between those that have, say, 2-3 such classes and those with 4-10. And I suspect that in the interest of standardization most or all of us would readily accept ‘noun class’ or ‘classifier’ as a generic term for such systems, with subtypes that could be distinguished in more structurally significant ways than one based on the sheer number of such classes within a given language.

Alan


From: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> on behalf of Guillaume Segerer <guillaume.segerer at cnrs.fr>
Date: Thursday, 23 March 2017 11:59 pm
To: "lingtyp at listserv.linguistlist.org" <lingtyp at listserv.linguistlist.org>
Subject: Re: [Lingtyp] genifiers (gender markers/classifiers)
Resent-From: <Alan.Rumsey at anu.edu.au>

Hello all,
This is a very inspiring conversation indeed ! May I jump in ?

Martin wrote :


A gender system (= a system of gender markers) is a system of genifiers which includes no more than 20 genifiers and which is not restricted to numeral modifiers.

First, it is obvious (as Martin himself admits) that this arbitrary limit of 20 is problematic. In Keeraak (Joola, Atlantic, Niger-Congo) my own description hesitates between 18 and 21 noun classes, depending on details that need not be mentioned here. Does it mean that this languages might (in a typological perspective) be called a gender language or not ? The Atlantic languages have noun classes systems whose size vary between 3 (Nalu) to 31 (Baïnounk Gubëeher). Following my own quick survey of 44 languages, 30 have 19 classes or less and 14 have 20 classes or more. But all of these systems show many structural affinities.

Second, I'd like to contribute a proposal : why not keep the term 'classifier', and add a one-letter prefix according to the following principle (which can of course be discussed) :
A-classifiers would be for the Amerindian type
B-classifiers would be for the Bantu type (that includes many other Niger-Congo branches of course)
C-classifiers would be for the Chinese type
etc...

One could even make up a F-classifier type (French-like classifiers) to include systems with morphological markers that add a semantic value and take concord in one of the available paradigms (in French for example, suffixes like -ette : diminutive, feminine concord ; -eur : agent nouns, masculine concord, etc.). this solution has the advantage to keep 'classifier' as a cover term without the objections mentioned earlier in this discussion.

Guillaume



--
Guillaume Segerer
LLACAN UMR 8135 - CNRS INALCO
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20170323/6fa27790/attachment.htm>