Talk:Classifier (linguistics)
From Wikipedia, the free encyclopedia
Contents |
[edit] need more
this article is obviously only about classificatory systems of the type found in Chinese & other E & SE Asian languages (i.e. it is clearly biased). Athabaskan languages (e.g. Western Apache, Navajo, Slavey, Koyukon) have another type. feel free to adapt what is at Navajo language or Southern Athabaskan languages. one linguist has identified 4 different types of classification. peace — ishwar (SPEAK) 23:33, 2005 May 7 (UTC)
I agree. The systems from austronesian languages, mayan languages should be included, and possibly integrated more with discussion of noun class systems
This article woefully lacks examples. The only example given is 'measure words' in English, which isn't a very fitting choice if the purpose is to help English-readers like me understand how classifiers work in languages where they *are* significant grammatical features (which they're not in English). So if somebody could edit in a few, more significant and/or exotic examples of classifiers, that would be great. Khalil.
Actually, not all linguists agree that classificatory verbs in Athabaskan languages contain classifier morphemes. I believe in Navajo, for example, the classificatory meaning is conflated with other aspects of the meaning of the verb stem, and no separate classifer can be identified. Thus Navajo is said to have 'classificatory verbs' but not 'classifiers' as such. Furthermore, the status of 'classifiers' in signed languages has also been disputed. Adam.
[edit] Number of classes
"...while gender systems vary from two to twenty classes at most." 20 classes for gender? I believe there are only 2 or 3 possible (masculine/feminine + neuter). If there were 20 classes, what would they refer to? Why would anyone call it "gender"? In that case, I think, it would simply be an regular class system. --213.6.70.30 11:21, 13 December 2005 (UTC)
-I believe the point is that so-called class(ifier) systems are simply extended gender systems; or putting it differently, that gender systems are impoverished classifier systems. If 'neuter' can be a gender, why not 'round' or 'pointed' or 'made of metal'?
Mcswell 19:01, 19 December 2006 (UTC)
[edit] Classes
Gender systems are simply a subtype of class systems. Some people say gender system for systems like those of the Bantu languages (which are obligatory and trigger agreement, ie behave just like gender systems but with a larger number of divisions). The article clearly needs to discuss the difference between classifier systems and general class systems (which may indeed have 20 different classes), rather than the difference between classifier systems and the more specific gender type systems. I'd also point out there there isn't a clearly demarked difference between the two: since classifiers often evolve into class systems like those of the Bantu languages, we have a cline rather than a binary opposition, with classifiers at one end and a class system at the other.
- I have doubts that classifier systems often evolve into a class system. Chinese had plenty of time to do so but did not. Doesn't going from a classifier system to a class system involve far more pervasive changes in a language? And probably could only happen with languages that actually inflect?
[edit] Discussion on proposal to merge "Classifiers" with "Measure Words"
Classifiers are distinct from measure words. Measure words allow us to quantify (or count) mass nouns. The measure word is itself the head of the Noun Phrase that contains it -- this is not the case with classifiers. (see section 8.1.2 of "Analyzing Grammar" by Paul R. Kroeger. Cambridge University Press 2005.) By Oborge
-This is not strictly true. Some noun classifiers are the heads of their NPs, depending on the language being considered (that is, whether the classifier or the noun is the head of the NP varies from language to language). See Aikhenvald, Classifiers: A Typology of Noun Classification Devices.
By Chrisdb
-I find both terms unclear, although that may just reflect my ignorance of the literature. My first exposure to these "things" (to use a generic term) was with the Tzeltal (Mayan) language, in which numerals take a suffix indicating the shape of the thing being counted. These were certainly not Measure _Words_, in that they were not _words_: they were suffixes.
My next exposure was to Tucanoan languages in Colombia (and Ecuador). Here the "things" appear after everything but verbs: nouns, numbers, adjectives (if you distinguish adjectives from nouns, a debatable point). Some are clearly suffixes, while others are clearly nouns (and not clitics), and the status (affix vs. word) of some is unclear.
Based on these two exposures (and not knowing anything about other languages that have such systems), it strikes me that "Measure Words" is an inappropriate term, at least if the term is supposed to refer to the morphemes in Tzeltal and Tucanoan. We always used the term "Classifier" in these languages. In Cubeo (one Tucanoan language), these classifiers can be attached to mass nouns, in which case the result is a count noun; but they also attach to count nouns. Some nouns obligatorily take classifiers, others take them optionally. Gender affixes, used on animate nouns, are probably best lumped together with classifiers, used on inanimate nouns (and on animate nouns, e.g. to represent "meat of X").
As for being the "head of the NP", I think that question depends largely on what you mean by "head", and IMHO that's a theory-dependent question.
In any case, the term "measure words" seems misleading to me if it is intended to apply to things that may be affixes.
Mcswell 18:58, 19 December 2006 (UTC)
[edit] Rewrite
This article had little in it, and it was unclear. I decided to be bold, and rewrite it. FilipeS 21:23, 31 October 2006 (UTC)
[edit] Noun classes vs classifiers
The lead section in this article reads:
- A classifier, in linguistics, is a word or morpheme used in some languages to classify a noun according to its meaning. Classifier systems should not be confused with noun classes, which often categorize nouns in ways independent from meaning, such as according to morphology.
Yet the noun class article gives the first of "three main ways by which natural languages categorize nouns into noun classes" as "according to similarities in their meaning (semantic criterion)". The classifier article also states that "there is no clearly demarked difference between the two: since classifiers often evolve into class systems, they are two extremes of a continuum."
I wonder if the second sentence of the lead section couldn't be modified a little, to something like "Linguists usually distinguish between classifier systems and noun classes, although the two terms may overlap in meaning". ??
ntennis 02:05, 3 May 2007 (UTC)
[edit] English
Should mention English analogues, such as "three head of cattle". AnonMoos 07:30, 28 August 2007 (UTC)