Talk:Noun class

Article split from Grammatical gender
This article has been split from Grammatical gender. See the relevant discussions at Talk:Grammatical gender, perticularly the discussions from this one on.

It's still pretty rough and with duplicates of the material in the other article, but it should improve with time. FilipeS 00:09, 8 November 2006 (UTC)

Difference between Noun class and Grammatical gender
The article needs to explain this better. Some ideas, from here:

FilipeS 20:54, 9 November 2006 (UTC)
 * Noun class, definition: noun classes exist in a language when nouns are divided into groups defined according to grammatical agreement. When the class of a noun changes, this entails morphological changes in other parts of speech that refer to that noun. Noun class may or may not be marked on the nouns themselves.
 * A very simple example of a 2-noun-class system is the singular/plural distinction, in languages that possess these two grammatical numbers.
 * When at least some of the noun classes in a language correlate to natural gender, it is common to speak of "grammatical gender". However, the notion of noun class is not a simple extension of that of grammatical gender, because noun classes traditionally include number distinctions, whereas gender is traditionally seen as a classification system orthogonal to number. Thus, for example, one might say that Spanish has 2 grammatical genders, masculine and feminine, and 2 numbers, singular and plural, but 2x2=4 noun classes, masculine singular, feminine singular, masculine plural, and feminine plural.
 * The article should explain why this difference in approach is considered relevant.

Korean
I have removed Korean from the list of "Languages without noun classes or grammatical genders", where it has been since 2004-04-21 (diff) – with no source given.

If I haven't misunderstood the concept, a single mutually exclusive pair such as Korean …에(-서) “to (from) …” for inanimate nouns vs. …에게(-서) “to (from) …” for animate nouns prove the existence of noun classes in a language. Moreover, with the use of 그녀+suffix (instead of 그+suffix) for female persons becoming increasingly mandatory in South (and North?) Korean, the language has also adopted grammatical gender, except perhaps in colloquial speech. I'm just saying this so you can revert and discuss if you disagree. Wikipeditor 18:00, 22 February 2007 (UTC)


 * Hi. I would not say that a single mutually exclusive pair (such as man/woman) establishes the existence of noun classes (or grammatical genders; I noticed that you've listed Korean in the other page, and I was wondering about that). While there is no strict line between the presence and the absence of grammatical gender, it needs to be a very prevalent feature in the language, and a productive one. Usually, each gender is associated with certain affixes (not necessarily unique) which appear in many, many words (like thousands), with the same value. FilipeS 00:16, 23 February 2007 (UTC)


 * What would be the male affix for German, then? Do articles such as dem count as male affixes? Which affix indicates an animate–inanimate distinction in Russian?
 * For Korean, I can only think of -nyeo for girls and women (leaving boys and men unmarked), which only occurs in the pronoun used for them, and -eg- for animate nouns (leaving inanimate nouns unmarked), which only occurs followed by -e to mean ”to” or by -eseo to mean “from”. If this is not enough, let's move Korean back to “Languages without noun classes or grammatical genders”. Wikipeditor 19:22, 24 April 2007 (UTC)


 * Who said there's only one masculine affix, or that all nouns must have one? You read much too much into my words.
 * I suggest you take a look at the discussion on Japanese, over at the Grammatical gender talk page. FilipeS 19:22, 24 April 2007 (UTC)


 * Sorry, I did not mean to imply there must only be one – are there several? I'm afraid I misunderstood “very prevalent” to mean that more or less all nouns should have one. I am a bit busy now, so I hope to have a more careful look at the relevant articles later. Wikipeditor 19:26, 24 April 2007 (UTC)


 * Hi. Suppose you have a language with noun classes. Then, there must always be some way to assign each noun to a particular class. Often, the class is marked through some sort of affix, but not necessarily. In those cases where there is no affix, the class must be memorized. This is what happens with most German nouns, from what I understand. Another thing about noun classes is that they trigger agreement in other parts of speech. For example, the adjectives that you apply to masculine nouns may need to be different from those you apply to feminine nouns. In the case of German, I know that noun gender triggers agreement in the definine article. To properly refer to a noun, you need to know whether it's a der-word, a die-word, or a das-word. A noun must always be one of these three. FilipeS 19:38, 24 April 2007 (UTC)

Japanese
I've removed Japanese from the list as the minimal pair iru/aru prove that Japanese carries an explicit distinction between animate and inanimate noun classes. --Puellanivis 19:13, 22 February 2007 (UTC)


 * See above. FilipeS 00:16, 23 February 2007 (UTC)

Swahilli example

 * Kitabu kikubwa kinaanguka. (cl.7-book cl.7-big cl.7-PRESENT-fall)

Is that correct? I don't speak kiSwahili, but I know that Kitabu is a loan-word from Arabic (al-kitaabu in the nominative definite), with "ki" being part of the noun. Is the "ki" reinforced somehow, or can the noun change its form (i.e. something other than ki at the beginning)? If it weren't class 7 would it be class whatever prefix+kitabu or would only the "tabu" element remain (which would be a little surprising...). &mdash; ዮም  |  (Yom)  |  Talk  • contribs • Ethiopia 07:24, 2 April 2007 (UTC)


 * Yeah, that's correct. Kitabu is a good example of a loan from Arabic that was reanalyzed in Swahili as starting with the noun class prefix 'ki'. That this is a true reanalysis can be seen from the plural, which is vi-tabu, analogous to ki-ti (pl. vi-ti) 'chair'. This is not an isolated case; another example is ki-oski (pl. vi-oski), borrowed from English 'kiosk'. Quite some time ago I had an interesting conversation with Kwamigami about this subject; some more examples are mentioned there. &mdash; mark &#9998; 08:40, 7 April 2007 (UTC)

declension
If a language has different declensions, does that mean that it has noun classes?--81.145.240.41 08:26, 7 April 2007 (UTC)


 * By most definitions, no. Noun classes are often defined as a categorization of nouns which is reflected on other words in the sentence, for example verbs and modifiers (adjectives, numerals, etc.). The classificatory particles (affixes) in the sentence then agree with the noun, as seen for example in the Swahili sentence above. Declensions on the other hand often signal different categories of meaning, like nominative or ergative case, instead of reflecting a category the noun is in. &mdash; mark &#9998; 11:36, 8 April 2007 (UTC)


 * I didn't mean cases - I was talking about things like Latin's 1st declension and 2nd declension. --81.145.242.99 18:46, 9 April 2007 (UTC)


 * Ah, I think I got confused there. One can certainly think of Latin's declensions as a classifier system, since they serve to mark different classes of nouns. &mdash; mark &#9998; 11:24, 10 April 2007 (UTC)


 * Not at all. Genders are not the same as declensions. In fact, Latin had both things. The difference is that declensions are purely morphological. They have no association to meaning (not even an imperfect association, as often happens with noun classes or gender). FilipeS 15:45, 11 April 2007 (UTC)


 * Now that I've thought about it, the language I was asking about does appear to have gender for nouns -2 mainly masculine declensions, 2 mainly feminine ones and 2 for inanimate objects. However, gender does not reguire grammatical agreement (as opposed to Spanish, for example, in which adjectives must agree with the gender of the noun), so can this language be said to have noun classes or gender? --81.145.242.35 13:41, 12 April 2007 (UTC)


 * I can't say anything without knowing the language. What is it? FilipeS 13:51, 12 April 2007 (UTC)


 * Malayalam --81.145.242.35 13:57, 12 April 2007 (UTC)


 * The Wiki article does not have much to say about grammar. In any case, I would not dare to make definitive statements about a language I know so little about. It seems that some of the Dravidian languages have grammatical genders (for example Tamil), but there is also a category of animacy which interacts with gender in a way I don't fully understand. We need a peer-reviewed source. FilipeS 14:06, 12 April 2007 (UTC)


 * Most grammar books say that there is gender, but I think that's a bit like saying "English has a dative case because Latin does" (in this case "Malayalam has grammatical gender because Sanskrit does"). Thanks for your help. --81.145.242.35 14:12, 12 April 2007 (UTC)

Languages without noun classes or grammatical genders
This article lists English as not having gender or noun classifiers. But, in my opinion, English does have some evidence of the latter. Consider: a "school" of fish, a "herd" of cows, a "flock" of geese, etc. Am I correct, in saying this? Le Anh-Huy (talk) 12:05, 25 November 2007 (UTC)


 * You are refering to English collective nouns. Your edfit to include Mandarin was wrong, because I suspect you were thinking about Chinese measure words. Both of these have nothing to do with grammatical gender. I shall revert your edit. Tebello TheWHAT!!?? 17:21, 25 November 2007 (UTC)


 * On that same note, would the pluralization of nouns in Persian/Farsi as being split along animate (-an) and inanimate (-ha) merit a note in this section? —Preceding unsigned comment added by 205.153.240.48 (talk) 16:14, 18 March 2009 (UTC)

3-level approach
What about mentioning a 3-level approach (for instance, see Johanna Nichols: Linguistic diversity in space and time, page 124 and following), where noun classes are separated in three levels:


 * concord subclass — strictly grammatical agreement, counting singular and plural separately. For instance, German has only one class in plural, the same goes for Russian, Ukrainian, and more classes in singular. This is the finest separation.


 * gender groups words according to what concord subclasses are used. In German, there's no distinction in plural, so gender is determined by singular class alone. In Croatian it's not so, 8 concord subclasses combine to 4 or 5 (traditionally 3) genders. In Chechen, 4 concord subclasses combine to 6 genders.


 * macrogender a set of genders grouped together (I find this quite arbitrary) "often sex of referent". In Croatian, and other Slavic languages this would be 3 traditional genders.

She writes:

"The three levels form a hierarchy when so defined but it should be noted that only concord subclass and gender are agreement classes, while macrogender is not an agreement class [...] My distinction of concord subclass vs. gender follows the Africanist tradition [...] Many recent works divide what I call gender into two categories, one called gender, which refers to systems including a masculine/feminine distinction, and the other called class, which refers to other system. I find this distinction hard to apply to the languages of my corpus, and not revealing of anything else..."

(page 127)

I find this very clear, convincing, and worth mentioning in Wikipedia, if nothing more.

dnik ► 10:04, 30 September 2009 (UTC)

"have elaborate systems of particles that nouns based on shape and function"
I can't parse this sentence. Can someone who understands what it's supposed to mean please rewrite it? Curly Turkey 🍁 ¡gobble! 03:15, 19 February 2016 (UTC)

Categorize under Nouns by type?
"a noun class is a particular category of nouns"

Should it be categorized under Category:Nouns by type? Dpleibovitz (talk) 18:32, 4 March 2018 (UTC)
 * Noun classes aren't types of nouns, but a grammatical category that some languages use to classify their nouns and then, within sentences, to index that class on other words (like verbs or adjectives) that agree with the noun. Maybe placing the article directly under Category:Grammatical categories? – Uanfala (talk) 18:48, 4 March 2018 (UTC)
 * This talk suggests it was split from grammatical category (which is about agreement systems). But I am also unclear about your word use - aren't types the result of classification (based on any reason)? PS. I'm using the words kind, type, class, group, etc. synonymously (wikt:Thesaurus:group). Dpleibovitz (talk) 19:02, 4 March 2018 (UTC)
 * Category:Nouns by type is for individual "subspecies" of nouns; the implicit classification there is unrelated to the noun classes described here. The topic of Noun classes is a certain structural feature of the grammars of some languages. I don't see the two things as congruous. But I'm only speaking out of a general linguistics background, I hope someone more knowledgeable will comment here. – Uanfala (talk) 19:24, 4 March 2018 (UTC)

Error in Article
Under a section, this is said:

Bora, a language of Peru and Colombia has gathered attention for the fact that it has 350 different noun classes, the most of any known language.

However, that's not what the citations say: they say it has over 350 classifiers, not 350 noun classes. So I will remove this section. - Mocha2007 (talk) 22:42, 12 November 2018 (UTC)