zompist bboard

What do different natlangs or conlangs that have genders or noun classes do with infinitives, action nominals, deverbal nouns and the like? How is the gender or noun class of those words decided? I just ran into this problem when I started working with one of my old conlangs that has eleven noun classes, and I don't know what to do.

One strategy could simply be to not mark the gender of deverbal nouns/infinitives/whatever; or if, for instance, your language marks noun class with class-sensitive articles, maybe infinitives/etc. don't take articles. Or you could either assign them gender arbitrarily, or based on semantics (e.g., Ojibwe nouns formed by adding the nominalizing suffix -win to verbs are always inanimate--since they refer to actions--rather than animate), or have one gender customarily used with (among other things) infinitives/etc. I don't know what kind of semantic distinctions your noun class system makes, though. You could maybe check into how Swahili and other Bantu languages handle this?

Infinitives belong to Bantu noun class 15 (prefix *ku-), e.g. kuishi kwema "live-INF good".

There's also the possibility of assigning the gender based on the phonological form of the word. For example in one of my conlangs, all nouns ending in a fricative are feminine (root only, not counting suffixes).

In another, there's a class which is basically a bin for cases like this. Maybe you could simply say that one of your classes also is the default class for compound words like this.

Whimemsz wrote:One strategy could simply be to not mark the gender of deverbal nouns/infinitives/whatever; or if, for instance, your language marks noun class with class-sensitive articles, maybe infinitives/etc. don't take articles. Or you could either assign them gender arbitrarily, or based on semantics (e.g., Ojibwe nouns formed by adding the nominalizing suffix -win to verbs are always inanimate--since they refer to actions--rather than animate), or have one gender customarily used with (among other things) infinitives/etc. I don't know what kind of semantic distinctions your noun class system makes, though. You could maybe check into how Swahili and other Bantu languages handle this?

The only thing right now that's causing problems in my language is that the passive voice requires a prefix that agrees with the object's class, so your first suggestion doesn't work. EDIT: Oh yeah, I forgot I also have demonstratives that agree with the noun. But maybe it would be possible to not have any verb > noun derivation. I haven't decided on how to handle, um complements I think it's called, as in "begin digging", "aid understanding", "resist changing", etc. The second verb in those cases could be kept as a verb. But I haven't thought yet about other cases where one'd might need to nounify a verb...

The noun classes I have are masculine, feminine, earth, water, fire, air, metal, wood, aether, light and dark. (Yeah I know, but it is a magical language.) I guess one could use aether as the default noun class because it stands for magical and non-material things. But I don't quite like doing that. :/

linguoboy wrote:Infinitives belong to Bantu noun class 15 (prefix *ku-), e.g. kuishi kwema "live-INF good".

Is that class associated with any kind of thing, or is it just something abstract?

Opera wrote:There's also the possibility of assigning the gender based on the phonological form of the word. For example in one of my conlangs, all nouns ending in a fricative are feminine (root only, not counting suffixes).

Hmmm, I guess I could do something like that, but I don't know... :/

Opera wrote:In another, there's a class which is basically a bin for cases like this. Maybe you could simply say that one of your classes also is the default class for compound words like this.

Yeah, Whimemsz also suggested that.

Okay, I'm gonna see if I can do without nounified verbs, otherwise I will have to choose one of the suggestions on this thread. Unless someone knows yet another solution...

I don't know if this is this has any precedent, but you could possibly assign verbs to these classes as well; while it probably wouldn't show in them while they were verbs, all of the more abstract nouns derived from the verb would have the same class. As an actual example, Latin's infinitives are all neuter, to the extent of my knowledge, and I believe the supines and gerunds are as well. Most nouns of the state/quality derivation, on the other hand, are feminine, including virtūs, virtūtem, manliness.

Qwynegold wrote:

linguoboy wrote:Infinitives belong to Bantu noun class 15 (prefix *ku-), e.g. kuishi kwema "live-INF good".

Is that class associated with any kind of thing, or is it just something abstract?

It's associated with infinitives.

Jetboy wrote:I don't know if this is this has any precedent, but you could possibly assign verbs to these classes as well; while it probably wouldn't show in them while they were verbs, all of the more abstract nouns derived from the verb would have the same class. As an actual example, Latin's infinitives are all neuter, to the extent of my knowledge, and I believe the supines and gerunds are as well. Most nouns of the state/quality derivation, on the other hand, are feminine, including virtūs, virtūtem, manliness.

Oh yeah, that's out of the question because it would be a pain in the ass. But if one could divide all verbs into a few different types where it is very obvious which verb belongs to which type, then one could assign noun classes to those types.

linguoboy wrote:

Qwynegold wrote:

linguoboy wrote:Infinitives belong to Bantu noun class 15 (prefix *ku-), e.g. kuishi kwema "live-INF good".

Is that class associated with any kind of thing, or is it just something abstract?

It's associated with infinitives.

Oh, okay.

Anyhow, I think I've come up with a solution. I shall have an infinitive marker, a postposition called eengchoe. I can assign this infinitive marker together with the infinitive to the dark class, because the dark class contain, among other things, things of non-existance. The ee in eengchoe means inactive, and inactivity is a kind of non-existance. Thanks everyone for the replies!