Polypersonalism

[Chengjiang]

I realized that polypersonalism features in more than one of the planned conlangs I have and that I don't actually know much about languages that feature it. What are common features of polypersonal languages? What also tends to be true about a language with this feature? I mean, I can surmise some things (e.g. it is likely to be pro-drop), but I want to be sure I know what I'm doing here. All I currently know on the subject is a smattering of information on Basque, Georgian, and Bantu languages.

[roninbodhisattva]

By "polypersonalism" are you referring to languages where multiple arguments in a clause trigger agreement on the predicate?

[Chengjiang]

By "polypersonalism" are you referring to languages where multiple arguments in a clause trigger agreement on the predicate?

Yes. I thought that was the standard definition of the term.

[roninbodhisattva]

By "polypersonalism" are you referring to languages where multiple arguments in a clause trigger agreement on the predicate?

Yes. I thought that was the standard definition of the term.

Well, I'm more used to the term "polypersonal agreement" than "polypersonalism", actually I've never seen the later used in the literature I read.

Anyway, back to the original question. I think one has to be very careful about generalizations that one makes over languages that cross-reference multiple arguments on the verb. Part of this has to do with the variety of ways that morphemes that look very similar on verbs can actually differ significantly. So for example in Bantu, many languages are able to cross-reference both subject and object morphologically on the verb, but only sometimes does object morphology behave like "agreement" (in many languages object markers have the distribution of pronouns, for example), while subject morphology always does. And this varies within Bantu.

*shrugs* I don't know, that makes it tough.

[Chengjiang]

in many languages object markers have the distribution of pronouns, for example

By this do you mean that they are used, like pronouns, only if an explicit object constituent is not present? Rather like French cliticized object pronouns, for instance?

[roninbodhisattva]

in many languages object markers have the distribution of pronouns, for example

By this do you mean that they are used, like pronouns, only if an explicit object constituent is not present? Rather like French cliticized object pronouns, for instance?

Yes, so they can't occur when there is an explicit NP object and leaving them out removes reference to an object, so kind of like French object clitics, yes. But note that there are Romance languages where those clitics can double explicit object NPs so it gets murkier.

[linguoboy]

Aren't there also varieties of Bantu where the inclusion of object markers expresses a definiteness distinction?

I'm most familiar with Osage, which seems a fairly typical polypersonal language with the possible caveat that it shows a marked tendency toward redundant inflection. For instance, the predicate ðahkišpižǫi "you learned it for yourself" is triple-inflected for a 2S agent. (ða, š, and ž are all allomorphs of the same agent marker. Cf. ahkihpimǫi "I learned if for myself".)

[zompist]

What are common features of polypersonal languages? What also tends to be true about a language with this feature? I mean, I can surmise some things (e.g. it is likely to be pro-drop), but I want to be sure I know what I'm doing here.

I'm not sure there are many common features. You could take a look at the sketch of Quechua on my site; also spoken French.

Polysynthetic languages usually have polypersonal agreement, but you can have polypersonal agreement and still be agglutinative (Quechua, Bantu) or fusional (French).

[Astraios]

There really isn't anything useful to say without looking at specific languages. How could Neo-Aramaic and Lakota (for example) be reasonably considered as having anything in common apart from agreement with multiple persons?

[jal]

How could Neo-Aramaic and Lakota (for example) be reasonably considered as having anything in common apart from agreement with multiple persons?

Well, there could be some universals governing these type of constructions, without the languages themselves being related.


JAL

[Chengjiang]

How could Neo-Aramaic and Lakota (for example) be reasonably considered as having anything in common apart from agreement with multiple persons?

Well, there could be some universals governing these type of constructions, without the languages themselves being related.


JAL

More or less what I was thinking of. I was wondering if there were any tendencies found in unrelated languages, similar to how, say, VSO languages tend to also be noun-adjective.