relative clause strategy

[Nemesis]

Does this strategy for creating relative clauses make sense to you? Is there any natlang doing this?
"In ckoerrolu relative clauses are formed by a correlative strategy (I don't know if this is the correct term), with the relative pronoun "vi" that is declined in case and number to reflect the functions of the upper clause: subject (vi/vir), direct object (vit/virt), indirect object (vitün/virtün), and so on. The relative clause precedes the noun it modifies, just like adjectives, and the relative pronouns comes in the first place. (note: Ckoerrolu is an SOV language, and adjectives come before the noun).

Ex_1: Vi paziher bogodos, mede mora sös.
That.NOM keys brought, man.NOM kind is.
The man that brought the keys is kind.

Ex_2: Virt žemak mygajav, pazihert bogodu.
That.ACC door open., keys.ACC I brought.
I brought the keys that open the door.

Ex_3: Virtün acarados, medetün pazihert bogodu.
That.DAT arrived, man.DAT keys I brought.
I brought the keys to the man that arrived.

[Khunjund]

It kind of reminds me of Quechua, in which relative clauses are transformed to take the form of a participle which modifies the main clause question.

This system seems perfectly intuitive to me, but in this case, and considering your language is head-final (adjectives, etc. before the nouns they modify), wouldn't it make more sense to directly chain together the relative clause and subject, rather than seperate it with a comma, especially considering your relative pronoun is declined like the modified clause?

Ex.

Virtün acarados medetün pazihert bogodu.

where your subject, object and verb are

[Virtün acarados medetün] [pazihert] [bogodu]

[Frislander]

It's not a correlative relative clause because it's externally headed; it's just a plain prenominal relative clause which uses the "pronoun retention" strategy ( see WALS Chapter 90) which is unusual, since all of the prenominal relatives I know of use a gapping strategy. A true correlative would look like this:

Mede paziher bogodos, vi mora sös.
man.NOM keys brought that.NOM kind is
The man that brought the keys is kind.

The most weird thing for me, though is the case agreement of the internal pronoun with the head noun. This makes no sense to me, because pretty much the whole point of retaining the pronoun is to show the role the head noun fulfills within the relative clause. Even the Indo-European relative pronouns strategy shows the head nouns role within the relative clause even though it's actually outside of it.

So by this logic your second and third examples would look like this:

Vi žemak mygajav, pazihert bogodu.
that.NOM door open keys.ACC I brought
I brought the keys that open the door.

Vi acarados, medetün pazihert bogodu.
that.NOM arrived man.DAT keys I brought
I brought the keys to the man that arrived.

Your strategy might make sense if relativisation is only possible when the head noun is the syntactical subject of the relative clause, but then in that case I have to ask why you even bother with having the pronoun there at all and don't use a gapping strategy instead?

It kind of reminds me of Quechua, in which relative clauses are transformed to take the form of a participle which modifies the main clause question.

And it's a nice strategy, but how is this relevant to the point in question?

This system seems perfectly intuitive to me, but in this case, and considering your language is head-final (adjectives, etc. before the nouns they modify), wouldn't it make more sense to directly chain together the relative clause and subject, rather than seperate it with a comma, especially considering your relative pronoun is declined like the modified clause?

Ex.

Virtün acarados medetün pazihert bogodu.

where your subject, object and verb are

[Virtün acarados medetün] [pazihert] [bogodu]

That's a matter of punctuation, not syntax per se, and it really depends on taste: note that German uses a comma with all subordinate clauses, even relative clauses.

[clawgrip]

I struggled to find justification for this unusual system, and I think I've come up with something nice, though it involves imposing rules on the language that probably weren't there before. Here's my idea:

The language has only one relative pronoun (vi). This relative pronoun is not a bound pronoun as it has been described (i.e. "that"); rather, it is a free relative pronoun (i.e. "that which"). Right away, this removes the relative pronoun from the clause itself, explaining why it declines according to the referent's position in the main clause rather than the case of the gap in the relative clause.

All relative clauses must be headed by relative pronouns; there being only one relative pronoun, vi, all relative clauses naturally must begin with vi. Consequently, nouns cannot be heads of relative clauses. The only way for a relative clause to modify a noun is through apposition of the relative pronoun and the referent noun. Conveniently, there is a comma separating the two elements of apposition, just as is done in English (cf. "chromic acid, a highly corrosive substance" and "that which brought the keys, the man"). Through convention, the order of constituents gravitated towards a system where the shortest of the two comes last (which seems nice in a head-final language).

Now, how could such a weird system where there are no bound relative pronouns have come about? My thoughts:

Originally, in some older form of the language the system was the same as in English. Using the same vocabulary:

Mede vi paziher bogodos, mora sös.
man.NOM REL keys brought, kind is.
"The man that brought the keys is kind."

Pazihert vi žemak mygajav, bogodu.
keys.ACC REL door open, I brought.
"I brought the keys that open the door."

At some point, an innovation occurred whereby it became possible to drop the head noun, with the relative pronoun itself implying its existence, or actually replacing it. This has precedence in e.g. Classical Japanese, where an attributive form of a verb (essentially a relative clause) can occur without its head noun, taking only the case particle and implying a generic 3rd person, e.g.:

Kikazaru hito ga ōshi.
listen-NEG-ATTR person NOM be.many
"There are many people who do not listen."

Kikazaru ga ōshi.
listen-NEG-ATTR NOM be.many
"There are many who do not listen."

It should be noted also that the English translations are pretty much doing the same thing, except that with a pronoun instead of a verbal suffix.

So this gives us these as valid sentences:

Vi paziher bogodos, mora sös.
REL.NOM keys brought, kind is.
"The one who brought the keys is kind."

Virt žemak mygajav, bogodu.
REL.ACC door open, I brought.
"I brought that which opens the door."

At some point this became established as a free pronoun, and it could no longer be used as a bound pronoun. The referent noun could be placed in apposition and, as mentioned above, eventually settled in second position since it is shorter. This results in the sentences that appear in the OP.

How does this sound?

[lsd]

I used a similar construction in araedeilẹ...
The linear language describing 3D writing can pronounce different branching in any order in absence of focus...
Everything can be seen as a relative clause on the other

[k1234567890y]

This works (: and has someone pointed out, it is a pronoun-retention strategy.

https://en.wikipedia.org/wiki/Relative_clause

And as it works, just go for it (:

[Nemesis]

Thanks a lot guys, it really helped!

I think I'll stick with the gapping strategy, with a prenominal position, and a relative pronoun that agrees only in number with the head noun, but retains the case of the head noun. Example:

Vir žemakt mygajav paziher, bogodu.
that-NOM-PL door-ACC open-3P key-PL, I brought.
I brought the keys that open the door.

Vitün acarados mede, pazihert bogodu.
that-DAT arrived man, keys I brought.
I brought the keys to the man that arrived.´

Does it make sense now? Or should the relative pronoun not decline at all by case?

[Frislander]

Thanks a lot guys, it really helped!

I think I'll stick with the gapping strategy, with a prenominal position, and a relative pronoun that agrees only in number with the head noun, but retains the case of the head noun. Example:

Vir žemakt mygajav paziher, bogodu.
that-NOM-PL door-ACC open-3P key-PL, I brought.
I brought the keys that open the door.

Vitün acarados mede, pazihert bogodu.
that-DAT arrived man, keys I brought.
I brought the keys to the man that arrived.´

Does it make sense now? Or should the relative pronoun not decline at all by case?

Well seeing as you have a pronoun of some sort actually inside the relative clause fulfilling the role of the head noun it's not gapping any more, unless you want to try and argue that it's some kind of external relative pronoun, in which case why isn't it next to the head noun. Other than that, I see no problem.

Further question for others to answer: do we know of any natlangs with prenominal relative clauses which also use relative pronouns?

[Creyeditor]

Thanks a lot guys, it really helped!

I think I'll stick with the gapping strategy, with a prenominal position, and a relative pronoun that agrees only in number with the head noun, but retains the case of the head noun. Example:

Vir žemakt mygajav paziher, bogodu.
that-NOM-PL door-ACC open-3P key-PL, I brought.
I brought the keys that open the door.

Vitün acarados mede, pazihert bogodu.
that-DAT arrived man, keys I brought.
I brought the keys to the man that arrived.´

Does it make sense now? Or should the relative pronoun not decline at all by case?

Well seeing as you have a pronoun of some sort actually inside the relative clause fulfilling the role of the head noun it's not gapping any more, unless you want to try and argue that it's some kind of external relative pronoun, in which case why isn't it next to the head noun. Other than that, I see no problem.

Even if it's not next to the pronoun it could be an external relative pronoun, if you say it always appears in first position of the complex noun phrase.

Further question for others to answer: do we know of any natlangs with prenominal relative clauses which also use relative pronouns?

I guess people would call it a participle prefix, if they encounter it. Maybe they would even invoke noun incorporation, depending on the phonological facts.

[Nemesis]

Well seeing as you have a pronoun of some sort actually inside the relative clause fulfilling the role of the head noun it's not gapping any more, unless you want to try and argue that it's some kind of external relative pronoun, in which case why isn't it next to the head noun. Other than that, I see no problem.

It's still a gapping strategy. English has relative pronouns but there's still a gap:

I ate the cake that my mom baked [the cake].