Chagen wrote: ↑14 Oct 2020 02:02
I find it odd that you're confused,
Update: reading this post, I'm now even more confused!
but I think I've figured out what's tripping you up: when I talk about participles agreeing with particular nouns, I'm referring to them agreeing with them in declension and case, similar to how adjectives in many European languages agree with them.
Yeah, sure. Got that.
[although:
are there European languages in which participles always take the declension of the noun they modify? I'm sure I'm just missing something, but this seems a little odd to me. Agreement in number, case and gender is of course perfectly ordinary]
Part of the reason this isn't obvious is because all of the nouns in my example are the same declension. Lemme provide you with the example sentence, but switching out "criminal" (a 1st-declension word) for "girl" (a 2nd-declension one):
Nayakā yīdasāya, cṛsūrē uźāmi qiḥāya kansēyṛtūrē, vīghevyū
NAME-NOM dagger-DEF.SG-ACC girl-DEF.SG.NOM REL.PRO-INSTR man-DEF.SG-ACC die-CAUSE-PTCPL.PST.ACT-2ND-DEF.SG.NOM find-PERF-3S
"Nayaka has retrieved the dagger with which the girl killed the man"
You will note here that the participle agrees with "girl" (cṛsūrē) and NOT "dagger" (yīdasāya).
Yeah, of course. Why would it ever agree with "dagger"? The participle is modifying its subject, just as in English.
The whole clause may be modifying "dagger", but the actual participial head of the clause agrees with its actual subject, "girl", in both declension and case.
Well yeah, of course. Just the same as in English. [well, English doesn't have to agree in 'declension', and it happens not to use a participle in relatives, but otherwise...]
So, in English, using present tense to induce non-zero agreement marking:
Nayaka sees the dagger, with which the criminal stabs the man (obviously 'is stabbing' is more natural, but shows the same agreeement)
vs
Nayaka sees the dagger, with which the criminals stab the man (again, same for 'are stabbing')
So, 'stab' agrees with the subject, 'criminal', not with 'dagger', the antecedent of the clause.
Other than that your language uses participle clauses for relatives (whereas English only uses them in an adverbial sense), and that modern English has lost all its agreement on participles by now, Pazmat seems to have exactly the same structure as English. It seems to me that you could construct, as Ser suggests, exactly the same construction in Latin, or to some extent even in literary Old English (it's debated to what extend absolute participle constructions were used in ordinary speech, and to what extent they were just a stylistic imitation of Latin in high-register and translated texts, but let's not quibble over details...).
This is why I'm confused. You're obviously aware that languages can have more or less non-zero agreement marking on participles, so that's presumably not what you're asking about. And you seem to be aware that languages can certainly use deranked constructions, including participles, in relative clauses, so I assume that's not what you're asking about either. And all the rest seems exactly the same as in English, so I presume you're not asking about any of that either. So what's the issue you'd like our opinion on?
There is a relative pronoun that refers to the dagger, of course, to explain what it's doing in the relative clause (being used as an instrument).
Well, not really 'of course', no - that's just how English happens to do it. WALS' survey came up with only 13 languages that do this, compared to 99 that don't. 12 of those 13 are from Europe (well, one is Georgian, but close enough).
Perhaps it would be clearer if you laid out:
- how Pazmat would translate a direct relative, like "Nakaya saw the criminal who killed the man" (A)
Nayakā swotāya, qiḥāya kansēyṛtāya vēttavyū.
NAME.NOM criminal-DEF.SG-ACC man-DEF.SG-ACC die-CAUS-PTCPL.PST.ACT-1ST-DEF.SG-ACC see-PERF-3S
"Criminal" is modified by a participial phrase whose participle agrees with it in both noun declension and case. Literally, this is "Nayaka saw the criminall, (he) having killed the man"
OK, pretty much like in English, yes. That makes sense.
- how Pazmat would translate an indirect relative, like "Nakaya saw the dagger with which the criminal killed the man" (B)
Nayakā yīdasāya, swotā uźāmi qiḥāya kansēyṛtā, vēttavyū
NAME-NOM dagger-DEF.SG-ACC criminal-DEF.SG.NOM REL.PRO-INSTR man-DEF.SG-ACC die-CAUSE-PTCPL.PST.ACT-1ST-.DEF.SG.NOM see-PERF-3S
In this sentence, the participle of the relative clause does not agree with the noun it is modifying in case. That's because it agrees with its own subject, "criminal" and thus is in the nominative case, whereas with the previous example, it agreed with the noun it was modifying because said noun was actually the subject of the clause.
So, the same as before, yes.
OK! So, the difference between these two sentences seems to be that when you relativise on an oblique, you need a relative pronoun, whereas when you relativise a subject, you can drop the pronoun. Is this the difference you're asking about?
If so, yes, this is perfectly ordinary. WALS suggests that Spanish does this? [English isn't far off either - direct relatives can be replaced by participial constructions ("I saw the man (who was) chasing you"), but indirect relatives can't, and there's an optional special pronoun-less constuction only when its the object that's the antecedent ("I saw the man (who) you married"). Pronoun dropping in certain sorts of clause but not others is something many languages do.
(btw, it's intuitively easier to understand glosses if you only include the parts that are relevant - like, you don't need to wite 'sg' for every singular, or 'def' for every definite, when we're not talking about those things; likewise, you can just say 'kill', you don't have to say 'die-CAUSE' every time...)
Put another way: participles in Pazmat agree in declension and case with their own subjects
Sure, of course.
, which usually are the same as the actual noun they are modifying,
Almost by definition, yes
but in this instance (a relative clause where the modified noun is an oblique), it isn't.
Huh? Now I'm confused again. The modified noun here is not an oblique, it's the subject. "killing" modifies "criminal", not "dagger". How could it modify "dagger"? In what language would it ever modify "dagger"?
[well, you could have a language where it did, where every relative was an adjectival, and there was incorporation, and "the dagger with which the criminal killed the man" was always translated "the man-killing dagger of the criminal", but that doesn't seem to be how Pazmat (or English) does it]
Similarly, in English, "kill" agrees with "criminal", not with "dagger" (as the above examples with number alternations demonstrate).
Perhaps you're confusing "verb" with "clause"?
The
verb, which is
inside the clause, modifies the
subject of
the relative clause. [In Pazmat, the verb is participial in form.]
The
relative clause as a whole modifies
the antecedent, which can be
any noun (in this case an object) in
the matrix clause.
But Pazmat seems to behave exactly the same way that English does in this regard, and that most languages do. Indeed, the only slight oddity seems to be that when the antecedent co-refers with the object of the relative clause, the relative clause uses the passive (in English, both the passive and the active are allowed, although only in the active is the relative pronoun droppable ("I saw the man (who) the muggers stabbed", vs "I saw the man who was stabbed by the muggers").
- how Pazmat WOULD translate B IF it did NOT have the restriction you're discussing (C)?
Absolutely nothing would change except the participle would be in the accusative to agree with "dagger".
But... why? Why would that ever happen? "Dagger" isn't the subject of the verb!
See, I think what's confusing me is that you're effectively saying "on Tuesdays, water runs downhill; God has put in place a special rule to make water run downhill specifically on Tuesdays". And I'm left saying: but water always runs downhill, on every day of the week. What would be really weird would be if, on Tuesdays only, it
didn't run downhill!
So, both Pazmat and Standard Average European have verbs agree with their subjects. Both Pazmat and SAE specifically have verbs agree with their subjects inside relative clauses. So far as I can see, both Pazmat and SAE have rules whereby the verb only EVER agrees with the subject.
But if the verb agreed with "dagger", which isn't only
not the subject, but
isn't even in the same clause!, then this would be something that doesn't happen anywhere else in the language. Either in Pazmat or in SAE.
So given that the behaviour when the antecedent of a relative is oblique within the relative is
exactly the same as when it's not oblique (other than the obligatory pronoun), and given that this behaviour is
exactly the same as in a standard average European language, including English... I'm confused why you're asking if it's naturalistic! And why you're describing with phrases like "the relative clause does not modify 'dagger'", and "it almost sounds more like..." and the like...
Not trying to be rude. It's just that I think either you must be misunderstanding something or I must be, and I'm not entirely clear which it is!