Very good, IMO.
So of course I'm only going to point out the flaws.
PTSnoop wrote:Məmiárua tnrikir nedu hirkatwki.
ERG-Miárua DECH-stone window 3.INANIM-throw.
Miárua threw a stone at the window.
Actually, "Miárua threw a stone to
"At the window" would be allative, a kind of locative.
You want a recipient, rather than a target.
PTSnoop wrote:.... more neat stuff not quoted in order to save space ....
PTSnoop wrote:*Tnrinedu Məmiárua kir qerkatwki.
DECH-window ERG-Miárua stone 3.INANIM-throw
It was the stone that Miarua threw at the window.
(But what this actually says is "Miárua threw the window at the stone").
Either interpretation is bad news for the window.
PTSnoop wrote:So let's introduce an applicative voice, which promotes the DECH argument to ABS and pushes the current ABS into a locative.
So let's introduce an applicative voice, which promotes the DECH argument to ABS and pushes the current ABS into a locative.
Czedanedu Məmiárua kir hirkatwkilri.
LOC-window ERG-Miárua stone 3.INANIM-throw-APPL
It was a stone that Miárua threw at the window.
That wouldn't be an applicative voice; that would be a circumstantial voice.
Applicative voice would promote an oblique argument to the primary-object position; in dative languages that's the position the direct object occupies, but in dechticaetiative languages, in default-voice ditransitive clauses, it's the position the recipient occupies.
Circumstantial voice promotes an oblique argument to the subject position (in accusative/nominative languages) or absolutive position (in ergative/absolutive languages).
Appplicative voices usually promote some oblique argument to the primary-object position.
Indirect objects, in dative languages, and secondary objects, in dechticaetiative languages, are core arguments rather than oblique arguments. But they're "closer" to "oblique" than either of the other two core arguments.
Is "dative movement", in dative/accusative/nominative languages, a kind of "applicative voice"? "Dative movement" in such languages promotes the indirect object to the direct-object position. Wikipedia seems to thing "dative movement" is a kind of applicativization.
If "dative movement" is applicative, then maybe what you describe is not too far from applicative.
Some applicative-voice morphology on the verb only says that some kind of applicative promotion takes place. OTOH, some gives a hint as to which former oblique argument is now the primary object. (For instance, you might have four applicative-voice morphemes to affix to a verb; one to tell you the location has been promoted to direct object, one to tell you the instrument has been promoted to direct object, one to tell you the beneficiary has been promoted to direct object, and one to tell you any other oblique argument has been promoted to direct object.)
Did you give enough examples to tell which kind happens in your language? It looks like whether the oblique argument that gets promoted is an instrument or a location, the verb still has the same applicative-voice suffix attached. Or did I read that wrong?
Are you promoting oblique arguments, such as location or instrument, to the secondary object position? If so that's not applicativization exactly, though it is a very similar thing. Applicativization promotes oblique arguments to the primary-object position.
PTSnoop wrote:And now, let's cram all the cases into one big applicative sentence for a causative construction.
Czedanedu Tnrimiárua məkåprwhpåkə kir hirkatwkilri.
LOC-window DECH-Miárua ERG-rage stone 3.INANIM-throw-APPL
Her rage made Miárua throw the stone at the window.
(or "rage used Miárua to throw the stone at the window").
I believe you're on a road parallel-ish to the right track.
But let me explain why I thought this challenge would be challenging.
First, why combining Ergative with Dechticaetiative is challenging.
"Ergative/Absolutive" is usually a term used to label languages that treat the Patient of a prototyical Transitive clause the same way as the Subject of an Intransitive clause ("Absolutive"), and use some other treatment ("Ergative") for the Agent of a prototypical Transitive clause. It contrasts with (among other possibilities) "Accusative/Nominative", a term used to label languages that treat the Agent of a prototyical Transitive clause the same way as the Subject of an Intransitive clause ("Nominative"), and use some other treatment ("Accusative") for the Patient of a prototypical Transitive clause.
For Accusative/Nominative languages, "Dechticaetiative" is usually a term used to label languages that treat the Recipient of a prototyical Ditransitive clause the same way as the Patient of a prototypical Monotransitive clause ("Primative"), and use some other treatment ("Secundative") for the Theme of a prototypical Ditransitive clause. It contrasts with "Dative", which is usually a term used to label languages that treat the Theme of a prototyical Ditransitive clause the same way as the Patient of a prototypical Monotransitive clause ("Accusative"), and use some other treatment ("Dative") for the Recipient of a prototypical Ditransitive clause.
But how would the cases/treatments line up for a language that's Ergative and Dechticaetiative?
The most syntactically privileged Morphosyntactically-assigned Argument Position (or Grammatical Relation) would be S=P=R; the Subject of a prototypical Intransitive clause, the Patient of a prototypical Monotransitive clause, and the Recipient of a prototypical Ditransitive clause, would all take the Absolutive treatment (or case).
The second-most privileged M.A.P. (or G.R.) would be A=D; the Agent of a prototypical Monotransitive clause and the Donor/Exhibitor/Narrator of a prototypical Ditransitive clause would both take the Ergative treatment (or case).
There would be a third G.R.; prototypical Ditransitive clauses would have a third core argument, the Theme (entity located or moved). This would take a Secundative treatment or case.
Or anyway that's what I had in mind.
You can see the difficulty, though; if a language is both Ergative and Dechticaetiative, does that mean S=P=R, A=D? Or what?
Ergative and Dative combine easily; S=P=T Absolutive, A=D Ergative, R Dative.
Accusative and Dechticaetiative combine easily; S=A=D Nominative, P=R Primative, T Secundative.
But can you be sure you're doing what you should be doing when you try to combine Ergative with Dechticaetiative?
Second, why combining Applicative Voices with any system with three or more GRs is challenging.
I thought of "Applicative Voice" as any way of promoting a non-Core item (i.e. an Oblique item) which was nevertheless an Argument rather than an Adjunct, into the second-most-privileged Core-Argument position; that is, into the Primary Object position (the most-privileged GR being "the Subject" and all other GRs being "Objects"; the most-privileged Object being "the Primary Object" and the/all other Object(s) being "Secondary Object(s)".)
If there is a third GR, that is, a Secondary Object (still a Core Argument, but not the Subject and not the Primary Object), then Applicative Movement must somehow leapfrog the promoted Oblique Argument over this Secondary Object Position to put it into the Primary Object Position. (What it does the the erstwhile Primary Object -- if there was one -- may vary. Maybe it will be demoted but remain in the Core, becoming a Secondary Object. Maybe it will be demoted all the way out of the Core, becoming an Oblique Argument. Maybe it will be so demoted it becomes implicit and is not spoken of.)
IMO this could be a challenge. There may be a different challenge in Applicativizing an Intransitive clause, from Applicativizing a Monotransitive clause, from Applicativizing a Ditransitive clause.
Third, why combining Applicative Voices with Ergativity might be challenging.
In a strictly morphologically-and-syntactically Ergative language (if there were any such natlangs), the Absolutive would be the syntactically-most-privileged GR and the Ergative would be the second-most-privileged GR.
But among natlangs Applicativization is usually described as promoting an Oblique Argument to the Direct Object slot.
OK; in a purely Ergative language's prototypical Monotransitive clause, which participant counts as the Direct Object?
The syntactically most-privileged Core Argument is the Absolutive.
The second-most-privileged Core Argument is the Ergative.
Depending on whose definitions you use, that would make the Absolutive be the Subject, and would make the Ergative be the Primary Object.
Does Applicativization, in Ergative languages, mean promoting an Oblique Argument to the Ergative slot?
Fourth, additional reasons combining Applicative Voices with Dechticaetiativity might be challenging.
Among natlangs Applicativization is usually described as promoting an Oblique Argument to the Direct Object slot.
But Dechticaetiative languages don't have Direct Objects. Instead they have Primary Objects and Secondary Objects.
The second-most-privileged Core Argument is the Primary Object.
In an Accusative/Nominative language that was also Dechticaetiative, we would have S=A=D (Nominative) and P=R (Primative or Accusative) and T (Secundative). The Primary Object, Patient of Monotransitives and Recipient of Ditransitives, would be the second-most-privileged GR. I was expecting and intending that Applicativization, in such a language, would promote an Oblique Argument to the Primary Object (P=R) slot.
In an Ergative/Absolutive language that was also Dative, we would have S=P=T (Absolutive) and A=D (Ergative) and R (Dative). The Ergative, Agent of Monotransitives and "Donor" of Ditransitives, would be the second-most-privileged GR.
In an Ergative/Absolutive language that was also Dechticaetiative, I think we would have S=P=R (Absolutive) as the most-privileged GR, and A=D (Ergative) as the second-most-privileged, and T (Secundative) as the 3rd-most-privileged. The Ergative, Agent of Monotransitives and "Donor" of Ditransitives, would again be the second-most-privileged GR.
I expected/intendid Applicativization, in such a language, to promote an Oblique Argument to the Ergative slot.
I can't decide whether IMO promoting the Secondary Object to the Ergative slot should also count as Applicativization in such a language; perhaps it should.
I hope some of that helps clear up some of the challenge?