Radius Solis wrote:What do you call a language that lacks morphosyntactic alignment in any real sense at all?
I call it "interesting".
Radius Solis wrote:My old project Naidda does not mark nouns for their roles, not by morphology and not by word order (excluding datives, for which there is a preposition). In transitive sentences, agents and patients are both agreed with by the verb, and this (plus context) is the main way they are told apart. There is a single set of fused subject-object agreement suffixes. For example, a sentence could be glossed "hit-AGR he I" and the AGR suffix is what tells which hit the other. -ai marks 3sg subject, 1st person object, so if the sentence is "hit-ai he I", he hits me. But if it's "hit-j@ he I", it's me that hits him.
That's neat! And I think it's an ANADEWism (that is, A Natlang Already Does that (Even Worse?)).
Radius Solis wrote:But if, for example, both the agent and patient are 3sg, ambiguity results and context helps determine which is doing what to the other. In the rare event even context fails, an intransitive clause can be brought in to clarify. "Hit Juan Pedro. Hit Juan." (where the intransitive subject is the agent of the former sentence)... but again, this is very rarely needed.
I've read of a South American language like that.
Radius Solis wrote:With intransitive verbs these suffixes don't occur and there's a different set of agreement suffixes for the single argument that bears little morphological resemblance to anything in the transitive set.
So I suppose you could call it tripartite, in the sense that S, A, and P are all dealt with separately, but really I find it a stretch to try to fit Naidda into such a mold. I prefer to analyze it as morphosyntactically unaligned.
No, I've seen this in a couple of Burmese languages and in some Iranian language as well. (The only one of them I can think of the name of is "Lolo"; and maybe that's wrong. Sorry.)
The Original Poster probably included a diagram for this system. Anyway I've seen such a diagram several times. I'll have to work to find the label that went with it, though.
Radius Solis wrote:Now, among natlangs, the system sometimes called "animacy hierarchy" is also morphosyntactically unaligned in some languages (though more often it coexists with or is part of an alignment system).
Even in the languages I mentioned above, which are not Hierarchical nor Inverse/Direct, if it happens that one of the 3sg participants is animate and the other isn't, then, depending on the verb, it's probably going to be quite likely that the animate one is the agent and the inanimate one is the patient. "Eat-AGR3sg3sg George Sandwich" is unlikely to be interpreted as "The sandwich is eating George".
Radius Solis wrote:But I don't know of any natural languages that fail to have either an animacy hierarchy or an alignment.
I'm pretty sure there are at least the four I mentioned above.
But you should know that there are situations in which Hierarchical MSA languages have trouble; specifically,
- when both of, or neither of, the participants in the transitive clause are Speech Act Participants (Speaker or Addressee), and also
- both of, or neither of, the participants in the transitive clause are Animate.
Since your language has the verb agree with the persons of the agent and patient, you won't have problems when one of them is 1st person and the other is 2nd person. But you'll still have problems whenever a transitive clause's participants are two 3rd-person inanimates or two 3rd-person animates of the same grammatical number. Most Hierarchical MSA languages solve that problem via "obviation". But that South-American language I mentioned above would have to disambiguate exactly as you said; "Hit Juan Pedro. Hit Juan.", unless, in context, it wasn't ambiguous for narrative or discursive or pragmatic reasons.
[EDIT]:I am sorry, I cannot find that diagram. I think it was in an article of Dryer's. Again, I could be wrong.[/EDIT]