Frislander wrote:Random idea as to a verb complex.
The verb phrase comes in two main parts: the auxiliary and the main verb. The auxiliary comes from a restricted set of semantically bleached verbs ("be", "go", "do" etc.) and is the part of the complex which can stand on its own. Each auxiliary encodes the valency of the whole verb complex, as well as taking polypersonal marking for the whole verb complex. The main verb selects its auxiliary based on its own inherent valency. This main verb encodes most of the other verbal categories: aspect, mood and evidentiality, which are all mixed into one giant heap of combined forms. Aspect is the most lexicalised of these, with each verb having its own inherent Aktionsart. Perfective verbs display the greatest number of evidentiality distinctions: seen first-hand, first-hand from another sense, reported and deduced from secondhand evidence. Imperfective and stative verbs only make a direct-indirect distinction. These evidentiality markers may instead be replaced by an optative marker pr subordinating affix, or they may coexist with an interrogative marker, where the evidential indicates the expected source of knowledge for the answer.
To me, this sounds more like light-verb + lexical content-word, rather than auxiliary word + main verb.
(Although it isn't
purely either of those.)
Light verbs are semantically "light" rather than semantically "empty"; they don't have much lexical content, but they have some. This may be because they have been semantically "bleached".
Since light verbs can stand by themselves, they have -- or get marked for -- all verbal categories (including valence, agreement, aspect, mood, evidentiality, illocutionary force (e.g. interrogative), polarity, mirativity, pluractionality, validationality) for which any of the language's verbs get marked (either lexically inherent, or derivational, or inflectional, or syntactic, or with an auxiliary word).
If lexical content-words can't stand by themselves as verbs -- that is, as nuclei of clauses -- they syntactically "aren't verbs", in a manner of speaking.
Verbal auxiliary-words, OTOH, are words that tell any of the following (and possibly other) verbal accidents of some syntactically related verb; aspect, modality/mode/mood (possibly including for instance realis/irrealis, declarative/interrogative/imperative, main clause vs subordinate clause), polarity (affirmative or negative), tense, and/or voice.
If an auxiliary word does not inflect, it's a particle, and its distribution and other syntactic behavior may be idiosyncratic.
But if it inflects, it usually behaves syntactically as if it were a verb, and the main verb were its object. This often happens because it is diachronically derived (via semantic bleaching) from a biclausal construction in which the now-auxiliary-word's ancestor was the verb of the main clause, and the now-main-verb's ancestor was the nucleus of a subordinate complement clause.
In your setup what you call "the auxiliary" can "stand on its own" (I assume that means, can be the nucleus of a clause), so it deserves to be called some kind of verb; since it's semantically bleached, it counts as a "light verb". It either has an inherent valency, or can be marked for valency; and it inflects for (polypersonal) agreement with (some -- at least two if there are two) of its participants. So it's a verb.
What you call "the main verb", has some inherent verb-accidents, in particular including valency and aktionsart (lexically-inherent aspectual class); and inflects for aspect, mood, and evidentiality. Apparently it can also inflect for optativity and for subordination, and can take an interrogative auxiliary particle -- unless that particle applies rather to the light-verb, or to the whole clause, instead of to "the main verb".
All of that makes it sound very verb-like.
But if it can't stand on its own as the nucleus of a clause, it can't be completely verb-like, can it?
I'd say the "verbishness" is distributed over the entire light-or-auxiliary-verb + main-verb complex.
[hr][/hr]
This reminds me a lot of a "Tibetan Dwarvish" conlang one of us (Vardelm) worked on, in which verb-complexes were simultaneously both lightverb+contentword and auxiliaryword+mainverb. The light verb carried all the grammatical marking, including all the verbal accidents (and thus also counted as "the auxiliary"); the content-word carried almost all of the lexical, semantic content.
Edit: Also see Ossicone's "Hjusan".
[hr][/hr]
I have more questions, because your idea sounds interesting. I should probably wait and read your next post, though.