Active alignment, head-marking, aktionsart, version, and other stuff

A forum for discussing linguistics or just languages in general.
Post Reply
User avatar
eldin raigmore
fire
fire
Posts: 6239
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

Active alignment, head-marking, aktionsart, version, and other stuff

Post by eldin raigmore » 03 Jun 2018 21:29

In Frans Plank’s “Universals Archives” (https://typo.uni-konstanz.de/archive/na ... number=257), universal #257 suggests the following features of various languages are statistically correlated with each other to a significant degree.

1. Active/stative morphosyntactic alignment.
2. Grammatical “voice” is more about introversive vs extroversive version than about other details of diathesis.
3. Possession is distinguished between alienable possession and inalienable possession.
4. Verbs are inflected (or derived or otherwise “graded”) for aktionsart* rather than tense. *(action-type or action-sort)
5. Possession (at least; maybe most of the language) is head-marked rather than dependent-marked. (E.g. construct-state rather than genitive-case)

I’m especially interested in #4. Can anyone illustrate, with an example from some natlang, a verb inflected for Aktionsart?

I’m also not very confident that I understand “version” all that well. And, why would feature #2 go with the others?

In general, does anyone know of any languages having, say, just three of these features but not the other two?

WALS.info has nothing about version nor aktionsart so I can’t answer thes questions for myself without help.

Thanks.
Last edited by eldin raigmore on 21 Aug 2018 02:47, edited 1 time in total.

User avatar
Omzinesý
runic
runic
Posts: 2452
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

Re: Active alignment, head-marking, aktionsart, version, and other stuff

Post by Omzinesý » 05 Jun 2018 16:02

eldin raigmore wrote:
03 Jun 2018 21:29
In Frans Plank’s “Universals Archives” (https://typo.uni-konstanz.de/archive/na ... number=257), universal #257 suggests the following features of various languages are statistically correlated with each other to a significant degree.

1. Active/stative morphosyntactic alignment.
2. Grammatical “voice” is more about introversion vs extroversion version than about other details of diathesis.
3. Possession is distinguished between alienable possession and inalienable possession.
4. Verbs are inflected (or derived or otherwise “graded”) for aktionsart* rather than tense. *(action-type or action-sort)
5. Possession (at least; maybe most of the language) is head-marked rather than dependent-marked. (E.g. construct-state rather than genitive-case)

I’m especially interested in #4. Can anyone illustrate, with an example from some natlang, a verb inflected for Aktionsart?

I’m also not very confident that I understand “version” all that well. And, why would feature #2 go with the others?

In general, does anyone know of any languages having, say, just three of these features but not the other two?

WALS.info has nothing about version nor aktionsart so I can’t answer thes questions for myself without help.

Thanks.
I haven't read the article, so my comments are just speculation.

Aktionsart = lexical aspect
When Aktionsart is grammaticalized, it's called aspect. I guess Aktiosart is just used to mean the semantics of aspectual distinctions, here.

I think 2. (voice) is quite well explained by 5. (head-marking)
There are roughly two means of coding semantic roles of the arguments: a) in the verb (voices, inversion markers etc.) or b) in the complements (cases). Head-marking languages do not have case marking, so they have to mark argument structure in the clausal head, the verb. "Case languages" do without voices because semantic roles are marked by cases and roles of information structure are marked by word order.
I haven't checked what they mean with " introversion vs extroversion version". It can be to do with animacy. I guess the language lacking cases, raises the probability of all means of coding semantic roles in the verb, including "introversion vs extroversion version alignment", to a statistically significantly.

I should read the papers I comment, before speculating.

User avatar
eldin raigmore
fire
fire
Posts: 6239
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

Re: Active alignment, head-marking, aktionsart, version, and other stuff

Post by eldin raigmore » 05 Jun 2018 21:48

as i inderstand it it’s an error to say “when aktionsart is grammaticized it’s aspect”.
However the remark that “aktionsart...is the semantics of aspectual distinctions” exactly describes what I thought until reading this “universal”. Now I don’t know what to think.

Lexically-inherent aspectual class (aka lexical aspect, or inherent aspect, or aspectual class), is an acknowledgment that the same aspect-marking on two different verbs often don’t mean exactly the same thing.
The meaning of an aspect-mark is, instead, an interaction between the specific aspect-mark, and the verb’s aspectual class. This aspectual class is (supposed to be) inherent to the verb-(root); some other verbs might be in the same class, and some might not be.

I had formerly understood aktionsart to be a near-synonym for “lexically-inherent aspectual class”; but if a verb can be inflected for it, maybe aktionsart means something else here? I can easily see lexical suppletion to change aktionsart. I can imagine being persuaded that derivation might change aktionsart. But I don’t know how anything that’s lexically inherent can be changed by inflection, and I have my doubts about derivation.

I’m hoping I don’t have to chase down the references and their references to find out. That could take months; only some of it would be reading; probably half of it would be waiting for inter library loans to arrive.

Elsewise, as you say, I’ll have to read the books.

—————

I well understand, and agree with, what you say about grammatical voice vs grammatical case.

Many active/stative languages have two suppletive nouns for the same thing, one when it’s an agent and the other for when it’s a patient. That’s theoretically why we have both “aqua” and “hydro” from pre-proto-IE.

One head-marking way a split-intransitive language could show whether the subject of an intransitive verb were agent-like or patient-like, would be to head-mark the verb; mark it unergative voice if the subject is an agent, or mark it unaccusative voice if the subject is a patient.

A dependent-marking way, OTOH, would be to case-mark the subject ergative if it’s an agent, but accusative if it’s a patient.

The “universal” seems to be saying that in active/stative alignment languages, the first way tends not to be used.
So the difference between “intransitive agent” and “intransitive patient” must either be
* inherent to (all) the intransitive verb(s) —— every intransitive verb is either unergative-only or unaccusative-only; or,
* marked on the subject-noun by case; or,
* be inherent to the subject-noun; the given noun is either ergative-only or accusative-only.

The second and third means are not inconsistent with each other. Some nouns could be inflected between the ergative and accusative cases; some nouns could have one of their cases morphologically derived from the other; and/or some nouns could have an ergative form and an accusative form lexically supplied.

—————

If the language requires speakers to indicate somehow whether the subject of an intransitive verb is a patient or an agent; but never does so by voice-marking the verb; that leaves “voice”-marking available to show something else.

An action is “introversive” if it is directed towards the agent, or is totally contained within the agent..
It is somewhat similar to “reflexive”, and somewhat similar to “middle voice”.

If an action is directed outside the agent that is “extroversive”.

If I run in place, that’s introversive running; if I run to the post office, that’s extroversive running.

If I paint my own house, that’s introversive painting; if I paint your house, that’s extroversive painting.

If on hearing that Trump pulled US out of the Iran nuclear agreement I quietly say to myself “that idiot!”, that’s introversive speech; if instead I throw open the door to my second-story patio and go out and shout “That idiot!” to all my neighbors, that’s extroversive speech.

The EUROTYP database records version as one kind of voice.

———

And I’m still about as confused as I guess you think you are, based on available evidence. (I may have misinterpreted and you aren’t confused.)

Thanks for your response!

—————————————————————

Update: I went to my public library today and the librarian found out U of Mich in Ann Arbor is the only library in the state that has the book I want. And, because “they” are busy “improving”(?) the inter library loan system, they can’t borrow it for me until October at the earliest.

See? I told you so!
Last edited by eldin raigmore on 21 Aug 2018 02:51, edited 1 time in total.

User avatar
Omzinesý
runic
runic
Posts: 2452
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

Re: Active alignment, head-marking, aktionsart, version, and other stuff

Post by Omzinesý » 06 Jun 2018 08:38

eldin raigmore wrote:
05 Jun 2018 21:48


—————

If the language requires speakers to indicate somehow whether the subject of an intransitive verb is a patient or an agent; but never does so by voice-marking the verb; that leaves “voice”-marking available to show something else.

An action is “introversion” if it is directed towards the agent, or is totally contained within the agent..
It is somewhat similar to “reflexive”, and somewhat similar to “middle voice”.

If an action is directed outside the agent that is “extroversion”.

If I run in place, that’s introversive running; if I run to the post office, that’s extroversive running.

If I paint my own house, that’s introversive painting; if I paint your house, that’s extroversive painting.

If on hearing that Trump pulled US out of the Iran nuclear agreement I quietly say to myself “that idiot!”, that’s introversive speech; if instead I throw open the door to my second-story patio and go out and shout “That idiot!” to all my neighbors, that’s extroversive speech.

The EUROTYP database records version as one kind of voice.

———
OK, I somehow messed "intorversion" with "inversion". So I guess nothing I said about it is meaningful.
Georgian grammars name a morpheme of the verb as "the version vowel". It seems to be a kind of applicative marker.

User avatar
Omzinesý
runic
runic
Posts: 2452
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

Re: Active alignment, head-marking, aktionsart, version, and other stuff

Post by Omzinesý » 06 Jun 2018 09:02

eldin raigmore wrote:
05 Jun 2018 21:48

I had formerly understood aktionsart to be a near-synonym for “lexically-inherent aspectual class”; but if a verb can be inflected for it, maybe aktionsart means something else here? I can easily see lexical suppletion to change aktionsart. I can imagine being persuaded that derivation might change aktionsart. But I don’t know how anything that’s lexically inherent can be changed by inflection, and I have my doubts about derivation.
I think it's a question of definitions if a derived verbs a "new" one, with its own Aktionsarts or if the Aktiosart of the original stem is somehow "changed". It's however evident that a derived noun and a stem noun can, and usually do, have different aspectual interpretations.

This "inflectional" Aktionsart could also mean an infectional class that also has some common semantics. Say stative verbs are inflected differently from other verbs in a language. I think that cannot be called aspect because it doesn't have to form a paradigm. Stative verbs don't have to be inflected as non-stative verbs, like say (nearly all) Slavonic verbs have both perfectgive and imperfective inflections.

----------------------------

The burden between ABS-ERG alignment and Split/fluid alignments seems to be quite blurred. I once read an article on Basque dialects. They seem to differ very much on what clauses really are intransitive'. 'I eat meat.' is a clear transitive clause but 'I eat.' without object can be interpreted as an elliptic transitive clause where 'I' is still in the ergative or an intransitive clause where 'I' is an S in the absolutive.

User avatar
eldin raigmore
fire
fire
Posts: 6239
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

Re: Active alignment, head-marking, aktionsart, version, and other stuff

Post by eldin raigmore » 14 Jun 2018 17:56

Omzinesý wrote:
06 Jun 2018 09:02
The burden between ABS-ERG alignment and Split/fluid alignments seems to be quite blurred. I once read an article on Basque dialects.
In early work on morphosyntactic alignment, the term “ergative” seems to have been used to mean just “anything not accusative/nominative”. Split-intransitive, split-S, fluid-S, split-transitive, neutral, hierarchical, tripartite, split-ergative, “monster raving loony”, Austronesian/Philippine, active/stative, etc. —— all got called “ergative”.

———

In response to your next sentence* (which I deleted — 🙊!) , I think 🤔 it’s sometimes better to think of valency, or of intransitive vs monotransitive vs ditransitive, as a feature of the whole clause or its core, rather than of its nuclear verb.

*(the one about the elliptical transitive “I eat” clause in Basque)

Post Reply