zompist bboard

Originally posted in C&C Quickies in relation to my conlang. I figured this might be useful to people so I'm putting in here. Please add, comment, question.



What is aktionsart? It is built-in a semantic and lexical quality, inherent in every verb. It explains a fundamental difference between a word such as "think" and "shoot". It is different from grammatical aspect which can be changed as needed.

First, we can categorize verbs into stative and dynamic verbs. Stative verbs describe a state, dynamic verbs describe a change. We can further subdivide dynamic verbs into four subtypes by their aktionsart. Together, we have five types of actionsart:

Stative:
state e.g. (atelic, durative) "know" or "be drunk"

Dynamic:
semelfactive (atelic, punctual), e.g. "shoot" or "knock"
achievement (telic, punctual*), e.g. "realize" or "die"
accomplishment (telic, durative), e.g. "teach" or "break"
activity (atelic, durative), e.g. "research" or "run"

*Certain achievement verbs can be durative in a sense, it can take time for someone to die, for example. But the process of dying isn't the same thing as actually dying.

Telicity refers to whether or not the verb has a defined or natural endpoint. If it does, it's telic, if it doesn't it is atelic. Durativity refers to whether the verb's action takes time. Punctuality is the opposite; a punctual verbs' result is more or less immediate.



The following is true of most European languages and many more:

States generally cannot be in a progressive aspect, e.g. *I am knowing it, *I am being drunk. They also cannot be described with modal adverbials: *I know it fiercly; *I think it with great care; *I am tall throughly. They generally cannot appear in imperatives either: *Know about aktionsart! *Be of that opinion! *Be tall! Since they are atelic, they cannot be described with period-descriptors: *I knew Latin in three months; *It took me three hours to be drunk (note however: "It took me three hours to become drunk").
They sort of work with these if more qualification is added and a non-present tense is used: I knew Latin for five years but then I forgot everything (the present is impossible however) I will be drunk for an hour but then I'll sober up

Semelfactives are punctual, atelic and nondurative. They do not work well progressive: *?he was knocking it on the floor. They do not work with most period-descriptors: *Knock it for an year. They do work as imperatives however: Knock it! and with modal adverbials He knocked it fiercly.

Achievements refer to changes in state, and are thus telic. Their subjects thematic role is usually that of theme or patient. However, they are not really durative and therefore cannot logically be made progressive. In English, when these verbs are made progressive with "be ...ing" they actually become habitual: She wasn't getting it, she was getting awards, she was realizing new things; or prospective: She is getting it (as in: She's about to get it), the sheets are drying (or: are about to get dry).
They do not work well with most modal adverbs: *The sheets dried fiercly, ?The sheets dried too much.. They rarely work when imperative, and then only as wishful thinking: ?Dry!, Break!, die!, *Realize!. They sometimes work with period-descriptors: She understood it in ten minutes; but often not: *She died for ten hours.

Accomplishments are durative and work with the progressive: I was teaching syntax. Some work badly with certain modal adverbials: ?I broke the glass fiercly. Since their subjects are willing participants in the action (θ-role: agent), these verbs can be imperative. Their objects often influence how amenable they are to period-descriptors: ?He taught syntax in ten months but I taught the entire book in ten months.

Activities are atelic because there is no obvious endpoint to something such as eat, browse the web, or rain. They are durative because they take time however. Most weather verbs are activities but they have no agents and are thus a bit different from others. Mostly, activities can be progressive: He was browsing the web and It was raining, though other weather verbs cannot: *It was being sunny. When activities have subjects, they are agents or agent-like and therefore the imperative works fine with the ones with subjects: Browse the web!, but *Rain!. Period-descriptors beginning with in do not work: *He browsed the web in ten minutes or *It rained in ten minutes.

_________________
vec

I thought that using "for" or "in" was a test for telicity. ( http://en.wikipedia.org/wiki/Telicity ) I don't understand your reasoning for classification. Going by my test for telicity, I can say the following:

"run" can be telic.
- I ran a mile in 7 minutes.
- *I ran a mile for 7 minutes.
- I walked to the store in 7 minutes.
- *I walked to the store for 7 minutes.
It can be atelic too: (Note how these sentences mean slightly different things.)
- I ran around the block in 7 minutes. (You went around the block exactly once.)
- I ran around the block for 7 minutes. (You went around the block an unspecified number of times.)


- I taught the entire "Intro to Java" book in 4 months.
- I taught "Intro to Java" for 5 years.
"teach" is atelic in the former, and telic in the latter.

"break" too, can be either:
- I broke my glasses twice in a week.
- I broke windows (as a job, at a window factory, to test them) for 5 years.

I admit that "research" doesn't sound good in a telic sense:
- *I researched 5 different kinds of cancer in 5 years.
- I researched 5 different kinds of cancer for 5 years.
I suppose that one would express the telic meaning with a different verb, "look up":
- I looked up 5 different kinds of cancer in 1 minute.
- *I looked up 5 different kinds of cancer for 1 minute.


They work just fine. The reason that "dry", "break", and "realize" sound weird is because they need an object. "realize" also sounds weird because it's usually static.
- Dry yourself off!
- Break it!
- Die!
- Realize it! (Telic. Could be rephrased as "Come to realize it!" if you want to respect "realize" being static.)


Of course they work with "in", but not with "for"; They're telic.


Because they're atelic.


"eat" can be either:
- I ate spaghetti (every day) for 5 years. (atelic)
- I ate a plate of spaghetti in 5 minutes. (telic)


I agree with this. Even then, people do say things like:
- How are you liking the class?
- I'm hating it.

However, it doesn't mean anything different from the non-progressive forms though:
- How do you like the class?
- I hate it.


What do you mean by "modal adverbials"?

This leaves punctuality/durativity. I think that a better way of understanding it is to draw an analogy from nouns: Count nouns are to punctual verbs as mass nouns are to durative verbs. Durative verbs are hard to break up into separate instances. The best test that I can think up to test for punctuality/durativity is to ask whether "is x-ing (the y)" is broadly equivalent to to "starting to x (the y)". If it's not, then ask whether it's iterative. If it's not, than it's durative.
Examples:
Group 1) die, (come to) realize, (come to) understand, break
Group 2) shoot, hit, knock, bite
Group 3) run, eat, teach, research, read, yell

I'd still classify the second group as punctual though. You can imagine watching a video that shows a gun shooting in slow motion, and say "It's shooting!" as it begins. Maybe it's something to do with the nature of the verbs in the first group that they change the state of the patient so drastically that it can't happen twice. (You can die only once, usually; A vase can break only once, etc.)

*****

So, to sum up my view of it:

Verbs
- Static : The progressive does *not* change meaning. (Aka: Generally cannot be in the progressive.)
- Dynamic : The progressive does change meaning.
-- Telic : Can take a temporal prep phrase that starts with "in".
-- Atelic : Can take a temporal prep phrase that starts with "for".
-- Punctual : Doing it repeatedly (if possible) is perceived as multiple actions. (iterative)
-- Durative : Doing it repeatedly is perceived as a single action.

*****

By the way, "knock" is a weird verb that has a strange restricted usage to:
1) I knocked on his door 5 times.
2) I knocked the cup off the table.
3) Knock it off! (What you say to somebody that's teasing you. However, it sounds wrong/odd to say ?"The bully finally knocked (it) off.". "stop" is properly orthogonal: "Stop it!" "The bully finally stopped.")
Saying "knock" without a preposition sounds wrong:
*I knocked his door.
*I knocked the cup.

That's because telicity is not a property of verbs, it's a property of predicates. "run" is atelic, but "run a mile" is telic. The difference between the two "run around the block"s is in fact telicity: for the former, the block is like a race, and has a goal; for the second, there is no goal, so there is no specified amount of times to run it.


Again, the difference is in the object. It is widely true cross-linguistically that definite/specific/quantified objects can make a predicate telic, while indefinite/unspecific/unquantified objects make it atelic. There are different theories to account for this, but all theories of aktionsart acknowledge that properties of the object can affect telicity.


You give good tests for the first two distinctions (static/dynamic; telic/atelic). I'm not sure what tests there are to distinguish punctuality from durativity.

_________________

Ah, yes, it makes more sense that way. Static/Dynamic and Punctual/Durative seem to be a property of verbs (and not predicates) though.


Yes, that's exactly what I was pointing out with those two examples. I didn't mean to imply that they were both atelic.

There is a decent one that Terra pointed at but didn't quite hit, only it's restricted to what vec listed as atelic verbs: the progressive easily (I want to say necessarily, but can't prove it) gives a repetitive-action reading with punctual verbs that is unavailable with durative verbs. Despite vec's claim, "I was knocking it on the floor" is perfectly good English, it just means that you are repeatedly knocking something on the floor. With "knock" this is only semantically compatible with doing so deliberately, but volition aside, the same repetition rule goes for numerous other punctual verbs: "I was sneezing", "he was blinking in the bright light", "that light is flashing", "we were shooting fish in a barrel to test out the old saying", and so forth.

--

I'm not confident that these semelfactive verbs actually possess any default telicity. You can't test for it because punctual-ness by its very nature permits neither a "for"- nor an "in"-phrase for time and I don't know any other tests that would apply. (The progressive-repetitive versions are clearly atelic but I think that's specified grammatically, not lexically.) Intrinsically the very definition of punctual-ness defies asking whether there's a natural endpoint - things that happen instantaneously do not begin or end, they happen all in that one instant.

Nevertheless they're clearly different than the achievement verbs, which also do not easily permit a repetitive reading in the progressive. The achievement verbs are clearly telic by default (i.e. predicates with nothing but the verb are telic), but I feel the punctual-durative dimension does not really apply to them very well. Vec was wrong that they just become habitual in the progressive: they can be habitual, but in all of the ones at hand, the progressive can also just refer to the time spent achieving the endpoint... which is clearly durative. And yet, the endpoint itself is often only what you mean when you use the verb, and endpoints are punctual by virtue of being points. The verb freely referring to either the time spent on achieving or the realization of the achievement, or both together as a whole, I do not see value in saying they have a punctual or a durative aktionsart.


Vec formed his four "dynamic" classes out of the intersection of [+/- punctual] and [+/- telic]. I agree the four are valid classes with regard to those features, but I do not agree that those are always in fact features. I would re-arrange his table thusly:

semelfactive (punctual, telicity undefined)
achievement (telic, durativity undefined)
accomplishment (telic + durative)
activity (atelic + durative)


Obviously this leaves open the possibility of more categories, if we start over from from telicity and durativity and allow each feature a third "undefined" category. Can anyone think of English verbs that fill other slots on it?

One of the difficulties with defining the Aktionsarten of predicates in the two-component theory of aspect Vec (and many linguists) are working in is that you can't get to them untouched: Aktionsart interacts with the category of viewpoint aspect in interesting ways. For examples, in many languages, stative predicates are incompatible with perfective aspect, while in other languages, adding perfective aspect to stative predicates makes them into achievements. Semelfactive predicates with progressive aspect (in languages where this is a licit combination) become iteratives, which can seem similar to durativity.


In this case, certain viewpoint aspects (e.g., the progressive) take an inherently punctual achievement predicate and turn it into a durative. Strictly speaking, under Vec's model, what an achievement predicate refers to is a change of state, which is punctual. However, as you point out, is it reasonable to infer that in the real world, achieving this change of state takes time, and with imperfective viewpoint aspect, which cannot take punctual predicates as-is, the viewpoint is shifted to this preparatory action.

_________________

I'm just seeing this now. Glad this sparked some discussion!

I myself question if the five categories are all. I find the names a bit confusing and I can never remember which is which with achievements and accomplishments.

Perhaps getting rid of those five categories is better and to simply use [±telic] [±durative] [±dynamic] since it seems there can be a great deal of context-based variance.

_________________
vec

Indeed. It's like how "aorist" means past-perfective.

It doesn't always. In fact outside of certain uses in Ancient Greek (and it's not even absolute there), I'd venture to say that's not the primary definition. Then again, saying it has a primary definition is probably delusional...IOW unless it's clearly defined in a paper I'd say it's a term best left unused.

_________________

Common Zein Scratchpad & other Stuffs! OMG AN ACTUAL CONPOST WTFBBQ

Formerly known as Drydic.

I know. The variable definition is part of the problem.

Btw, I think reading the verb section of this conlang would be helpful to somebody trying to make sense of verbs: http://www.rickmor.x10.mx/lexical_semantics.html#S2_0