Re: Vardelm's Scratchpad (NP: Devani OVS order)

[Vardelm]

Maybe just call the episodic pluractionality "neutral" or "default" or something like that? There doesn't have to be a fancy latinate term for everything, does there?

True. That's definitely an option.

[Vardelm]

Classes start back up (online) on Monday, so pushing through to type up as much as I can while I have time & energy!


Dwarven: Voice/Trigger

The "nominative" case indicates either a subject (when preceding the verb phrase) or a verb complement when it immediately follows the verb. As mentioned, it is defined only by a lack of a case particle and indicates no thematic relationship or any other semantic information. That's because the subject's role (the noun before the verb phrase) is marked by the verb phrase. Dwarven therefor has - depending on what you want to call it - Austronesian alignment, a trigger system, or Philippine-type voice.

As mentioned in an above post, Dwarven uses particles to indicate what "voice" or "trigger" is being used. Philippine languages use affixes, but I wanted Dwarven to be fairly analytic, so only nouns & verbs only inflect for number. Additionally, the particles used for the voice are the same as the noun case particles for the "locational" series of noun cases. There are then 7 groups of 2 noun cases each that are associated with 1 verb voice. The voices are:

Control	æq
Method	nyð
Manner	åvu
Origin	an
Position	e
Destination	lah
Affect	yñ

The Default "Middle Voice"

Verbs can go unmarked by voice. This can be viewed as a sort of middle voice. It's actually ambiguous what role the subject would be, but the assumption is that the subject is somehow affected by the verb, so it could be seen as a patient, experiencer, or theme. This depends on the verb and context to a large extent. If those don't make sense in context, then the subject might be an agent or some other role. Take the previous example:

Vardelm chæmayd.

Vardelm

Vardelm

chæm-ayd

walks-SNG.EPI

Vardelm walks.

• A voice particle doesn't appear in a preverbal slot, so the voice is technically ambiguous. However, because the verb "walk" is used, the subject is an agent. It could still be seen as "affected", like a patient, experiencer, or theme in that the subject is doing the action and is moving himself.

jastyr qis fjunayd.

jastyr

bread

qis

upward

fjun-ayd

bake-SNG.EPI

Bread is baking (up).

Vardelm qis fjunayd.

Vardelm

Vardelm

qis

upward

fjun-ayd

bake-SNG.EPI

Vardelm is baking (up).

• Here, the 1st example is relatively clear. Bread is being baked using an oven or whatever. It's a patient, which is the default.

• "Vardelm is baking" is less obvious. Is he making some kind of bread or cookies and therefor an agent performing an action? (Answer: in my house, probably not.) Or, is he a patient and is in some kind of situation comparable to bread being baked, such as lying out in the sun? Here, context is everything.

The Voice Particles

jastyr qis yñ fjunayd.

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

Bread is being baked (up) (by something/someone).

Vardelm qis æq fjunayd.

Vardelm

Vardelm

qis

upward

æq

CONTROL

fjun-ayd

bake-SNG.EPI

Vardelm is baking (up) (something).

• These examples offer more information. The bread is being baked and Vardelm is baking something. So, the bread is explicitly a patient, and Vardelm an agent.

• It's worth noting I might change my voice labels at some point, but I'm trying to be consistent with "affect", "control", etc.

• In both cases, there is an implication that there is an interaction with another referent. The bread is baked by something and Vardelm is baking something. Not every verb will include this implication, but it's there for those that it applies to.

Here's a few more basic examples showing the other voices. I'm sticking with simple intransitives for now. Transitivity is coming, maybe in the next installment.

filnjåd lah chæmayd.

filnjåd

park

lah

DESTINATION

chæm-ayd

walks-SNG.EPI

A park is walked-to.
A park is where (someone) walks-to.

ðjirok e chæmayd.

ðjirok

street

e

LOCATION

chæm-ayd

walks-SNG.EPI

A street is walked-at/along.
A street is where (someone) walks-at/along.

tunych an chæmayd.

tunych

north

an

ORIGIN

chæm-ayd

walks-SNG.EPI

North is walked-from.
North is where (someone) walks-from.

yåshot åvu chæmayd.

yåshot

stealth

åvu

MANNER

chæm-ayd

walks-SNG.EPI

Stealthy is the walking.
With stealth is how the walking is done.

tjönoch nyð chæmayd.

tjönoch

cane

nyð

INSTRUMENT

chæm-ayd

walks-SNG.EPI

A cane is walked-with.


The translations can be a bit rough. I'm trying to roughly keep the order of words from the original but have a half-way acceptable English sentence as well. Hopefully, these at least show that the voice particle is allowing the verb phrase to indicate what the role of the sentence subject is.

These are, admittedly, not the greatest or most interesting examples. I think once I go over transitivity, which might not be much more info than this post, I'll be able to make some examples that illustrate various roles moving around more.

[Vardelm]

I'm straining my tongue trying to pronounce palatalized uvulars... is there any interaction between consonant palatalization and the vowel system? Ubykh apparently had palatalized uvulars, but Ubykh also didn't have a frontness distinction in vowels. Distinguishing /qi qy qʉ qu qʲi qʲy qʲʉ qʲu/, if all of those combinations do occur, seems like a challenge unless there's some allophony that makes it easier.

Sorry, I finally noticed this reply!

Unfortunately, all I have at this time is "haven't considered it much". Still, I think it's a good point. I looked at the Ubykh phonology page on Wikipedia and found this: "Palatalisation of the uvular consonants is no longer phonemic, also being replaced in many instances by gemination (standard /qʲa/ ('to cough') → Karacalar /qːa/)". That might be something I steal since gemination is phonemic in this 'lang, although I haven't used it much in the bit of vocab I've made up. The other thing to consider is that this is intended as a proto-language, so depending on what I want to do with the descendant langs may decide whether this is an issue at all.

[Vardelm]

Dwarven: Transitivity

As mentioned in the last post, adding a voice particle can add an implication of transitivity. An explicit patient can suggest an agent at work.

jastyr qis fjunayd.

jastyr

bread

qis

upward

fjun-ayd

bake-SNG.EPI

Bread is baking (up).

jastyr qis yñ fjunayd.

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

Bread is being baked (up) (by something/someone).


When the subject is a patient (the "affect" voice is used with the verb), agents can be added just like other adjuncts after the verb. Prototypical agents will appear in the ergative or genitive case. Natural causes/forces might be in the comitative/instrumental cases.

jastyr qis yñ fjunayd æq Vardelm.

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

æq

by.ERG

Vardelm

Vardelm

Bread is being baked (up) by Vardelm.


When the subject is an agent (the "control" voice is used), patients are in the absolutive or partitive case.

Vardelm qis æq fjunayd yñ jastyr.

Vardelm

Vardelm

qis

upward

æq

CONTROL

fjun-ayd

bake-SNG.EPI

yñ

ABS

jastyr

bread

Vardelm is baking (up) bread.


Verbs of Perception

Verbs of perception, emotion, or other mental activity treat the experiencer as the "patient". The thing "causing" the stimulus takes the comitative/instrumental cases.

Vardelm yñ gjovayd nyð jastyr.

Vardelm

Vardelm

yñ

AFFECT

gjov-ayd

love-SNG.EPI

nyð

with.INST

jastyr

bread

Vardelm loves bread.


And the "passive" version of that, using the "method" voice:

jastyr nyð gjovayd yñ Vardelm.

jastyr

bread

nyð

with.METHOD

gjov-ayd

love-SNG.EPI

yñ

ABS

Vardelm

Vardelm

Bread is loved by Vardelm.

[evmdbm]

Could I say "Vardelm filnjåd lah chæmayd"?

to mean Vardelm walks to the park, or indeed just add Vardelm to the start of all these sentences - or do I need an extra particle (æq I assume)? As it stands these sentences are not merely intransitive they're also impersonal

[Vardelm]

Could I say "Vardelm filnjåd lah chæmayd"?

Nope. One of the upcoming posts is that this is also a V2 language. Your example would would have 2 nouns - "Vardelm" and "filnjåd" - ahead of the verb "chæmayd". One of them will have to go after the verb, and that noun will require a case particle as a result.

As it stands these sentences are not merely intransitive they're also impersonal

I assume you're talking about the sentences like the one (copied below) with the location voice where an actor isn't specified? If so, yes, you are correct.

ðjirok e chæmayd.

ðjirok

street

e

LOCATION

chæm-ayd

walks-SNG.EPI

A street is walked-at/along.
A street is where (someone) walks-at/along.

[Vardelm]

Dwarven: Telicity

As might be guessed, Dwarven uses the partitive case particle in much the same manner as Finnish. The partitive is used for atelic events. I still have to decide about some of the other uses, such as negatives.

Where Dwarven differs from Finnish is that the partitive is not the only case used for atelic events. The entire series of "path" cases are used as well. That's the partitive, lative, prolative, ablative, essive, commitative, and genitive.

This is the first I'm writing this down, aside from working out a few example sentences. Hopefully it makes sense. I've learned to understand telicity a bit over the past couple years, although I know I don't fully grasp it. One difficulty here is going to be that Dwarvish doesn't have perfective vs. imperfective aspects, and so I think the language might be learning more towards using case to indicate perfectivity rather than telicity. Or it could be me not understanding, or just not translating well. I'll give it a go & see what happens.


Vardelm yñ uljegayd nyð vollyr. (TELIC)

Vardelm

Vardelm

tjar

via

yñ

AFFECT

uljeg-ayd

read-SNG.EPI

nyð

with.INST

vollyr

book

"Vardelm is reading (via/with) a book (to finish it)."
"Vardelm read (via/with) a book (and finished it)."


Vardelm yñ uljegayd acho vollyr. (ATELIC)

Vardelm

Vardelm

tjar

via

yñ

AFFECT

uljeg-ayd

read-SNG.EPI

acho

with.COM

vollyr

book

"Vardelm is reading (via/with) (from) (some of) a book."
"Vardelm read (via/with) (from) (some of) a book."

• Unlike Finnish, where the object/patient case determines telicity, here it's always the 1st adjunct after the verb phrase.
• In Dwarven, uljeg "to read" is a mental/perception verb, so the experiencer is the patient & the verb gets yñ, the "affect" voice particle which indicates a patient subject.
• Likewise, given the verb, the book is the instrument that "causes" the perception, so it uses either the commitative or instrumental case.
• The translations show both a perfective and imperfective version since Dwarven doesn't make the distinction.

Like the last post on transitivity, we can take the "passive" version of the above sentences. (I say "passive" just because that's what English would use for this word order.)

vollyr tjar nyð uljegayd yñ Vardelm. (TELIC)

vollyr

book

tjar

via

nyð

with.INST

uljeg-ayd

read-SNG.EPI

yñ

ABS

Vardelm

Vardelm

"The book is being read with by Vardelm (to finish it)."
"The book was read with by Vardelm (and finished)."


vollyr tjar nyð uljegayd ja Vardelm. (ATELIC)

vollyr

book

tjar

via

nyð

with.INST

uljeg-ayd

read-SNG.EPI

ja

PART

Vardelm

Vardelm

"The book is being read with by Vardelm (some of it)."
"The book was read with by Vardelm (some of it)."

• Here, what English (and I think Finnish) would consider to be the agent is treated as the patient, and so get either the absolutive case yñ for the telic sentence or the partitive case ja for the atelic.

Let's take verb that shows a more familiar pattern.

jastyr qis yñ fjunayd æq Vardelm. (TELIC)

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

æq

by.ERG

Vardelm

Vardelm

"Bread is being baked up by Vardelm (a complete loaf/batch)."


jastyr qis yñ fjunayd nyð pjokkün. (TELIC)

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

nyð

with.INST

pjokkün

oven

"Bread is being baked up by an oven (a complete loaf/batch)."


jastyr qis yñ fjunayd acho pjokkün. (ATELIC)

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

hu

of.GEN

Vardelm

Vardelm

"(Some) Bread is being baked up by Vardelm."


jastyr qis yñ fjunayd acho pjokkün. (ATELIC)

jastyr

bread

qis

upward

yñ

AFFECT

fjun-ayd

bake-SNG.EPI

acho

with.COM

pjokkün

oven

"(Some) Bread is being baked up by an oven."

VS.

Vardelm qis æq fjunayd yñ jastyr. (TELIC)

Vardelm

Vardelm

qis

upward

æq

CONTROL

fjun-ayd

bake-SNG.EPI

yñ

ABS

jastyr

bread

"Vardelm is baking up bread (a complete loaf/batch)."


Vardelm qis æq fjunayd ja jastyr. (ATELIC)

Vardelm

Vardelm

qis

upward

æq

CONTROL

fjun-ayd

bake-SNG.EPI

ja

PART

jastyr

bread

"Vardelm is baking up (some) bread."

• I think the translations are a bit rough, but hopefully it gets the point across.
• I'm aware that definiteness plays some role in determining the telicity of a statement. I haven't included definite articles here. I need to try & clarify how that will play a role exactly. I assume the case particle might partly be determined by the definiteness of the patient, be that in subject or adjunct position.

[aporaporimos]

As it stands these sentences are not merely intransitive they're also impersonal

I assume you're talking about the sentences like the one (copied below) with the location voice where an actor isn't specified? If so, yes, you are correct.

ðjirok e chæmayd.

ðjirok

street

e

LOCATION

chæm-ayd

walks-SNG.EPI

A street is walked-at/along.
A street is where (someone) walks-at/along.

Are they necessarily impersonal? As in, does the agent have to be vague/non-definite, or could an agent be inferred from context? Or to put it another way, could that sentence mean "they were walking along a street" (given an appropriate context)?

By the way, I think you can get away with being freer in your translations. The gloss is there to show us the syntax of the original, so the translation doesn't have to convey all of that information too.

[Vardelm]

Are they necessarily impersonal? As in, does the agent have to be vague/non-definite, or could an agent be inferred from context? Or to put it another way, could that sentence mean "they were walking along a street" (given an appropriate context)?

I would say that yes, context could make the agent apparent, but of course it would also be used in situations where an agent is unknown or only inferred.

This is interesting because I haven't thought much about it yet, but it fits into the language concept. Using an Austronesian alignment was one of the 1st features I wanted to include for it's ability to have locative, benefactive, etc. subjects. I picture my Dwarves as focusing much more on patients and the circumstances of an event than agents. Their interest is more in objects, locations, and methods than in people. So, use of an impersonal might be fairly common, such as an answer to the question "Which way did they walk?", or even "Which way was the walking?" Answer: "That street was walked along."

By the way, I think you can get away with being freer in your translations. The gloss is there to show us the syntax of the original, so the translation doesn't have to convey all of that information too.

Thanks! Good to know.

[Vardelm]

Dwarven: V2 order & focus

For convenience, here's the word order again, but I'm describing it slightly differently.

Subject Noun Phrase:

(Case particle)₁ + Determiner + Noun + Adjective

1. The case particle is included only when the subject noun phrase follows the verb due to an adjunct noun phrase being fronted for focus.

Adjunct Noun Phrase:

Preposition + Case particle + Determiner + Noun + Adjective

Verb Phrase

Aspect Particle + Directional Particles + Voice/Focus Particle + Verb

Sentence Order:

Subject Noun Phrase + Verb Phrase + Adjunct Noun Phrase 1 + Adjunct Noun Phrase 2 + ...

V2 Word Order

Dwarven looks like it has SV & SVO word order, or perhaps OVS depending on how you look at the standard sentences I've provided so far. However, it actually has V2 word order, at least as far as I understand V2 so far. I haven't worked out dependent clauses yet, but I plan to make those V2 order as well. I just don't know how that will work out, exactly. Right now, phrase fronting is used for emphasis/focus like is seen in Hungarian. I'm open to other suggestions.

An adjunct noun phrase can be fronted in order to place it in focus, which is to emphasize it. I think this should be very similar to Hungarian. Because Dwarven is V2, the subject noun phrase moves to immediately follow the verb. It also needs to use a case particle, unlike when it comes before the verb.

Focused Sentence Order:

Adjunct (Focused) Noun Phrase + Verb Phrase + Subject Noun Phrase + Adjunct Noun Phrase 1 + Adjunct Noun Phrase 2 + ...

Vardelm tjar yñ uljegayd acho vollyr e ðjodaim.

Vardelm

Vardelm

tjar

via

yñ

AFFECT

uljeg-ayd

read-SNG.EPI

acho

with.COM

vollyr

book

e

at.LOC

ðjodaim

morning

"Vardelm was reading a book in the morning."


e ðjodaim tjar yñ uljegayd yñ Vardelm acho vollyr .

e

at.LOC

ðjodaim

morning

tjar

via

yñ

AFFECT

uljeg-ayd

read-SNG.EPI

yñ

AFFECT

Vardelm

Vardelm

acho

with.COM

vollyr

book

"It was in the morning that Vardelm was reading a book."


This works for the impersonal statements as well.

e ðjirok e chæmayd.

e

at.LOC

ðjirok

street

e

LOCATION

chæm-ayd

walks-SNG.EPI

It's a street that is walked-at/along.


I might allow the 2nd case particle to be dropped when clarity isn't sacrificed.

e ðjodaim tjar yñ uljegayd Vardelm acho vollyr .

e

at.LOC

ðjodaim

morning

tjar

via

yñ

AFFECT

uljeg-ayd

read-SNG.EPI

Vardelm

Vardelm

acho

with.COM

vollyr

book

"It was in the morning that Vardelm was reading a book."

e ðjirok chæmayd.

e

at.LOC

ðjirok

street

chæm-ayd

walks-SNG.EPI

It's a street that is walked-at/along.


Subject nouns for intransitive verbs take the absolutive case particle by default.

e ðjodaim ðayd qjuralg yñ Vardelm

e

at.LOC

ðjodaim

morning

ðayd

is.SNG

qjuralg

warrior

yñ

ABS

Vardelm

Vardelm

It's in the morning that Vardelm is a warrior. (He's a bastard before he gets his coffee.)

[aporaporimos]

Can you expand on the difference in meaning between fronting an adjunct phrase, and making that phrase be the subject (and the original subject an adjunct)? This might just be me not being very familiar with topic-focus terminology, but I don't really grasp the distinction between

1. e ðjodaim tjar yñ uljegayd yñ Vardelm acho vollyr. ("In the morning" as fronted adjunct)
2. ðjodaim tjar e uljegayd yñ Vardelm acho vollyr. ("In the morning" as subject)

The second sentence is my own construction, hopefully I formed it right.

[Vardelm]

s[/url]t_id=29199 time=1589910525 user_id=361]

1. e ðjodaim tjar yñ uljegayd yñ Vardelm acho vollyr. ("In the morning" as fronted adjunct)
2. ðjodaim tjar e uljegayd yñ Vardelm acho vollyr. ("In the morning" as subject)

The second sentence is my own construction, hopefully I formed it right.

I believe you did, yes! I'm still learning this language as well, but it looks corrrect to me.

Can you expand on the difference in meaning between fronting an adjunct phrase, and making that phrase be the subject (and the original subject an adjunct)? This might just be me not being very familiar with topic-focus terminology, but I don't really grasp the distinction between...

Sure. I think it's a matter emphasis. Here's a link to the relevant Wikipedia article/section:

Hungarian grammar: Emphasis

I think that article states it fairly well when it says:

The focus attracts the attention to an element of the event that is considered to be unknown, or it may be a refutation to a possible opposing belief. It excludes the validity of the statement for all other individuals in question and is equivalent to "it was X and nothing else that...".

I would say that it could also be a re-affirmation that the element is true. So, sentence #2, where "morning" is the subject, is a neutral phrase. The event happened in the morning. Maybe it also happened in the afternoon, evening, etc. Sentence #1 says the event definitely happened in the morning. That could mean that it could be in opposition to a belief that it did not happen in the morning, or that it happened some other time.

I don't know if Austronesian languages have a similar feature, but I don't think so. For them, simply making a location (or whatever role) the subject puts focus on that element and emphasizes it. They may be able to add stress to the word to add more emphasis, like we do in English.

Does that help?

[Ares Land]

HELP!?!?! : I'm using "episodic" here because, AIUI, episodic contrasts with gnomic. However, I am also using the term "episodic" for one of my verb pluractionality numbers: momentane, episodic, iterative, and habitual. There, "episodic" is basically neutral w/ regards to aktionsart. Momentane is just just 1 instance of an activity (jumping), while iterative would be several episodes (jumping for a short bit, stopping for a while, then jumping again, etc.). Any suggestions for alternative labels to avoid confusion? Delimitative, perhaps?

I'd use "episodic" for the opposite of gnomic and change the label for your other 'episodic'. Can you expand on your episodic aspect? Maybe we can figure out another label for it.

(Very nice work, BTW!)

[Vardelm]

I'd use "episodic" for the opposite of gnomic and change the label for your other 'episodic'. Can you expand on your episodic aspect? Maybe we can figure out another label for it.

Yep. I need to go back & add some more detail on a bunch of things. Noun and verb number are both on that list. I need to think through the "episodic" issue more to define it. It's quite possible that over-thought it and don't need a distinction between that & momentane.

(Very nice work, BTW!)

Thank you much!

[aporaporimos]

...
Does that help?

Yeah, it does. I was just uncertain since the resulting constituent order is the same in either case.

Yet another question: why does the case particle have to be provided with a post-verbal subject? Isn't it always identical to the voice marker on the verb?

[Vardelm]

Yet another question: why does the case particle have to be provided with a post-verbal subject? Isn't it always identical to the voice marker on the verb?

You are correct that it will be identical. It would be feasible to not use it. The post-verbal noun phrases would be interpreted such that the one that didn't use a case particle would be the one that is the subject. However, there is at least 1 situation - equational sentences - where there is already a noun in the nominative "case" (in that it doesn't use a case particle), so adding the case particle to the subject becomes something that would help keep things organized. True, even then, it's maybe not absolutely required. As I make more & more examples, maybe I'll get to the point that I don't feel like adding it is ever needed. To some extent, though, I was taking a cue from Tagalog which always used its case particles, for both subject & object. The particles there are different from the markers on the verbs, though, which are affixes.

[aporaporimos]

Yet another question: why does the case particle have to be provided with a post-verbal subject? Isn't it always identical to the voice marker on the verb?

You are correct that it will be identical. It would be feasible to not use it. The post-verbal noun phrases would be interpreted such that the one that didn't use a case particle would be the one that is the subject. However, there is at least 1 situation - equational sentences - where there is already a noun in the nominative "case" (in that it doesn't use a case particle), so adding the case particle to the subject becomes something that would help keep things organized. True, even then, it's maybe not absolutely required. As I make more & more examples, maybe I'll get to the point that I don't feel like adding it is ever needed. To some extent, though, I was taking a cue from Tagalog which always used its case particles, for both subject & object. The particles there are different from the markers on the verbs, though, which are affixes.

Fair enough, and of course there's nothing wrong with a little redundancy. Come to think of it, you equally well go the other direction and relax the requirement that the subject immediately follow the verb, since the case marker identifies it as such.

[Vardelm]

Fair enough, and of course there's nothing wrong with a little redundancy. Come to think of it, you equally well go the other direction and relax the requirement that the subject immediately follow the verb, since the case marker identifies it as such.

That's actually a good point. I had in mind to define the typical order for the other cases: time-manner-place type issues. Thinking about a freer order for subjects, agents, patients, etc. would be worthwhile.

[Vardelm]

Maybe just call the episodic pluractionality "neutral" or "default" or something like that? There doesn't have to be a fancy latinate term for everything, does there?

I'd use "episodic" for the opposite of gnomic and change the label for your other 'episodic'. Can you expand on your episodic aspect? Maybe we can figure out another label for it.

OK, to respond to the replies about verbal number....


Dwarven: Verbal Number

Repeating the verb number table for reference:

	Singular	Plural	Distributive
Momentane	-ys	-essa	-ysach
Durative	-ayd	-ið	-ach
Iterative	-ry	-deyr	-rych
Habitual	-ym	-maan	-mach

I reviewed lexical aspect/aktionsart a bit. Since Dwarven will use noun cases to indicate telicity, that lets me group verbs into 3 categories that in a useful manner.

• States
• Activities & accomplishments
• Semelfactives & achievements

From there, I made a chart to figure out how each categories might interact with verbal number. Here I'm only concerned with the momentane, durative, iterative, and habitual numbers, since singular, plural, & distributive are basically just agreement w/ the subject.

	States	Semelfactives, Achievements	Activities, Accomplishments
momentane	briefly in a state	emphasis on one time only	do the activity briefly & suddenly one time
durative	in a state	do the action, possibly several times in immediate succession	do the activity
iterative	in a state repeatedly (the state comes & goes)	do the action repeatedly with some time interval in between	do the activity repeatedly with some time interval in between
habitual	in a state repeatedly over a long period of time	do the action repeatedly over a long period of time	do the activity repeatedly over a long period of time

To state it differently:

• Momentane indicates states/events of a very short amount of time and only 1 "burst" of the state/event occurs. For semelfactives & achievments, such as "to knock", there would be only 1 knock, as in 1 single tap.

• Durative is neutral regarding the length of time the verb takes, but there is still only 1 event. For states, activities, & accomplishments, this just says the event is happening, for whatever period of time. With "knock", there may be 1 knock (a single tap) or there could be several knocks in quick succession with no break in between. The totality of that succession is seen as 1 event of some duration. You could say, for example, "he is knocking", indicating (probably) a repetition of knocks. "He knocks" could be 1 or more individual taps.

• Iterative says that there are several episodes that occur and that there is probably some amount of time that occurs between each episode. This is particularly true for 1 episode of the instants. The subject may knock 1-5 times in rapid succession, wait for several seconds, and then repeat. It may be that the change between each iteration is not time but a change in objects, such as knocking on several different doors.

• Habitual is perhaps the most straightforward in that it's an event that the subject regularly is engaged in over some long period of time, which is relative to each event. If I go for a run several times during an afternoon, that's iterative. If I do it every day for a week and plan on continuing, that's habitual.

So, thoughts on what labels to use? General comments welcome, too!


EDIT: I updated the chart & descriptions since I decided last night to go with "durative" instead of "episodic" for the aspect.

[aporaporimos]

I think this set of aspect terms is fine... as long as you don't also use "episodic" to contrast with "gnomic" for verbs marked / not marked with a direction particle, since that seems to be an orthogonal distinction.