Struggling to grasp cases and voice

If you're new to these arts, this is the place to ask "stupid" questions and get directions!
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Struggling to grasp cases and voice

Post by Doranwen »

Keenir wrote:no, you wouldn't. I can't say it's looking odd, but then, I need a bo"le of root beer. :)
{would not \ can not \ it is \ bottle}
Hee, I see what you did there. ;)
one thing to remember: with your conlang, you are the deity: you have the final word (and the right to have a final word later on as well)
Yeah, I know. I'd like it not to look completely outlandish, though, and have some logic that made some sense to other people. :) And sometimes I'm just curious if any other language has ever done something, because it's nice to know if I'm following in someone else's footsteps or whether it's totally new (which will help me be sure to explain that clearly, if no one will have ever encountered it before). And partly since I was totally not satisfied with my last result--partially because I didn't understand things, and partially because I did go overboard with a few features, and the words of wisdom here will help me create a much better result. :) Believe me, if I'm really not happy with someone's advice, I will just chuck it and do what I want, lol, but at this point everything has been very helpful and helped me feel like I understand my own conlang better.
User avatar
kiwikami
roman
roman
Posts: 1203
Joined: 26 May 2012 17:24
Location: Oh, I don't know, I'm probably around here somewhere.

Re: Struggling to grasp cases and voice

Post by kiwikami »

Doranwen wrote:Ooh, I like that! (Thanks for the examples, I would have stared at the words "nominal constructions" and not really been able to picture what you meant.) And could you give me an example of the verb markers you were referring to?
No problem! Marking direction/location on a verb might mean either morphemes added to a verb that show a location or direction of motion, or completely stems for different kinds of motion. The former might look something like the following:

bob-SUB hill-OBJ go-ABOVE
Bob goes up the hill.
Marking telicity (whether or not an action is completed) on the verb could distinguish then between that and Bob goes to the top of the hill.

The latter would be something like:
bob-SUB hill-OBJ go_up
Bob goes up the hill.

Where go_up is its own verb stem, separate from others like go_down, go_around, etc. (consider English "ascend" and "descend"). The first option, with separate morphemes, is perhaps more versatile here, because you could theoretically use it on any verb, possibly using the verb to make a locative-marked noun more specific. That is, something like this:

bob-SUB hill-LOC dance-ABOVE
Bob dances on top of the hill.

I'm a wee bit biased when it comes to marking direction and motion on verbs, because it's a feature I like to incorporate in my own conlangs, but it can certainly help make prepositions less of a thing without adding new cases for every possible method of getting from point A to point B. This could likely extend to non-location-related cases as well, but I'm not super certain how, or how common/likely this is. BUT it can also essentially lead to shifting a still-rediculously-complicated case system over to the verb from the noun, rather than actually reducing case complexity, so maybe take everything I say with a grain of salt... :roll:
Edit: Substituted a string instrument for a French interjection.

:eng: :mrgreen: | :fra: [:)] | ASL [:S] | :deu: [:|] | :tan: [:(] | :nav: [:'(]
Trailsend
moderator
moderator
Posts: 1451
Joined: 18 Aug 2010 05:22

Re: Struggling to grasp cases and voice

Post by Trailsend »

Doranwen wrote:Ahh, so the passive/antipassive is markers on the verb itself (I so did not understand that part!) that allows the verbs to change the transitivity (valency?) from transitive to intransitive while still sticking with the rule that a verb has to have a subject.
Strictly speaking, only the bold part is true; you could have a passive and antipassive voice in a language that didn't require verbs to have subjects. A required subject is just one way in which they might be useful.

You could certainly imagine a tripartite language that does allow omitting arguments from transitive verbs. Maybe the following two statements are both grammatical:

1. na ruff-mi ak-in ketek
my dog-ERG you-ACC bite
my dog will bite you

2. na ruff-mi ketek
my dog-ERG bite
my dog bites (somebody)

So if you also have an antipassive voice construction:

3. na ruff-ta i-ketek
my dog-ABS ANTIPASSIVE-bite
my dog bites

Now you get to decide what that means. Why would a speaker choose to say (3) instead of (2)? What additional flavor, connotation, nuance does the antipassive voice convey in this language, which just leaving out the accusative argument would not?

(For example, maybe omitting the accusative argument like in (2) can be used to describe specific events, while the antipassive is used to describe general truths or ongoing states. Thus, (2) would mean My dog bit somebody, but I'm not mentioning who, while (3) would mean My dog, as a general rule, bites.)
Doranwen wrote:And then of course there are verbs like "eat" which can be either transitive or intransitive, right? Well, in English they do. "She ate" vs. "She ate the apple". Are there languages that require the "ate" from "She ate" to be marked antipassive, then? I can't think of an example the other direction, because English would toss in a dummy subject. The closest I can think of is something like "It hurts" which is really "It hurts me" with the "me" part being implied. That's a verb where you can leave out the agent (who or what is doing the hurting) but the patient is sort of required (someone or something is being hurt, and it doesn't make sense not to say who/what). Am I sort of accurate there? Or getting confused, lol?
Sounds like you've got it. I can't name a language that requires "She ate" to use the antipassive, but only because I don't know many languages by name that have an antipassive; I would be completely unsurprised to find a language with such a rule.
Doranwen wrote:In a tripartite system, then what does one do about impersonal verbs, like the classic "to rain"? It's sort of both passive and antipassive at the same time, because it has no arguments at all. Does it get a marker all its own, or no marker at all, or something I haven't thought of? Obviously, if this is my conlang, I can do whatever, but if I take that attitude to the extreme, it'll make no sense to anyone, much less myself, so I'm trying to at least understand how natural languages do it (when possible), or what might make the most sense.
Languages handle this in a variety of ways; English's approach of sticking in a meaningless subject is probably the weirdest I'm aware of. A few languages I know simply allow rain as a verb with no subject. Japanese and Mandarin treat "rain" as the subject and add a verb, so instead of "It's raining" you say "Rain is falling."
Doranwen wrote:You mentioned "things that aren't cases"; what other options for expressing things like that are there, without resorting to prepositions?
Participial clauses are one high-mileage way. Spanish, being verb-framed instead of satellite-framed, does things like this:

Salimos corriendo del cuarto.
we.exited run.PTCP of.the room
We ran out of the room.

Spanish happens to tag the noun with "of", but you could easily imagine a language that doesn't, and instead just places the noun as the participle's complement without any preposition. So what English does with the preposition out of, Spanish instead encodes directly on the verb, salimos. Or, if you want your language to still use "run" as the main verb in situations like this, you could do something like:

oss-ta mar tar-us harek
we-ABS room exit-PTCP run
We ran out of the room.

Here, "tarus" is a participial form of a verb (say, "tarek") meaning to exit. This language handles what English would do with a preposition by allowing verbs in participial form to be dropped into phrases that already have main verbs, to provide extra meaning.

(There's interesting details to work out, here. I have the object of the participle, mar, taking no case marker; so I would have to decide, in creating this language, how you know which argument of a participle-form verb the object plays. Is it the ergative? Accusative? Absolutive? Any of the three, and you just infer which one from context? Can you have more than one argument in a participial phrase? Are there, perhaps, different participle forms that you can pick from depending on which argument you want to use?)

A similar example might look like:

oss-ta ranu istel-us hospek
we-ABS family circle-PTCP speak
We talked about our families.

Here, istelus is a participle form of istelek, meaning "to circle around". It, perhaps, gets paired frequently with hospek to indicate topics of conversation.
Doranwen wrote:Regarding cases, the next challenge I found was how to express the differences between sentences such as the following:

A: He left without a horse.
B: He left without his horse.

Assuming that one uses the abessive case, how would I distinguish them? I would rather combine affixes to create one long word than have multiple small ones, but for sentence B you've got really three nouns, right? (counting pronouns as nouns) He, horse, him. Mind you, I rather liked the clitic idea, throwing shortened or special versions of the pronouns onto the verb to get rid of pronouns except for emphasis, so it would then turn out: he-left horse-ABE he-GEN
I'm not sure I see the conflict; even without marking person on the verb, what's to stop you from having something like this?

he-ABS left horse-ABE he-GEN
Doranwen wrote:So I guess what I'm asking is, I know that for possessive I'd use the genitive case; would there be a way to attach a possessive *pronoun*, sort of like the clitic on the verb, to the noun it's possessing?
Certainly; Arabic has possessive suffixes, for example.
Doranwen wrote:Is there a name for that? (Wikipedia had something called the "possessive affix", is that it?)
Yep. There is also the idea of possessive head-marking, but that may carry more typological baggage than you want.
Doranwen wrote:How do I express things like conjunctions such as "and" when there's free word order? I'm assuming cases help. So if I have the English sentence "Alice loves Bob and Chris", how do I mark it to express that? The only thing I can come up with is: Alice-ERG love Bob-ACC Chris-AC
(Note that I'm sticking with the tripartite system here, and I'm typing the case abbreviations just to keep things straight in my head--I can sort out the actual marking, whether it's ergative or absolutive that's unmarked. And obviously I can stick Alice after the verb, if I so choose; it appears that the only tripartite languages I can find anything about are either SOV or have no dominant word order.) Essentially keeping them both in the same case and letting one assume that it means they're both the "object" of her love. Is this something that's done? Is it doable? Or is there an existing standard way to do this?
There are many ways to do this; what you've described (where there is no overt morpheme tying Bob-ACC and Chris-ACC together) is "juxtaposition", which is one way that languages coordinate constituents.

You could also have a particle that you place with each coordinated noun:

Alice-ERG love also Bob-ACC also Chris-ACC

Or do something like Latin -que, and have a suffix that you can add to a noun for coordination:

Alice-ERG love Bob-ACC Chris-ACC-also

And, like some others have said, you could consider how "free" you want your word order to be. Maybe placement of noun phrases relative to the verb is free, but coordinated noun phrases have to go together. (This would probably make the juxtaposition strategy more tenable in complex sentences.)
Doranwen wrote:Then I come to how to combine two sentences, and wonder if there I need a particle or whether an affix can be applied to the beginning of the second phrase (this may only be tangentially related to the case thing, but it's all mixed up in my mind). So if I had the sentence "Mary ate the apple and went into the house", this is as far as I can figure out how to mark: Mary-ERG ate apple-ACC "and"(?) went house-ILL. Are there any other ways to express the "and" in this sentence besides a particle/word like the English "and"?
Certainly. Different languages would do this differently, depending on what you intend it to mean. Did Mary start by eating an apple, and once finished, go into the house? Did she both eat an apple and go into the house, though not necessarily in that order? Did she go into the house while eating an apple?

Some possibilities off the top of my head:

Mary-ERG first ate apple-ACC next went house-ILL

Mary-ERG ate apple-ACC went house-ILL

Mary-ERG eat-PTCP apple-ACC went house-ILL
任何事物的发展都是物极必反,否极泰来。
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Struggling to grasp cases and voice

Post by Doranwen »

kiwikami wrote:Marking direction/location on a verb might mean either morphemes added to a verb that show a location or direction of motion, or completely stems for different kinds of motion. The former might look something like the following:

bob-SUB hill-OBJ go-ABOVE
Bob goes up the hill.
Marking telicity (whether or not an action is completed) on the verb could distinguish then between that and Bob goes to the top of the hill.

The latter would be something like:
bob-SUB hill-OBJ go_up
Bob goes up the hill.

Where go_up is its own verb stem, separate from others like go_down, go_around, etc. (consider English "ascend" and "descend"). The first option, with separate morphemes, is perhaps more versatile here, because you could theoretically use it on any verb, possibly using the verb to make a locative-marked noun more specific. That is, something like this:

bob-SUB hill-LOC dance-ABOVE
Bob dances on top of the hill.
Yeah, I see what you mean about the versatility. To use the second method is to exchange a set of morphemes which can apply to a large range of verbs, for a set of new vocabulary, which means a lot more vocabulary creation.
kiwikami wrote:I'm a wee bit biased when it comes to marking direction and motion on verbs, because it's a feature I like to incorporate in my own conlangs, but it can certainly help make prepositions less of a thing without adding new cases for every possible method of getting from point A to point B. This could likely extend to non-location-related cases as well, but I'm not super certain how, or how common/likely this is. BUT it can also essentially lead to shifting a still-ridiculously-complicated case system over to the verb from the noun, rather than actually reducing case complexity, so maybe take everything I say with a grain of salt... :roll:
Hee, well, I do like the verb marker thing a lot too. Now I just have to decide whether I want to go with the nominal constructions or the verb markers. Well, one could always use the nominal constructions even if there were the verb markers, that would add variety, but I guess it's really a "do I want these verb markers" question. Something to sort out when I'm looking at the verbs, which I just started doing (and I'm getting confused again, so will probably start another thread when I get around to it).

The big thing I can see is, yeah, it just makes the verbs more complicated, and you still have to either combine things all over the place or end up with a huge set of markers, due to all the different prepositions there are. Not saying I might not go for it, but it's something to consider. Plus it doesn't really get rid of the cases, because wouldn't there be situations where you'd need them and the verb couldn't carry the meaning? I'm blanking on those right now, though, so maybe I'm mistaken.
Trailsend wrote:
Doranwen wrote:Ahh, so the passive/antipassive is markers on the verb itself (I so did not understand that part!) that allows the verbs to change the transitivity (valency?) from transitive to intransitive while still sticking with the rule that a verb has to have a subject.
Strictly speaking, only the bold part is true; you could have a passive and antipassive voice in a language that didn't require verbs to have subjects. A required subject is just one way in which they might be useful.

You could certainly imagine a tripartite language that does allow omitting arguments from transitive verbs. Maybe the following two statements are both grammatical:

1. na ruff-mi ak-in ketek
my dog-ERG you-ACC bite
my dog will bite you

2. na ruff-mi ketek
my dog-ERG bite
my dog bites (somebody)

So if you also have an antipassive voice construction:

3. na ruff-ta i-ketek
my dog-ABS ANTIPASSIVE-bite
my dog bites

Now you get to decide what that means. Why would a speaker choose to say (3) instead of (2)? What additional flavor, connotation, nuance does the antipassive voice convey in this language, which just leaving out the accusative argument would not?

(For example, maybe omitting the accusative argument like in (2) can be used to describe specific events, while the antipassive is used to describe general truths or ongoing states. Thus, (2) would mean My dog bit somebody, but I'm not mentioning who, while (3) would mean My dog, as a general rule, bites.)
Ooh, I like this. No new morphemes, but adds new levels of meaning possibilities. The only problem is that now that you gave that example, I can't think of any other ways to use it, lol. Do you have any more possibilities that come to mind? I find that if I'm presented with one example of something, my brain has a very hard time thinking of another way it could be, but if I'm presented with two or three, then I start to see a pattern in how they work, and I have a chance of coming up with something that hasn't been shown to me already.
Trailsend wrote:
Doranwen wrote:And then of course there are verbs like "eat" which can be either transitive or intransitive, right? Well, in English they do. "She ate" vs. "She ate the apple". Are there languages that require the "ate" from "She ate" to be marked antipassive, then? I can't think of an example the other direction, because English would toss in a dummy subject. The closest I can think of is something like "It hurts" which is really "It hurts me" with the "me" part being implied. That's a verb where you can leave out the agent (who or what is doing the hurting) but the patient is sort of required (someone or something is being hurt, and it doesn't make sense not to say who/what). Am I sort of accurate there? Or getting confused, lol?
Sounds like you've got it. I can't name a language that requires "She ate" to use the antipassive, but only because I don't know many languages by name that have an antipassive; I would be completely unsurprised to find a language with such a rule.
Your earlier example of how to use the antipassive marker to encode extra meaning (in which case is it still called the antipassive?) reminds me of a form of aspect; using your example, I could imagine "She ate" without the antipassive being used to refer to a specific time she ate (something, we aren't told what), whereas "She ate" with the antipassive would mean that generally she ate and wasn't trying to starve herself? (It's a bit of a stretch with some verbs but I can see how that might work.)
Trailsend wrote:
Doranwen wrote:In a tripartite system, then what does one do about impersonal verbs, like the classic "to rain"? It's sort of both passive and antipassive at the same time, because it has no arguments at all. Does it get a marker all its own, or no marker at all, or something I haven't thought of? Obviously, if this is my conlang, I can do whatever, but if I take that attitude to the extreme, it'll make no sense to anyone, much less myself, so I'm trying to at least understand how natural languages do it (when possible), or what might make the most sense.
Languages handle this in a variety of ways; English's approach of sticking in a meaningless subject is probably the weirdest I'm aware of. A few languages I know simply allow rain as a verb with no subject. Japanese and Mandarin treat "rain" as the subject and add a verb, so instead of "It's raining" you say "Rain is falling."
I like the verb without a subject deal. Gets to the heart of things and is the most compact way that I can think of. I like succinctness. :)
Trailsend wrote:
Doranwen wrote:You mentioned "things that aren't cases"; what other options for expressing things like that are there, without resorting to prepositions?
Participial clauses are one high-mileage way. Spanish, being verb-framed instead of satellite-framed, does things like this:

Salimos corriendo del cuarto.
we.exited run.PTCP of.the room
We ran out of the room.

Spanish happens to tag the noun with "of", but you could easily imagine a language that doesn't, and instead just places the noun as the participle's complement without any preposition. So what English does with the preposition out of, Spanish instead encodes directly on the verb, salimos. Or, if you want your language to still use "run" as the main verb in situations like this, you could do something like:

oss-ta mar tar-us harek
we-ABS room exit-PTCP run
We ran out of the room.

Here, "tarus" is a participial form of a verb (say, "tarek") meaning to exit. This language handles what English would do with a preposition by allowing verbs in participial form to be dropped into phrases that already have main verbs, to provide extra meaning.

(There's interesting details to work out, here. I have the object of the participle, mar, taking no case marker; so I would have to decide, in creating this language, how you know which argument of a participle-form verb the object plays. Is it the ergative? Accusative? Absolutive? Any of the three, and you just infer which one from context? Can you have more than one argument in a participial phrase? Are there, perhaps, different participle forms that you can pick from depending on which argument you want to use?)

A similar example might look like:

oss-ta ranu istel-us hospek
we-ABS family circle-PTCP speak
We talked about our families.

Here, istelus is a participle form of istelek, meaning "to circle around". It, perhaps, gets paired frequently with hospek to indicate topics of conversation.
Wow, that's a way I never would have thought of. I don't think I'll use it myself (I really like the two examples kiwikami came up with--the nominal constructions and/or verb markers), but thank you for explaining it! For all that I can speak Spanish, I did not "see" that usage.
Trailsend wrote:
Doranwen wrote:Regarding cases, the next challenge I found was how to express the differences between sentences such as the following:

A: He left without a horse.
B: He left without his horse.

Assuming that one uses the abessive case, how would I distinguish them? I would rather combine affixes to create one long word than have multiple small ones, but for sentence B you've got really three nouns, right? (counting pronouns as nouns) He, horse, him. Mind you, I rather liked the clitic idea, throwing shortened or special versions of the pronouns onto the verb to get rid of pronouns except for emphasis, so it would then turn out: he-left horse-ABE he-GEN
I'm not sure I see the conflict; even without marking person on the verb, what's to stop you from having something like this?

he-ABS left horse-ABE he-GEN
I think I was asking how to handle possessives and didn't realize I was asking that, lol. My brain is like that sometimes. Though now that brings a question to mind: since I don't really want to mix the clitics with whatever suffixes I choose (because I think I already know I want suffixes on my verbs for the TAM stuff), I would put them as prefixes. Which means, how do I distinguish between ergative and accusative clitics vs. absolutive ... do I need different forms of each one? Do I then need to use the passive or antipassive markers to help distinguish? Obviously prefix order will help if both ergative and accusative are pronouns ("He loves her"), and the other words in the sentence would be marked in situations where only one or the other were pronouns ("He loves the dog" vs. "The dog loves him"), and if I had things like "He is loved" and I use the passive marker then it's obvious that the "he" clitic is meant in the accusative and any other). I guess that would interact with the whole "is the passive or antipassive marker explicitly required" question. So maybe sometimes it's required for clarity, whereas other times it can be left out and then when it's included it encodes extra meaning...?
Trailsend wrote:
Doranwen wrote:So I guess what I'm asking is, I know that for possessive I'd use the genitive case; would there be a way to attach a possessive *pronoun*, sort of like the clitic on the verb, to the noun it's possessing?
Certainly; Arabic has possessive suffixes, for example.
Yeah, I like that. :)
Trailsend wrote:
Doranwen wrote:Is there a name for that? (Wikipedia had something called the "possessive affix", is that it?)
Yep. There is also the idea of possessive head-marking, but that may carry more typological baggage than you want.
Wow, that is confusing. Happy to avoid that if I can.
Trailsend wrote:
Doranwen wrote:How do I express things like conjunctions such as "and" when there's free word order? I'm assuming cases help. So if I have the English sentence "Alice loves Bob and Chris", how do I mark it to express that? The only thing I can come up with is: Alice-ERG love Bob-ACC Chris-AC
(Note that I'm sticking with the tripartite system here, and I'm typing the case abbreviations just to keep things straight in my head--I can sort out the actual marking, whether it's ergative or absolutive that's unmarked. And obviously I can stick Alice after the verb, if I so choose; it appears that the only tripartite languages I can find anything about are either SOV or have no dominant word order.) Essentially keeping them both in the same case and letting one assume that it means they're both the "object" of her love. Is this something that's done? Is it doable? Or is there an existing standard way to do this?
There are many ways to do this; what you've described (where there is no overt morpheme tying Bob-ACC and Chris-ACC together) is "juxtaposition", which is one way that languages coordinate constituents.

You could also have a particle that you place with each coordinated noun:

Alice-ERG love also Bob-ACC also Chris-ACC

Or do something like Latin -que, and have a suffix that you can add to a noun for coordination:

Alice-ERG love Bob-ACC Chris-ACC-also

And, like some others have said, you could consider how "free" you want your word order to be. Maybe placement of noun phrases relative to the verb is free, but coordinated noun phrases have to go together. (This would probably make the juxtaposition strategy more tenable in complex sentences.)
Yeah, I'll have to work that out as I come to them. I should probably scale off on thinking about those till I have some of the more fundamental building blocks sorted, then I can start trying basic sentences on for size. I do like the way Latin does it, actually, and had come up with something similar on my own for the old version of this conlang, so I will probably stick with it. Something like:
Alísas sar Babét Crísét'si. or Alísas sar Babét Crísét-si. (but I like the apostrophe better than the hyphen)
Alice-ERG love Bob-ACC Chris-ACC-and.
*eyes sentences and decides she really needs to get enough stuff developed so she can tweak and play around with things more*
Doranwen wrote:Then I come to how to combine two sentences, and wonder if there I need a particle or whether an affix can be applied to the beginning of the second phrase (this may only
Trailsend wrote:be tangentially related to the case thing, but it's all mixed up in my mind). So if I had the sentence "Mary ate the apple and went into the house", this is as far as I can figure out how to mark: Mary-ERG ate apple-ACC "and"(?) went house-ILL. Are there any other ways to express the "and" in this sentence besides a particle/word like the English "and"?
Certainly. Different languages would do this differently, depending on what you intend it to mean. Did Mary start by eating an apple, and once finished, go into the house? Did she both eat an apple and go into the house, though not necessarily in that order? Did she go into the house while eating an apple?

Some possibilities off the top of my head:

Mary-ERG first ate apple-ACC next went house-ILL

Mary-ERG ate apple-ACC went house-ILL

Mary-ERG eat-PTCP apple-ACC went house-ILL
Interesting! I'll mull these over some more. :) I think at this point they aren't interacting with the cases questions so much, and I need to figure out more what I can combine for the purposes of fusional, inflections and all that (case + number? you can see my ramblings on that here). And then figure out everything I want on the verb besides the passive/antipassive markers (whenever those get used). :)
Last edited by Doranwen on 20 Jul 2016 03:22, edited 5 times in total.
Keenir
mayan
mayan
Posts: 2399
Joined: 22 May 2012 03:05

Re: Struggling to grasp cases and voice

Post by Keenir »

Doranwen wrote: Plus it doesn't really get rid of the cases, because wouldn't there be situations where you'd need them and the verb couldn't carry the meaning? I'm blanking on those right now, though, so maybe I'm mistaken.
there probably maybe are such situations...but do you want to have something made especially for those rare instances? or can other things handle the weight of those instances? (maybe a dummy like the it of "it rains"?)
At work on Apaan: viewtopic.php?f=6&t=4799
Trailsend
moderator
moderator
Posts: 1451
Joined: 18 Aug 2010 05:22

Re: Struggling to grasp cases and voice

Post by Trailsend »

Doranwen wrote:Ooh, I like this. No new morphemes, but adds new levels of meaning possibilities.
To give you a glimpse of the future: one common progression you see as conlangers gain more experience is that they start with a very heavy focus on morphology (ways to change the forms of words) and only later begin to explore syntax (ways to manipulate the structure of phrases).

In part, this is natural, because while morphology tends to be self-contained, syntax necessarily involves the interactions between other systems of the language. E.g., the passive/antipassive examples above involve an interaction between the morphology of verbs (adding the passive and antipassive prefixes), the morphology of nouns (switching the case marking), and the transitivity of verbs.

But you've already caught on to a very powerful feature of syntax: it allows you to take just the morphemes you already have, and use them to open up whole new dimensions of meaning.
Doranwen wrote:The only problem is that now that you gave that example, I can't think of any other ways to use it, lol. Do you have any more possibilities that come to mind? I find that if I'm presented with one example of something, my brain has a very hard time thinking of another way it could be, but if I'm presented with two or three, then I start to see a pattern in how they work, and I have a chance of coming up with something that hasn't been shown to me already.
Sure. Are you looking for more examples of this one possible connotation of the passive/antipassive (a "general rule" instead of a "specific event")? Or, are you looking for examples of more possible connotations the passive/antipassive could have?

Here's another "general rule" vs. "specific event" possibility:

hu-mi tenu-se lestek
he-ERG wood-GEN build
He built (it) with wood. (Perhaps in response to the question, "What did he use to build the house?")

hu-ta tenu-se o-lestek
he-ABS wood-GEN ANTIPASSIVE-build
He builds with wood (generally speaking).

Here's another possible connotation that the passive voice could have:

hu-in ketek
he-ACC bite
He was bitten.

hu-ta i-ketek
he-ABS PASSIVE-bite
He was bitten, poor thing! (Perhaps the passive voice adds a connotation that there were negative consequences for the subject.)

Or maybe the different voices play a role all the way up into discourse structure, and phrases in the passive or antipassive voice are used to introduce new entities into the conversation, while phrases in the active voice are used to talk about things that have already been introduced.

ruff-mi na-in ketek
dog-ERG me-ACC bite
The dog bit me. (You know which dog I'm talking about because I've mentioned it already.)

ruff-ta na-ku o-ketek
dog-ABS me-LAT ANTIPASSIVE-bite
A dog bit me. (This is the first time I've mentioned the dog.)

(There's an interesting twist: in the last example, even though it's in the antipassive voice, I included both arguments. Languages typically have a way to do this; for example, while the English passive voice allows you to omit the agent of a transitive verb, you can certainly add it back in with a by phrase like I was bitten by a dog. For this example I decided that the way you add a patient back into an antipassive phrase is by putting it in lative case.)
Doranwen wrote:Your earlier example of how to use the antipassive marker to encode extra meaning (in which case is it still called the antipassive?)
Certainly, unless there was some other name that you found more useful. If you're not sure, you can always post a bundle of example data of how the structure gets used, and see if anyone thinks it could be better analyzed as something besides antipassive voice.
Doranwen wrote:reminds me of a form of aspect
True! Many languages would convey this information with an aspect marking on the verb; an aspect for statements of "general rules" is often called the "habitual" aspect.
Doranwen wrote:Though now that brings a question to mind: since I don't really want to mix the clitics with whatever suffixes I choose (because I think I already know I want suffixes on my verbs for the TAM stuff), I would put them as prefixes. Which means, how do I distinguish between ergative and accusative clitics vs. absolutive ... do I need different forms of each one? Do I then need to use the passive or antipassive markers to help distinguish? Obviously prefix order will help if both ergative and accusative are pronouns ("He loves her"), and the other words in the sentence would be marked in situations where only one or the other were pronouns ("He loves the dog" vs. "The dog loves him"), and if I had things like "He is loved" and I use the passive marker then it's obvious that the "he" clitic is meant in the accusative and any other). I guess that would interact with the whole "is the passive or antipassive marker explicitly required" question. So maybe sometimes it's required for clarity, whereas other times it can be left out and then when it's included it encodes extra meaning...?
All of this sounds vaguely plausible, but I'm not entirely sure what you mean, in particular concerning clitics. A "clitic" is an affix that, despite being phonologically dependent on another word, plays a syntactic role at the phrase level.

For example, English's possessive 's is a clitic. It's phonologically dependent on the preceding word, which you can tell because it undergoes the same phonological changes as the plural -s suffix (it gets pronounced as z or s or ez depending the ending of the host word). But its syntactic role is at the phrase level, which you can tell because in this phrase:

The queen of England's hat

The 's doesn't just mark "England" as the possessor; it marks the entire phrase "The queen of England" as the possessor.

Are the clitics you're talking about prefixes on verbs which indicate a particular pronoun is an argument of the verb? If so, they have one half of what you need to be a clitic (they're phonologically dependent), but it seems like you could more simply analyze them as a person conjugation on the verb. Could you give an example?
任何事物的发展都是物极必反,否极泰来。
Doranwen
rupestrian
rupestrian
Posts: 19
Joined: 18 Jul 2016 02:49

Re: Struggling to grasp cases and voice

Post by Doranwen »

Trailsend wrote:
Doranwen wrote:Ooh, I like this. No new morphemes, but adds new levels of meaning possibilities.
To give you a glimpse of the future: one common progression you see as conlangers gain more experience is that they start with a very heavy focus on morphology (ways to change the forms of words) and only later begin to explore syntax (ways to manipulate the structure of phrases).

In part, this is natural, because while morphology tends to be self-contained, syntax necessarily involves the interactions between other systems of the language. E.g., the passive/antipassive examples above involve an interaction between the morphology of verbs (adding the passive and antipassive prefixes), the morphology of nouns (switching the case marking), and the transitivity of verbs.

But you've already caught on to a very powerful feature of syntax: it allows you to take just the morphemes you already have, and use them to open up whole new dimensions of meaning.
Yeah, I'm only halfway a beginner--I have attempted conlanging before (and some linguistic concepts did stick in my head fairly OK)--but I've been posting in this forum because I feel like I'm still a beginner, I have TONS of questions, and I tend to need things explained on a simpler level with concrete examples when it comes to linguistics, apparently (and it took me until the other day to realize that every time I saw an example with a ton of phonemes I couldn't pronounce, my mind really did start shutting down and I would get nowhere on trying to process the example). My previous conlang attempts were very much focused on the morphology. This time I am still trying to sort it out some--figure out what needs to actually be there, but keep in mind ways that I can make it richer and develop more layers of meaning.
Trailsend wrote:Are you looking for more examples of this one possible connotation of the passive/antipassive (a "general rule" instead of a "specific event")? Or, are you looking for examples of more possible connotations the passive/antipassive could have?
More the latter.
Trailsend wrote:(Perhaps the passive voice adds a connotation that there were negative consequences for the subject.)
Oh fascinating! I wish my brain could imagine these things. I really can't--if I see clear examples of all these different sorts of ideas, I can start to mimic it and include things similar, but without the examples, it's like it can't even conceive that it's possible. I can learn them when I'm taught them in the course of learning a language, but creating a language is a whole different ball game. (I blame being somewhere on the autism spectrum.) So I appreciate your examples very much; it's like opening up a new world.
Trailsend wrote:Or maybe the different voices play a role all the way up into discourse structure, and phrases in the passive or antipassive voice are used to introduce new entities into the conversation, while phrases in the active voice are used to talk about things that have already been introduced.

ruff-mi na-in ketek
dog-ERG me-ACC bite
The dog bit me. (You know which dog I'm talking about because I've mentioned it already.)

ruff-ta na-ku o-ketek
dog-ABS me-LAT ANTIPASSIVE-bite
A dog bit me. (This is the first time I've mentioned the dog.)

(There's an interesting twist: in the last example, even though it's in the antipassive voice, I included both arguments. Languages typically have a way to do this; for example, while the English passive voice allows you to omit the agent of a transitive verb, you can certainly add it back in with a by phrase like I was bitten by a dog. For this example I decided that the way you add a patient back into an antipassive phrase is by putting it in lative case.)
Aha, yes, I'd run into something referring to that, vaguely. I take it the choice of which case to add it back in would vary by the language. And that of course if I'm using both passive and antipassive in this one, that I'd have to set different cases for each of those situations?
Trailsend wrote:
Doranwen wrote:Your earlier example of how to use the antipassive marker to encode extra meaning (in which case is it still called the antipassive?)
Certainly, unless there was some other name that you found more useful. If you're not sure, you can always post a bundle of example data of how the structure gets used, and see if anyone thinks it could be better analyzed as something besides antipassive voice.
:D I'll keep that in mind. As soon as I can get a little of the basic ideas for noun and verb worked out, I'm going to start playing with the sentences in more depth. Then I'll see how this all shakes out. Right now I'm storing a ton of these ideas in a document for retrieval later.
Trailsend wrote:
Doranwen wrote:reminds me of a form of aspect
True! Many languages would convey this information with an aspect marking on the verb; an aspect for statements of "general rules" is often called the "habitual" aspect.
I was thinking it was the habitual... So basically it could substitute for an aspect that's missing (to a limited degree--unless I start using it all the time, even on sentences with no absolutive, in which case it would really stop being the antipassive and just be an unusual way of marking an aspect), or as you gave examples of, it could encode how the speaker feels towards the situation ("poor thing") or introduce new entities... I'll keep brainstorming and see what I settle on. :) (I haven't even taken a thorough look at the aspects area so it's too early for me to decide exactly what I want here.)
Trailsend wrote:
Doranwen wrote:Though now that brings a question to mind: since I don't really want to mix the clitics with whatever suffixes I choose (because I think I already know I want suffixes on my verbs for the TAM stuff), I would put them as prefixes. Which means, how do I distinguish between ergative and accusative clitics vs. absolutive ... do I need different forms of each one? Do I then need to use the passive or antipassive markers to help distinguish? Obviously prefix order will help if both ergative and accusative are pronouns ("He loves her"), and the other words in the sentence would be marked in situations where only one or the other were pronouns ("He loves the dog" vs. "The dog loves him"), and if I had things like "He is loved" and I use the passive marker then it's obvious that the "he" clitic is meant in the accusative and any other). I guess that would interact with the whole "is the passive or antipassive marker explicitly required" question. So maybe sometimes it's required for clarity, whereas other times it can be left out and then when it's included it encodes extra meaning...?
All of this sounds vaguely plausible, but I'm not entirely sure what you mean, in particular concerning clitics. A "clitic" is an affix that, despite being phonologically dependent on another word, plays a syntactic role at the phrase level.

For example, English's possessive 's is a clitic. It's phonologically dependent on the preceding word, which you can tell because it undergoes the same phonological changes as the plural -s suffix (it gets pronounced as z or s or ez depending the ending of the host word). But its syntactic role is at the phrase level, which you can tell because in this phrase:

The queen of England's hat

The 's doesn't just mark "England" as the possessor; it marks the entire phrase "The queen of England" as the possessor.

Are the clitics you're talking about prefixes on verbs which indicate a particular pronoun is an argument of the verb? If so, they have one half of what you need to be a clitic (they're phonologically dependent), but it seems like you could more simply analyze them as a person conjugation on the verb. Could you give an example?
Yeah, sorry, I think I am using the word incorrectly, now that I look at it; I got it from one of the language construction guides and I don't think I knew the full way to use it (plus the way I used it for the previous language isn't quite the same here). Yes, I mean a prefix on a verb which means it is one of the arguments of the verb. And *headdesk* of course it's conjugation, I've thought that the other day--but my mind did not put two and two together and switch terminology. So yes! Conjugation! But in a way that allows a pronoun in any case to be on the verb. I guess if it's conjugation, then I might want to combine person with anything else I want to mark on the verb, in order to have just one inflection, otherwise I'm leaning towards the agglutinative side in that respect. (Which I don't suppose is a problem to have one area where the language is more agglutinating, it just breaks the mold a bit.) So let me reform the question: Does it make more sense to combine the person with any of the TAM bits (that I haven't sorted out yet), or are there natural languages that conjugate for other things as inflections but then use prefixes for the pronouns? Because while I love the idea of conjugating person as well, to make the verb as compact as possible... my pronoun system is slightly complicated and I can't see how easy it would be to make THAT many conjugations to still respect some of the distinctions I want encoded.

(Lol, this thread has gone from cases to verbs now, in some ways. I should start a new one on the verb questions, but they all are interacting with each other...)
Post Reply