Conlang Random Thread

[Ahzoh]

That kinda contradict each other .

The verb is the head of the phrase, not the subject. So, VSO or VOS word order.

I have some difficulty with an animate-inanimate triggered pattern to have nom/abs and acc/erg. In fact, I don't think there's any language that has all four like this?

It's not nom/abs and acc/erg. If noun is animate gender it takes nom/acc cases. If noun is inanimate it takes erg/abs cases. Also do note the absolutive case marker would be the same as the accusative marker, and the ergative marker the same as the instrumental marker.

No matter what, there isn't a change in alignment, only word order.

[jal]

The verb is the head of the phrase, not the subject. So, VSO or VOS word order.

Yes, but the object is closer to the verb than the subject, so VOS makes it more head-initial than VSO. "Eating a peach" vs. "(Eating a peach) by me" (where "by" is the ergative).

It's not nom/abs and acc/erg.

Except you wrote:

eat\REAL-3sg Sam-NOM orange-ABS (VSO)
fall_on\REAL-3sg mouse-ACC pillar-ERG (VOS)

So don't blame me for misunderstanding :D.

If noun is animate gender it takes nom/acc cases. If noun is inanimate it takes erg/abs cases.

Ok, so you have split ergativity based on animacy?

Also do note the absolutive case marker would be the same as the accusative marker, and the ergative marker the same as the instrumental marker.

If they are equal, I wouldn't name them differently. Also, the absolutive is typically unmarked, as is the nominative. Accusative and ergative are typically marked.

No matter what, there isn't a change in alignment, only word order.

That doesn't make sense to me. You literaly change the alignment from nom/acc to abs/erg, so how is this not a change in alignment??

What would make sense to me, is that given the strong head-initiality, combined with the desire to have a word order be animate - non-animate, that in case the animate participant is the patient, one would use abs/erg, so that the less marked participant is closer to the verb. Would also be swell if you then would have head-marking, instead of dependent marking, of the animate participant :).


JAL

[Ahzoh]

The verb is the head of the phrase, not the subject. So, VSO or VOS word order.

Yes, but the object is closer to the verb than the subject, so VOS makes it more head-initial than VSO. "Eating a peach" vs. "(Eating a peach) by me" (where "by" is the ergative).

What? VSO is equally as head-initial as VOS, just that the object in VSO is decoupled from the verb.

Ok, so you have split ergativity based on animacy?

Like Hittite

Also do note the absolutive case marker would be the same as the accusative marker, and the ergative marker the same as the instrumental marker.

If they are equal, I wouldn't name them differently. Also, the absolutive is typically unmarked, as is the nominative. Accusative and ergative are typically marked.

I wouldn't because calling the absolutive an accusative is confusing as fuck and will confuse other people when they encounter an accusative-marked noun in subject position. Also it fucks with the way I set up my declension tables. If I had only animate and inanimate gender, I'd probably have a table with, in descending order, nominative, accusative, and ergative. But I have three animate genders and one inanimate gender, so it would just result in an unnecessary and asymmetrical amount of empty space.

Also, my case system originates from a tripartite system, so all the of the cases have suffixes (-m [animate nominative] -s [accusative/absolutive], and -n [ergative-instrumental]) and there are no null-morpheme marked cases.

No matter what, there isn't a change in alignment, only word order.

That doesn't make sense to me. You literaly change the alignment from nom/acc to abs/erg, so how is this not a change in alignment??

This was in reference to the aforementioned "hierarchical alignment"

What would make sense to me, is that given the strong head-initiality, combined with the desire to have a word order be animate - non-animate, that in case the animate participant is the patient, one would use abs/erg, so that the less marked participant is closer to the verb. Would also be swell if you then would have head-marking, instead of dependent marking, of the animate participant .

This is more complicated than it's supposed to be. I just wanted to make animate constituents more topical than inanimate ones.

Word-order apparently has these principles:

• Topic-first principle (more topical NPs come before less topical NPs)
• Animate-first principle (NPs with more animate referents come before NPs with less animate referents)
• Verb-object bonding principle (objects are more closely linked to the verb than subjects are)

Vrkhazhian default verb order is VSO, which satisfies the topic-first principle and animate-first principle, but not verb-object bonding principle

Way I see it, requiring animate objects to precede inanimate subjects is a way to satisfy verb-object bonding principle and animate-first principle, though it would be ambiguous if it would satisfy the topic-first principle, since I reason that the subject is the topic and the verb phrase is the comment in a topic phrase, but inanimates would be less topical than animates.

[bradrn]

The verb is the head of the phrase, not the subject. So, VSO or VOS word order.

Yes, but the object is closer to the verb than the subject, so VOS makes it more head-initial than VSO. "Eating a peach" vs. "(Eating a peach) by me" (where "by" is the ergative).

What? VSO is equally as head-initial as VOS, just that the object in VSO is decoupled from the verb.

You are correct here. VSO and VOS are equally ‘head-initial’.

Also do note the absolutive case marker would be the same as the accusative marker, and the ergative marker the same as the instrumental marker.

If they are equal, I wouldn't name them differently. Also, the absolutive is typically unmarked, as is the nominative. Accusative and ergative are typically marked.

I wouldn't because calling the absolutive an accusative is confusing as fuck and will confuse other people when they encounter an accusative-marked noun in subject position.

What’s more confusing is giving two names to what is essentially a single case with a single function. Besides, what do you call it in intransitive sentences? I’d just choose a single name and stick to it, whether that single name is ‘accusative’ or ’absolutive’ or ‘unmarked’ or ‘sole’ or whatnot.

(Also, there are plenty of languages with accusative-marked subjects. Icelandic comes to mind.)

Word-order apparently has these principles:

• Topic-first principle (more topical NPs come before less topical NPs)
• Animate-first principle (NPs with more animate referents come before NPs with less animate referents)
• Verb-object bonding principle (objects are more closely linked to the verb than subjects are)

Vrkhazhian default verb order is VSO, which satisfies the topic-first principle and animate-first principle, but not verb-object bonding principle

Yet again, I will repeat: don’t rely on Wikipedia’s linguistics article! At least this part you’ve quoted isn’t so bad, in that it isn’t totally incorrect… it’s just frustratingly imprecise. It doesn’t define ‘topic’, and if one assumes the usual definition, it’s not necessarily true that ‘subjects are generally more likely than objects to be topics’.

In any case, your word order system isn’t topic-sensitive in any way: it’s about as pure an expression of this ‘animate-first principle’ as you could find.

[Ahzoh]

What’s more confusing is giving two names to what is essentially a single case with a single function.

But it doesn't have a single function. In animate nouns, it marks only an object of a transitive, not a subject of a transitive or instransitive. In inanimate nouns, it marks both the object of a transitive and the subject of an intransitive. So it is either a case with two functions or two separate cases with a shared origin. It's not like this hasn't been done before where people decide to interpret a single marker as signifying two separate accusative and absolutive cases.

Besides, what do you call it in intransitive sentences?

Absolutive. And animate nouns would take a different marker, the nominative.

(Also, there are plenty of languages with accusative-marked subjects. Icelandic comes to mind.)

Quirky subjects are not the same as absolutives. They tend to exist as a function of the verb's lexical or whatever meaning (such as being unergative or whatever).

Yet again, I will repeat: don’t rely on Wikipedia’s linguistics article! At least this part you’ve quoted isn’t so bad, in that it isn’t totally incorrect… it’s just frustratingly imprecise. It doesn’t define ‘topic’, and if one assumes the usual definition, it’s not necessarily true that ‘subjects are generally more likely than objects to be topics’.

It isn't Wikipedia making these claims, it's Russell Tomlin, the paper writer from which these claims are taken. Also it says "tends to be more topical", not "is the topic".

[bradrn]

What’s more confusing is giving two names to what is essentially a single case with a single function.

But it doesn't have a single function. In animate nouns, it marks only an object of a transitive, not a subject of a transitive or instransitive. In inanimate nouns, it marks both the object of a transitive and the subject of an intransitive.

Yeah, so its single function is ‘object marking’. And similarly the single function of your nominative and ergative cases are ‘subject marking’. It just so happens that an intransitive argument can align as either ‘subject’ or ‘object’, depending on animacy. That is, what you have here is a split-intransitive system.

Having thought through it a bit more, the name I’d suggest is simply ‘accusative’. We have many well-known cases where ostensibly ‘ergative’ cases are extended to intransitive arguments (e.g. in Tibetan and Basque); this is simply the dual case. Alternately, some more modern Sino-Tibetan grammars have used the term ‘agentive’ for that not-quite-ergative case, so you could similarly call yours the ‘objective’ case.

(Also, there are plenty of languages with accusative-marked subjects. Icelandic comes to mind.)

Quirky subjects are not the same as absolutives. They tend to exist as a function of the verb's lexical or whatever meaning (such as being unergative or whatever).

Indeed, it’s not the same thing. I’m just pointing out that precedent exists for such terminological confusions.

Yet again, I will repeat: don’t rely on Wikipedia’s linguistics article! At least this part you’ve quoted isn’t so bad, in that it isn’t totally incorrect… it’s just frustratingly imprecise. It doesn’t define ‘topic’, and if one assumes the usual definition, it’s not necessarily true that ‘subjects are generally more likely than objects to be topics’.

It isn't Wikipedia making these claims, it's Russell Tomlin, the paper writer from which these claims are taken. Also it says "tends to be more topical", not "is the topic".

I’m not even sure that claim is correct. (Note that topic-prominent languages are well-known for topics which don’t even participate in the argument structure at all.)

(Also, you linked to Wikipedia, not Tomlin’s paper, so I think I’m justified in mentioning that.)

[Jonlang]

In a conlang with cases, what consideration should be given to adpositions and which cases they're used with? I've seen European languages which will tend to always use prepositions with, say, the genitive or the dative. Now, my plan is to have adpositions give context to case, e.g. where the locative case doesn't make it clear whether something is on/in/on top of/near when there is very real ambiguity. But why then would a preposition be used with a genitive or a dative?

[keenir]

In a conlang with cases, what consideration should be given to adpositions and which cases they're used with? I've seen European languages which will tend to always use prepositions with, say, the genitive or the dative. Now, my plan is to have adpositions give context to case, e.g. where the locative case doesn't make it clear whether something is on/in/on top of/near when there is very real ambiguity.

Thats a nice plan, a good one.

But why then would a preposition be used with a genitive or a dative?

um...because those European languages use their prepositions with them?

[Glass Half Baked]

Europe does indeed have a paucity of examples to look to. Latin used few prepositions with the genitive and dative, and most modern languages in Western Europe don't have a separate dative anyway. The best example is probably German, which uses in differently for accusative and dative. We get more examples further afield. For example, Finnish uses kesken with the genitive to mean basically "between." Romanian has some prepositions that pair with the genitive because they used to be nouns, something like "top (of) the dresser." And I skipped the Slavic languages, because it sounds like you've already gone down that rabbit hole.

[Ahzoh]

In a conlang with cases, what consideration should be given to adpositions and which cases they're used with? I've seen European languages which will tend to always use prepositions with, say, the genitive or the dative.

The concept is known as "case government". From my understanding, the adposition itself determines what case a noun can/should be in. Origin of the adposition could be a factor.

[Travis B.]

One thing to remember is the case of StG, where a number of adpositions change meaning depending on whether the noun you use them with is in the accusative or dative case.

[Glass Half Baked]

You're both right. The noun and the adposition together form a unit of meaning. One preposition can mean different things with different cases, and one case can mean different things with different prepositions. A good example is Latin in tabernā. The preposition means location in, not motion towards as it would if it were used with the accusative. But it's not like the ablative by itself is a locative case. Except for a few words, the bare ablative of a noun is very rarely used to indicate location. Neither _ tabernā nor in tabern_ would get the job done. The two together indicate the total meaning. Languages with case suffixes and postpositions, like a lot of "Uralic" languages for lack of a better term, make this even more clear by blurring the line between the two.

[Travis B.]

Languages with case suffixes and postpositions, like a lot of "Uralic" languages for lack of a better term, make this even more clear by blurring the line between the two.

The only reason why Finnic languages are treated as having a large number of cases rather than a smaller number of cases and a large number of postpositions is that adjectives agree with their nouns with regard to case, incorporating such endings, presumably under IE influence.

[hwhatting]

The only reason why Finnic languages are treated as having a large number of cases rather than a smaller number of cases and a large number of postpositions is that adjectives agree with their nouns with regard to case, incorporating such endings, presumably under IE influence.

It's not the only reason; the other is that the endings are fused to the nouns. As another example, Hungarian doesn't have agreement on adjectives, but it's usually described as having cases and case suffixes (as e.g. in Wikipedia, but I've seen that in other descriptions as well), because the endings are fused and show vowel harmony.

[jal]

What? VSO is equally as head-initial as VOS, just that the object in VSO is decoupled from the verb.

You are correct here. VSO and VOS are equally ‘head-initial’.

I disagree, though I may well be in disagreement with current linguistic theory. In my opinion, verb + object = predicate, and a predicate is the head of the subject. So I would consider VOS head-initial, and VSO not, or at least, not fully.


JAL

[bradrn]

What? VSO is equally as head-initial as VOS, just that the object in VSO is decoupled from the verb.

You are correct here. VSO and VOS are equally ‘head-initial’.

I disagree, though I may well be in disagreement with current linguistic theory. In my opinion, verb + object = predicate, and a predicate is the head of the subject. So I would consider VOS head-initial, and VSO not, or at least, not fully.

‘Verb + object’ being a single constituent is very much a language-dependent phenomenon, not a universal. In some languages, like English, you can make a strong case for the ‘verb phrase’ being a single entity. In many others, the evidence is much weaker. And for VSO languages, it’s quite difficult to make this claim at all.

The term ‘predicate’ is also a difficult one. R.M.W. Dixon, for instance, restricts ‘predicate’ to refer to the verb alone (plus its immediate modifiers), which to me seems a sensible approach. If you define it as ‘verb + object’, you’re forced to conclude that VSO languages don’t have predicates — and therefore their clauses don’t have any head at all! ‘Head’ may be an ill-defined term, but to me, that’s going a bit too far.

[Travis B.]

The only reason why Finnic languages are treated as having a large number of cases rather than a smaller number of cases and a large number of postpositions is that adjectives agree with their nouns with regard to case, incorporating such endings, presumably under IE influence.

It's not the only reason; the other is that the endings are fused to the nouns. As another example, Hungarian doesn't have agreement on adjectives, but it's usually described as having cases and case suffixes (as e.g. in Wikipedia, but I've seen that in other descriptions as well), because the endings are fused and show vowel harmony.

Oh I completely forgot about vowel harmony! D'oh!

[Ahzoh]

Yeah, so its single function is ‘object marking’. And similarly the single function of your nominative and ergative cases are ‘subject marking’. It just so happens that an intransitive argument can align as either ‘subject’ or ‘object’, depending on animacy. That is, what you have here is a split-intransitive system.

But it has two functions, one being "marks O/P" for animate and inanimate and the other being "marks S" if inanimate.

My understanding is that split-intransitive would have only two cases, agentive and patientive and would not split along animacy lines and would have it that all verbs would require all nouns to decide if the intransitive is volitional or involitional (I fell vs me fell)

Having thought through it a bit more, the name I’d suggest is simply ‘accusative’. We have many well-known cases where ostensibly ‘ergative’ cases are extended to intransitive arguments (e.g. in Tibetan and Basque); this is simply the dual case. Alternately, some more modern Sino-Tibetan grammars have used the term ‘agentive’ for that not-quite-ergative case, so you could similarly call yours the ‘objective’ case.

There is just no good way I could reorganize the table without creating a lot of unnecessary empty space.

Vrkhazhian basically does what Hittite does but apparently Hittite also allows animates to be marked with an ergative even though Vrkhazhian would not. So my grammar table would just have a blank space for the three animate genders.

Nounsinfullness.png (35.18 KiB) Viewed 2710 times

[bradrn]

Yeah, so its single function is ‘object marking’. And similarly the single function of your nominative and ergative cases are ‘subject marking’. It just so happens that an intransitive argument can align as either ‘subject’ or ‘object’, depending on animacy. That is, what you have here is a split-intransitive system.

But it has two functions, one being "marks O/P" for animate and inanimate and the other being "marks S" if inanimate.

I’m quite thoroughly confused now. What do S/O/P stand for here? (The system I’m familiar with is A/O/P.)

And anyway, what you previously said was that it marks the object of a transitive verb for animate nouns, and the object of a transitive verb or the sole argument of an intransitive verb for inanimate nouns. Now you seem to be saying something very different.

My understanding is that split-intransitive would have only two cases, agentive and patientive and would not split along animacy lines and would have it that all verbs would require all nouns to decide if the intransitive is volitional or involitional (I fell vs me fell)

‘Split intransitive’ just means that there is some kind of split in the marking of the sole intransitive argument. The most common subtypes of split-intransitive systems are active-stative systems, which are the kind you describe. But there are other systems as well: e.g. Yapen and Choctaw have a three-way split.

There is just no good way I could reorganize the table without creating a lot of unnecessary empty space.

I don’t see why that matters…?

[Ahzoh]

I’m quite thoroughly confused now. What do S/O/P stand for here? (The system I’m familiar with is A/O/P.)

Subject or Sole of intransitives; A for agent of transitives, O/P for object/patient of transitives.

S is what you call the sole argument of intransitives
A and P (or O) are what you call the arguments of transitives
That's why people say nominative is S=A and absolutive is S=P/O

Case -m marks animate S of intransitives and animate A of transitives
Case -s marks both animate and inanimate O/P of transitives but also inanimate S of intransitives
Case -n marks inanimate A

And anyway, what you previously said was that it marks the object of a transitive verb for animate nouns, and the object of a transitive verb or the sole argument of an intransitive verb for inanimate nouns. Now you seem to be saying something very different.

No, that's exactly what I just said. But your saying it marks only the object when it clearly doesn't since it also marks the subject/sole of an intransitive if and only if the subject/sole is an inanimate. That is by definition two functions.

I don’t see why that matters…?

Choices (like case terminology) are sometimes made for the sake of informational simplicity and aesthetics. It is easier to say "animates have a nominative and accusative case while inanimates have an absolutive and ergative case" and explain what each case does than to explain 1) why only inanimates take the ergative and animates do not and 2) why inanimates do not take the nominative when they are agent of transitive verbs. It also makes the organization of grammar easier.

like it's easier to say:
animates
one case is S=A
one case is O/P
inanimates
one case is S=O/P
one case is A

instead of being like:
one case is S=A
one case is S=O/P or just O/P
one case is A

There is no better and more unambiguous way to convey how the cases work except to split the accusative into animate accusative and inanimate absolutive. There just isn't.

And also I'll see in some grammars that they'll list an otherwise distinct nominative and accusative as simply "direct" case when a specific noun class or grammatical situation (e.g. construct state) syncretizes them.