Random ideas: Morphosyntax

[Creyeditor]

That sounds like a good idea.

[TBPO]

I have an idea for a visual conlang, where sentences are images. Do it make sense?

[Arayaz]

I have an idea for a visual conlang, where sentences are images. Do it make sense?

I believe we call this "writing."

Or do you mean that one would draw the action described by the sentence?

[TBPO]

I have an idea for a visual conlang, where sentences are images. Do it make sense?

I believe we call this "writing."

Or do you mean that one would draw the action described by the sentence?

I mean that every image is a message in this conlang. What matters more is the color, and its relationship to colors elsewhere in the image, rather than a predefined set of symbols. This allows the images to be messages in that language.

[Arayaz]

I have an idea for a visual conlang, where sentences are images. Do it make sense?

I believe we call this "writing."

Or do you mean that one would draw the action described by the sentence?

I mean that every image is a message in this conlang. What matters more is the color, and its relationship to colors elsewhere in the image, rather than a predefined set of symbols. This allows the images to be messages in that language.

Ohhhh, that's cool ─ after making that post, I saw your other post where you mentioned the idea. I made a language conveyed through color as well, but it would be a mammoth effort to translate a whole image from it into English.

[Creyeditor]

How about crossing marked nominative alignment with split ergativity? Let's say the split is based on animacy. Animate subjects and inanimate objects of transitive verbs would be marked with the direct (?) case.

Wece-sung tuba-sung diku.
woman-DIR stone-DIR admire.
The woman admires the stone.

Both inanimate subjects and animate objects of transitive verbs would appear unmarked even though they have different zero-marked cases.

Tuba wece nake.
stone[ERG] woman[ACC] hit.
The stone hit the woman.

Of course, you can also have situations with one marked case if the arguments of the verb have the same animacy.

Tuba nata-sung nake.
stone[ERG] ground-DIR hit.
The stone hit the ground.

Kila-sung wece diku.
man-DIR woman[ACC] admire
The man admires the woman.

Intranstive verbs would also take the direct case.

Tuba-sung tuja.
stone-DIR fall.
The stone falls.

This combines two existing patterns but the result looks much less naturalistic to me.

[Omzinesý]

How about crossing marked nominative alignment with split ergativity? Let's say the split is based on animacy. Animate subjects and inanimate objects of transitive verbs would be marked with the direct (?) case.

Wece-sung tuba-sung diku.
woman-DIR stone-DIR admire.
The woman admires the stone.

Both inanimate subjects and animate objects of transitive verbs would appear unmarked even though they have different zero-marked cases.

Tuba wece nake.
stone[ERG] woman[ACC] hit.
The stone hit the woman.

Of course, you can also have situations with one marked case if the arguments of the verb have the same animacy.

Tuba nata-sung nake.
stone[ERG] ground-DIR hit.
The stone hit the ground.

Kila-sung wece diku.
man-DIR woman[ACC] admire
The man admires the woman.

Intranstive verbs would also take the direct case.

Tuba-sung tuja.
stone-DIR fall.
The stone falls.

This combines two existing patterns but the result looks much less naturalistic to me.

It could look more natural if the nominative marker had less phonemic material, say one vowel.

[Flavia]

How about crossing marked nominative alignment with split ergativity? Let's say the split is based on animacy. Animate subjects and inanimate objects of transitive verbs would be marked with the direct (?) case.

Wece-sung tuba-sung diku.
woman-DIR stone-DIR admire.
The woman admires the stone.

Both inanimate subjects and animate objects of transitive verbs would appear unmarked even though they have different zero-marked cases.

Tuba wece nake.
stone[ERG] woman[ACC] hit.
The stone hit the woman.

Of course, you can also have situations with one marked case if the arguments of the verb have the same animacy.

Tuba nata-sung nake.
stone[ERG] ground-DIR hit.
The stone hit the ground.

Kila-sung wece diku.
man-DIR woman[ACC] admire
The man admires the woman.

Intranstive verbs would also take the direct case.

Tuba-sung tuja.
stone-DIR fall.
The stone falls.

This combines two existing patterns but the result looks much less naturalistic to me.

Interesting. Why is DIR the marked form though? Shouldn't animates be prototypical agents, likewise for inanimates and patients?
edit: Oh, I see, marked nominative alignment was an inspiration. Sorry for that.

[Creyeditor]

How about crossing marked nominative alignment with split ergativity? Let's say the split is based on animacy. Animate subjects and inanimate objects of transitive verbs would be marked with the direct (?) case.

Wece-sung tuba-sung diku.
woman-DIR stone-DIR admire.
The woman admires the stone.

Both inanimate subjects and animate objects of transitive verbs would appear unmarked even though they have different zero-marked cases.

Tuba wece nake.
stone[ERG] woman[ACC] hit.
The stone hit the woman.

Of course, you can also have situations with one marked case if the arguments of the verb have the same animacy.

Tuba nata-sung nake.
stone[ERG] ground-DIR hit.
The stone hit the ground.

Kila-sung wece diku.
man-DIR woman[ACC] admire
The man admires the woman.

Intranstive verbs would also take the direct case.

Tuba-sung tuja.
stone-DIR fall.
The stone falls.

This combines two existing patterns but the result looks much less naturalistic to me.

It could look more natural if the nominative marker had less phonemic material, say one vowel.

True!

[Pabappa]

Tuba wece nake.
stone[ERG] woman[ACC] hit.
The stone hit the woman.

Well, this is interesting, I'll say that. But it seems that inanimate objects are agents by default while humans and other animates are patients. i've explored similar ideas, such as having noun classes that are higher on the animacy hierarchy than humans, and for an entirely separate reason, in my main project (Play), the 2nd person is forced into the patient role by default in a way similar to what I see here ... a 2nd person agent must take a marked form, even if the patient is an inanimate object. The only difference is that in Play this marker is on the verb. I consider this to be a flaw in Play, but the morphology is so "tight" that it's simply not possible for me to change it ... so I've decided to make Play the language of verbal aggression, where the listener is quite often the receiving partner of some action, even in cases where other languages would word the whole sentence differently.

If you decide to go with this idea, I'd be interested to see if you'd create metaphors with inanimate objects acting on human beings to express concepts that we'd handle differently in English, since the resulting sentences would be shorter by perhaps two syllables. To start with a simple idea, would a sentence like "I saw the trees" in this language be better expressed by the equivalent of "the trees showed (to) me"?

[Creyeditor]

If you decide to go with this idea, I'd be interested to see if you'd create metaphors with inanimate objects acting on human beings to express concepts that we'd handle differently in English, since the resulting sentences would be shorter by perhaps two syllables. To start with a simple idea, would a sentence like "I saw the trees" in this language be better expressed by the equivalent of "the trees showed (to) me"?

Interesting idea(s). Would you expect something similar/analogous in marked nominative languages? So instead of using intransitives sentences using a transitive sentence with agent drop? Maybe instead of 'I fall' it would be '(It) causes.to.fall me' or instead of 'I fall asleep' you could say '(it) causes.to.sleep me'. Sounds somewhat plausible. I think some languages do something similar with locative verbs. So instead of 'The stick is leaning against the tree.' it is '(Somebody) caused.to.lean the stick against the tree.'.

[lurker]

I have an idea for a visual conlang, where sentences are images. Do it make sense?

Like this?