Zompist Bboard Again

masako wrote: ↑Sat Sep 26, 2020 6:50 pm Same for me.

Yep, glad to see it! There's been a good bit of conlanging this year. It's fun to see.

masako wrote: ↑Sat Sep 26, 2020 6:50 pm And this thread is far from boring.

Aw, shucks. Thanks!

masako wrote: ↑Sat Sep 26, 2020 6:50 pm You had my curiosity, now you have my attention.
/meme

You had me at "phonology"!!!

The next item up for bids is....

Jin: Adjunct Incorporation!!!

Adjunct incorporation would be incorporation.... of.... adjuncts. Yeah. Basically, it incorporates oblique arguments in a sentence. I've seen at least some polysynthetic languages can incorporate instruments, and even mark them as such with an affix that is suspiciously case-like. In Jin, I'll probably be using this type of incorporation for more than just instrumentals, but in the interest of a short, easy post, I'll stick with that since others would work the same.

General Notes:

• Adjuncts are introduced with the "linker" clitic tied to a verb. See this post for info (same as for genitive incorporation above).
• This construction incorporates both the linked verb and its argument, which makes it different from argument & genitive incorporation
• Word order changes: the linked verb's argument is prefixed to the incorporated linked verb instead of following it
• Because an adjunct is being incorporated, the voice of the main verb isn't an issue since the main arguments are not involved

Non-incorporated	k'alisi qam umbabiwi ja nsatimi alwika ja jambi k'alis-i cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG al-wika of.LNK-use ja 3P.STA.DET jambi knife The merchant cut the cloth with a knife.
Incorporated	jambiwikak'alisi qam umbabiwi ja nsatimi jambi-wika-k'alis-i knife-use-cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant knife-cut the cloth.

The sentence's patient (stative argument) could also be incorporated (per argument incorporation):

jambiwikansatimak'alisanzuni qam umbabiwi jambi-wika-nsatima-k'alis-anzun-i knife-use-cloth-cut-ANP.ACT-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG The merchant knife-cloth-cut.

Yokai Noun Classes

This will be the first I have ever posted about my ideas for this 4th language I have been working on: Yokai. Yeah, I haven't even posted a phonology! (OMG, am I doing it wrong? ) I actually have a phonology mostly worked out, but figuring out noun classes has been job #1 for this language since they will be so central to its identity.

What I have so far is still quite sketchy. However, I finally have something that doesn't just feel like a straight rehash of Bantu noun classes. I took some inspiration from Dyirbal & other Australian langs, as well as Pedant's Proto-Haradrimic (still love those noun classes!). Based on Ares Land's Tarandim and advice, I also looked at some languages with classifiers, such as Kuuk Thaayorre, Diyari, & Ngalakgan.

I had the idea to use noun classes based on the 4 typical elements, but that felt a little too tropey. What I have now - celestial, aqueous, terran, & nether - are still very close to that, but I think have just enough variance to be less artificial (feedback welcome on that, though!). The nether "realm" of classes still has fire, but it incorporates some of Dyrbal's "things to be feared", for example.


Yokai Noun Classes

#	Realm	Description
1	Celestial People	elders, leaders, ancestors, helpful spirits
2	Aqueous People	women
3	Terran People	men
4	Celestial Animals	birds, flying insects
5	Aqueous Animals	fish, amphibians, marine mammals & reptiles
6	Terran Animals	mammals, reptiles
7	Aqueous Plants	fruits, vegetables,seaweed, food (aqueous sources)
8	Terran Plants	trees, bushes, grass, moss, food (celestial, terran, nether sources)
9	Celestial Objects & Materials Aqueous Objects & Material Terran Objects & Material Nether Objects & Material	clouds, moon, sun, stars, air, smoke, light liquids, pastes, shells, corals tools, stone, metal, wood weapons, coffins, darkness, fire
10	Celestial Places Aqueous Places Terran Places Nether Places	sky, heaven rivers, lakes, water mountains, hills, forests, valleys caves, graves
11	Celestial Abstractions	ideas, concepts
12	Aqueous Abstractions	states, qualities
13	Terran Abstractions	actions, events, times
14	Nether Abstractions	death, fear, disease
15	Nether Beings (people, animals, plants)	untouchables, maleficient spirits, crawling insects, bats, mushrooms, mold

Notes:

• Most of these would have 1 affix per class for nouns. I'm thinking the objects & materials class might have several different affixes. One might be used across all of the "realms", while others might be used for just 1 or 2.
• I don't know how much I will actually mention the "realms" as I develop the language. It might still feel a bit artificial to do so. If they aren't specifically stated, the themes should still come through, but it wouldn't feel as forced.
• The descriptions are still fairly preliminary, but hopefully they give a good image of what I'm going after for each one.

Looks good so far! Some comments/questions:

• Is the ‘people/animals/plants/objects/places/abstractions’ distinction marked morphologically? If so, what is the nature of that marking?
• Terminologically, I’d prefer terms which are less ‘poetic’ and more ‘academic’. You could use ‘heavenly’, ‘oceanic’, ‘earthly’, ‘danger/maleficient’, perhaps? Your categories are abstract enough that you could probably even get away with ‘Classes I, II, III, IV’, if you want. (An unimportant point, I know, but it’s bugging me )
• This point is one I’m a little uncertain about, but some parts of your classification strike me as somewhat Eurocentric. (Well, Middle-Eastern-centric, really.) For instance, how common exactly is it to associate heaven with good things and authority, while underground is associated with dangerous things? I can easily think of alternate assignments; for instance, a culture might associate the sky with wind, storm and lightning, while conceiving underground as the source of nourishment. (Note that that’s not a rhetorical question; I’m genuinely interested in knowing this!)

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• Is the ‘people/animals/plants/objects/places/abstractions’ distinction marked morphologically? If so, what is the nature of that marking?

Not specifically, no. Or, it's not separate from the celestial/aqueous/terran/nether distinction. Each of the 15 classes I listed would have it's own pair of singular & plural affixes. Class 9 - the objects/materials - will probably have multiple affixes.

The language is going to be entirely or very close to entirely suffixing, so the noun class affixes will feel different than Bantu classes. I'm thinking about making the plural affixes for each class be a circumfix. That would be another nod to Pedant's Proto-Haradrimic. That would be the only non-suffixes, unless I toss just a few prefixes into the derivation mix.

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• Terminologically, I’d prefer terms which are less ‘poetic’ and more ‘academic’. You could use ‘heavenly’, ‘oceanic’, ‘earthly’, ‘danger/maleficient’, perhaps? Your categories are abstract enough that you could probably even get away with ‘Classes I, II, III, IV’, if you want. (An unimportant point, I know, but it’s bugging me )

Oh, you traditional purist!

I'm quite sure the names will change. I came up with this by stepping back from a simple list, which turned out too Bantu-ish. This is still close, but has at least some difference. I created a table with people/animals/plants/objects/places/abstractions on one axis and celestial/aqueous/terran/nether on the other. (Those were revised several times as well.) Once I had the table filled it, I changed it back to the list that you see here. This is why you're seeing "celestial people", etc.

Speaking of that, I'm toying with the idea of having differences in how people in the different classes utilize the other classes. For example, women might use different affixes or just 1 affix for everything in the nether "realm", while men might do the same for the celestial realm. There might be syntactic differences as well, or even different vocabulary, so this would be avoidance speech built into the classes. I haven't worked on those ideas much beyond what I just described. That might make the celestial/aqueous/terran/nether distinction more defined even if the distinction is not marked in the noun classes themselves.

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• This point is one I’m a little uncertain about, but some parts of your classification strike me as somewhat Eurocentric. (Well, Middle-Eastern-centric, really.) For instance, how common exactly is it to associate heaven with good things and authority, while underground is associated with dangerous things? I can easily think of alternate assignments; for instance, a culture might associate the sky with wind, storm and lightning, while conceiving underground as the source of nourishment. (Note that that’s not a rhetorical question; I’m genuinely interested in knowing this!)

This is actually quite helpful and a very good point! I've thought a little bit about the culture behind this language, but not with this crystal-clear type of focus. That's probably why I have had a hard time coming up with noun classes, since I don't have a super clear set of values & beliefs to base it on. It might be time to do a little research for that purpose.

I can say the culture will be influenced by the Pacific Rim. Chinese and Korean were primary inspirations for the phonology. There will definitely be a caste system per India. No idea on specifics, but I'd like it to not be a carbon copy. Holism will inform the philosophies, along with ideas about the world being a living entity itself: basically an "Earth Mother" or similar. They will practice some form of druidic, animistic, nature-based magic as well.

It's not a complete picture, but it at least gives a framework that some well-focused research could build upon.


Thanks for the feedback!

Vardelm wrote: ↑Fri Oct 09, 2020 8:27 am

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• Is the ‘people/animals/plants/objects/places/abstractions’ distinction marked morphologically? If so, what is the nature of that marking?

Not specifically, no. Or, it's not separate from the celestial/aqueous/terran/nether distinction. Each of the 15 classes I listed would have it's own pair of singular & plural affixes. Class 9 - the objects/materials - will probably have multiple affixes.

Thanks for clarifying! But in that case, is there any particular reason to analyse this as a system of 4x6 ‘superclasses’ rather than just as a collection of 15 noun classes?

(Actually, now that I think of it, what would you call those individual composable features? ‘Superclasses’ doesn’t sound right, but neither does anything else…)

The language is going to be entirely or very close to entirely suffixing, so the noun class affixes will feel different than Bantu classes. I'm thinking about making the plural affixes for each class be a circumfix. That would be another nod to Pedant's Proto-Haradrimic. That would be the only non-suffixes, unless I toss just a few prefixes into the derivation mix.

Hmm, that ‘circumfix’ reference gives me an idea: why not make the plural affix a prefix, which cooccurs with the class suffix but itself varies with noun class? You could then merge some of the noun classes in the plural prefix, so the classifying suffix marks all classes but the plural prefix takes on a more limited set of values. This lets you do some interesting things: for instance, it could give you a way to morphologically mark your celestial/aqueous/terran/nether distinction, while keeping the fusional suffixes as is. That is, if your singular noun classes are something like this (making up some random affixes):

Celestial People: -ya
Aqueous People: -gi
Celestial Animals: -de
Aqueous Animals: -no
Celestial Places: -ŋi
Aqueous Places: -za

Then your plurals could be something like this:

Celestial People: yo--ya
Aqueous People: ji--gi
Celestial Animals: yo--de
Aqueous Animals: ji--no
Celestial Places: yo--ŋi
Aqueous Places: ji--za

Or, you could go even further, and merge some classes in the singular and others in the plural. You could even go as far as marking the celestial/aqueous/terran/nether distinction in one affix and the people/animals/plants/objects/places/abstractions distinction in the other, like this:

Class	Singular	Plural
Celestial People	-ya	yo--ya
Aqueous People	-ya	ji--ya
Celestial Animals	-de	yo--de
Aqueous Animals	-de	ji--de
Celestial Animals	-ŋi	yo--ŋi
Aqueous Animals	-ŋi	ji--ŋi

And even this is tame compared to what some languages do… for instance, here are the noun classifier suffixes of Luwo, in singular and plural:


You could consider each line here to be a separate noun class, with its own singular/plural marking… and in fact this is an entirely reasonable position to take (I believe zompist endorses it in the ALC). I’m essentially describing a system such as this, but where the plural markers are prefixes which occur in addition to the singular suffix. So e.g. where Luwo has singular/plural pairs like rìŋ-ɔ̀/rìŋ-∅, gwàŋ-∅/gwáŋ-ɛ́, àdʊ́ʊ́l-ɔ̀/àdʊ́ʊ̀l-ɛ́, I’m describing a system which would have *rìŋ-ɔ̀/∅-rìŋ-ɔ̀, *gwàŋ-∅/ɛ́-gwáŋ-∅, *àdʊ́ʊ́l-ɔ̀/ɛ́-àdʊ́ʊ́l-ɔ̀.

(You don’t have to do any of this, of course; I’m just trying to give some ideas to think about.)

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• Terminologically, I’d prefer terms which are less ‘poetic’ and more ‘academic’. You could use ‘heavenly’, ‘oceanic’, ‘earthly’, ‘danger/maleficient’, perhaps? Your categories are abstract enough that you could probably even get away with ‘Classes I, II, III, IV’, if you want. (An unimportant point, I know, but it’s bugging me )

Oh, you traditional purist!

I'm quite sure the names will change. I came up with this by stepping back from a simple list, which turned out too Bantu-ish. This is still close, but has at least some difference. I created a table with people/animals/plants/objects/places/abstractions on one axis and celestial/aqueous/terran/nether on the other. (Those were revised several times as well.) Once I had the table filled it, I changed it back to the list that you see here. This is why you're seeing "celestial people", etc.

Yes, that’s pretty much what I assumed when I saw that list. Maybe try delineating exactly which words fall into which categories? That might help with getting some better names. (For inspiration, I can suggest Dixon’s Dyirbal grammar; given ‘prototypical’ noun class members, Dyirbal has a set of metaphorical ‘transfer’ processes to assign class membership to non-prototypical nouns.)

Speaking of that, I'm toying with the idea of having differences in how people in the different classes utilize the other classes. For example, women might use different affixes or just 1 affix for everything in the nether "realm", while men might do the same for the celestial realm. There might be syntactic differences as well, or even different vocabulary, so this would be avoidance speech built into the classes. I haven't worked on those ideas much beyond what I just described. That might make the celestial/aqueous/terran/nether distinction more defined even if the distinction is not marked in the noun classes themselves.

Other than zompist’s Barakhinei, I don’t think I’ve seen any other conlang with gender differences before… I’d love to see more of this!

bradrn wrote: ↑Thu Oct 08, 2020 10:23 pm

• This point is one I’m a little uncertain about, but some parts of your classification strike me as somewhat Eurocentric. (Well, Middle-Eastern-centric, really.) For instance, how common exactly is it to associate heaven with good things and authority, while underground is associated with dangerous things? I can easily think of alternate assignments; for instance, a culture might associate the sky with wind, storm and lightning, while conceiving underground as the source of nourishment. (Note that that’s not a rhetorical question; I’m genuinely interested in knowing this!)

This is actually quite helpful and a very good point! I've thought a little bit about the culture behind this language, but not with this crystal-clear type of focus. That's probably why I have had a hard time coming up with noun classes, since I don't have a super clear set of values & beliefs to base it on. It might be time to do a little research for that purpose.

Huh, that’s new… for someone who has almost never gotten to the ‘conculture’ stage of his conlanging, I find it somewhat strange to be complimented on my ‘crystal-clear focus’ on the topic!

(I’m not sure I would be able to help any more with this, though… I know too little about the cultures you cite as inspirations to be able to be of much use. I guess it’s time for me to buy more of zompist’s books then!)

Thanks for the feedback!

As always, you’re very welcome! I’m always happy to give feedback. (Some would say I give too much of it! )

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Thanks for clarifying! But in that case, is there any particular reason to analyse this as a system of 4x6 ‘superclasses’ rather than just as a collection of 15 noun classes?

(Actually, now that I think of it, what would you call those individual composable features? ‘Superclasses’ doesn’t sound right, but neither does anything else…)

I'm not sure there's a reason to do so, no, but see comments below about the affix forms.

The only reason I did it this way was that I needed to sort of reframe things and look at it from a different angle. My intention has been to organize them into a more traditional list, just because that's what you see with natlangs.

The question of what to call them is interesting. Either set could be the "superclass". Maybe both sets are "subclasses", and the combination of them leads to the actual "classes". I'll probably worry about that more when I have to write a grammar.

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Hmm, that ‘circumfix’ reference gives me an idea: why not make the plural affix a prefix, which cooccurs with the class suffix but itself varies with noun class? You could then merge some of the noun classes in the plural prefix, so the classifying suffix marks all classes but the plural prefix takes on a more limited set of values. This lets you do some interesting things: for instance, it could give you a way to morphologically mark your celestial/aqueous/terran/nether distinction, while keeping the fusional suffixes as is. That is, if your singular noun classes are something like this (making up some random affixes):

...

(You don’t have to do any of this, of course; I’m just trying to give some ideas to think about.)

Pedant did this a little bit with his class affixes, but the plurals were suffixes. So yeah, something in this regard is at least on my radar. What you list here is all really good variations to consider, so that will be quite useful.

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Yes, that’s pretty much what I assumed when I saw that list. Maybe try delineating exactly which words fall into which categories? That might help with getting some better names. (For inspiration, I can suggest Dixon’s Dyirbal grammar; given ‘prototypical’ noun class members, Dyirbal has a set of metaphorical ‘transfer’ processes to assign class membership to non-prototypical nouns.)

This confuses me a little bit since I do have example (English) words in each category. Granted, it's sketchy, but it's there. More example words would of course maybe indicate names, but I have a fairly firm idea of what each category means. The Dyirbal transfer thing sounds interesting, though!

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Other than zompist’s Barakhinei, I don’t think I’ve seen any other conlang with gender differences before… I’d love to see more of this!

Happy to accommodate.... eventually! I think this may be a primarily matriarchal society, with maybe 60-70% of it's leadership being female.

BTW, I think there would be additional registers as well: there will probably be an "untouchable" caste that uses avoidance speech with regards to the celestial classes, and maybe the aqueous classes as well.

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Huh, that’s new… for someone who has almost never gotten to the ‘conculture’ stage of his conlanging, I find it somewhat strange to be complimented on my ‘crystal-clear focus’ on the topic!

(I’m not sure I would be able to help any more with this, though… I know too little about the cultures you cite as inspirations to be able to be of much use. I guess it’s time for me to buy more of zompist’s books then!)

Let me rephrase that: at the very least, it brings the issue into crystal-clear focus for me, even if it isn't clear for you.

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am As always, you’re very welcome! I’m always happy to give feedback. (Some would say I give too much of it! )

Nope, it's all appreciated!

Vardelm wrote: ↑Fri Oct 09, 2020 2:37 pm Maybe both sets are "subclasses", and the combination of them leads to the actual "classes".

Interestingly, this was the terminology I originally used in my post, but then when I re-read it I changed it to ‘superclass’ since I thought that seemed more sensible.

bradrn wrote: ↑Fri Oct 09, 2020 9:20 am Yes, that’s pretty much what I assumed when I saw that list. Maybe try delineating exactly which words fall into which categories? That might help with getting some better names. (For inspiration, I can suggest Dixon’s Dyirbal grammar; given ‘prototypical’ noun class members, Dyirbal has a set of metaphorical ‘transfer’ processes to assign class membership to non-prototypical nouns.)

This confuses me a little bit since I do have example (English) words in each category. Granted, it's sketchy, but it's there. More example words would of course maybe indicate names, but I have a fairly firm idea of what each category means.

Ah, sorry, they just sounded very vague from your brief distinctions.

bradrn wrote: ↑Fri Oct 09, 2020 7:27 pm Ah, sorry, they just sounded very vague from your brief distinctions.

Quite all right; no apology necessary! This tells me I definitely need to elaborate more in future posts as I iterate. Of course, there's a balance to the presentation of the ideas since too much info can turn people away, too.

Yokai Phonology

Consonants: IPA

	Labial	Coronal	Palatal	Dorsal	Gutteral
Stops	b p pʰ	d t tʰ		g k kʰ
Affricates		d͡z t͡s t͡sʰ	d͡ʑ t͡ɕ t͡ɕʰ
Fricatives	f	z s	ʑ, ɕ	x	h
Nasals	m	n		ŋ
Continuants		ɾ l	j	w

Consonants: Orthography

	Labial	Coronal	Palatal	Dorsal	Gutteral
Stops	b p ph	d t th		g k kh
Affricates		dz ts tsh	j c ch
Fricatives	f	z s	zh, sh	x	h
Nasals	m	n		ng
Continuants		r l	y	w

Vowels

	Front	Mid	Back
Closed	i		u
Mid	e		o
Open		a

diphthongs

• ie ia io iu
• ei eo
• ai ao
• ui ue ua uo

tripthongs

• iao
• uai

EDIT: I'm considering adding nasalization to the vowel system as well, but not sure to what extent. Maybe nasalized versions of all the vowels here, or perhaps just those that have <a u o>.


Syllable Structure

• CGVN
  • C = any consonant except <ng>
  • G = < y w >
  • N = < m n l ng >

Onset		Nucleus Vowel
C	+	i e a o u
Cw	+	i ie ia io iu iao
CG	+	e a o ei eo ai ao
Cy	+	u ui ue ua uo uai

Stress

• penultimate

Jin: Adjunct Antecedent Incorporation

This type of incorporation is similar to the type #4 described by Mithun. However, it's not used for primary arguments of a verb; it's just for highlighting or emphasizing adjuncts, which appear after the main arguments.


General Notes:

• This type works almost identically to the "adjunct incorporation" shown before
• The adjunct itself is not incorporated; instead, a more generic noun that is a class or group the adjunct belongs to is incorporated.
• The same verb used in the adjunct is incorporated as well
• I think there will be some kind of pragmatics in play for whether this construction is used, which I haven't specifically worked out yet. I see it as wanting to emphasize the adjunct to some extent, but the agent & patient also are in their non-incorporated state. I will probably write a post comparing the various incorporations after I describe the 5th & final type.

lit’udawikamimutaqi qam ndaluwi ja untaqi alja umbabiwi alwika ja jambi lit’uda-wika-mimut-aq-i weapon-use-dead-DIR.ACT-SNG qam 3P.ACT.DET ndalu-wi man-the.SNG ja 3P.STA.DET untaq-i camel-the.SNG al-ja of.LNK-3P.STA.DET umbabi-wi merchant-the.SNG al-wika of.LNK-use ja 3P.STA.DET jambi knife The man weapon-killed the merchant’s camel with a knife.

Jin: Adverbs, Serial Verbs, & Incorporation

To write about this final type of incorporation, I had to first figure out how I was going to handle adverbs! As a result, I'll cover that a little bit here as well, although it's not a comprehensive treatment of the topic. It's just comprehensive for what I've come up with on it so far.

Starting Point:
Take the following sentence from one of the previous posts:

k'alisi qam umbabiwi ja nsatimi k'alis-i cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant cut the cloth.

Main verb that can be (obviously) adverbialized:
Now we have the stative, intransitive verb miba "quick" appearing as the predicate of a sentence.

miba ja umbabiwi miba quick ja 3P.STA.DET umbabi-wi merchant-the.SNG The merchant is quick.

Coordinated sentence with the linker al-
As seen in previous posts, the linker al- can be used to form adjunct phrases, which are really coordinated sentences. In this example, almiba doesn't show the subject since it is the same as the sentence with k'alisi. I need to do some work on this to nail down the rules for coordinated sentences & omitting subjects, but I think Jin is syntactically accusative.

k'alisi qam umbabiwi ja nsatimi almiba k'alis-i cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG al-miba of.LNK-quick The merchant cut the cloth and was quick.

Adverb derivation:
Adverbs can be derived from content words by the suffix -(n)un. (This might change to a prefix later since I want Jin to be primarily prefixing.) Adverbs primarily appear at the start or end of a sentence.

mibanun k'alisi qam umbabiwi ja nsatimi mibanun quickly k'alis-i cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG Quickly, the merchant cut the cloth.

k'alisi qam umbabiwi ja nsatimi mibanun k'alis-i cut-SNG qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG mibanun quickly The merchant cut the cloth quickly.

Adverb interjection:
Adverbs can also appear between a verb and its arguments as an interjection. This is less common than starting or ending a sentence.

k'alisi mibanun qam umbabiwi ja nsatimi k'alis-i cut-SNG mibanun quickly qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant - quickly - cut the cloth.

Serial verbs:
Verbs can appear as serial verbs. They appear right after each other at the start of the sentence; otherwise, they need to be connected with the linker al-.

The first verb in the serialization determines the required arguments for the sentence. Any successive verbs must refer to the same subject and (if transitive) to the same object. I'm not sure if successive verbs will be allowed to take only the 1st verb's object as an argument.

k'alisi miba qam umbabiwi ja nsatimi k'alis-i cut-SNG miba quick qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant was quick to cut the cloth.

Adverb (Serial verb / Adjunct) Incorporation
Finally, with all of that in place, we can incorporate adverbs into the main verb. Now that I'm writing this, I can see that it's pretty much the same as adjunct incorporation. It's just that here there is no object (as in "with a knife") being incorporated along with the serial/adjunct verb.

"Adverbs" are incorporated, but not in their derived adverbial form. With that said, I think there will be some adverbs that are not derived with the suffix -nun. Those can probably be incorporated as-is since they aren't derived from other content words. There will probably be some rules about which adverbs can incorporate or not.

mibak'alisi qam umbabiwi ja nsatimi miba-k'alis-i quick-cut-SNG miba quick qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant quick-cut the cloth.

Vardelm wrote: ↑Fri Oct 16, 2020 11:51 pm Serial verbs:
Verbs can appear as serial verbs. They appear right after each other at the start of the sentence; otherwise, they need to be connected with the linker al-.

The first verb in the serialization determines the required arguments for the sentence. Any successive verbs must refer to the same subject and (if transitive) to the same object. I'm not sure if successive verbs will be allowed to take only the 1st verb's object as an argument.

k'alisi mibanun qam umbabiwi ja nsatimi k'alis-i cut-SNG miba quickly qam 3P.ACT.DET umbabi-wi merchant-the.SNG ja 3P.STA.DET nsatim-i cloth-the.SNG The merchant was quick cut the cloth.

Quick terminological note: Usually, serial verbs are defined as sequences of verbs which (amongst other properties) cannot have a morphological marker of dependency inserted between them, so by most definitions these aren’t really serial verbs.

And, as usual, I have a question: are these ‘serial verbs’ used only for adverbial meanings, or do they have other uses as well?

(Oh, and a possible typo: was the translation supposed to be ‘The merchant cut the cloth quickly’?)

bradrn wrote: ↑Sat Oct 17, 2020 7:05 am
Quick terminological note: Usually, serial verbs are defined as sequences of verbs which (amongst other properties) cannot have a morphological marker of dependency inserted between them, so by most definitions these aren’t really serial verbs.

Are you referring to the -i in k'alisi? If so, that's not a marker of dependency, but it's still a fair point since serial verbs should have the same TAM markings (although this isn't even technically TAM!).

This is going to be interesting to work out due to the amount of inflection Jin will have. Since the -i singulative marker codes a verb to be episodic, both verbs in a serial construction should probably have that. Also, the tense marking that I haven't talked about yet will need to be the same. There's a slew of inflectional prefixes, which I might only mark on the 1st verb in the chain, with the understanding that it affects both. Voice is partly derivational, so that will be allowed to be different, but with restrictions. I don't think you'll be able to combine a direct voice with a passive or inverse voice, for example. It remains to be worked out, of course.

bradrn wrote: ↑Sat Oct 17, 2020 7:05 am And, as usual, I have a question: are these ‘serial verbs’ used only for adverbial meanings, or do they have other uses as well?

Answer: yes, but I haven't thought through it much. I thought that coordinated sentences would use a particle that was different than the clitic al-, but realized al- did what I needed. I love these moments in conlanging when you "discover" something about your language!

I think they will be used for verbs that are parts of a whole, connected event. That could be simultaneous actions or where one is a direct result of the other. My understanding is that - in some languages at least - serial verbs can have different objects. That's not the case here, so the verb will have to be very closely connected.

bradrn wrote: ↑Sat Oct 17, 2020 7:05 am (Oh, and a possible typo: was the translation supposed to be ‘The merchant cut the cloth quickly’?)

It was almost 1:00 AM here, so the brain was starting to shut down. Yeah, that was a typo of omission, which I just corrected. I changed it to "The merchant was quick to cut the cloth", but I also considered "The merchant was quick and cut the cloth". Either work. Actually, all of these examples could easily be translated as "The merchant cut the cloth quickly", but I want to show a difference between them in English where possible.

EDIT: I just noticed & fixed a couple other problems as well. Nothing to see here: move along, move along!

Vardelm wrote: ↑Sat Oct 17, 2020 8:31 am

bradrn wrote: ↑Sat Oct 17, 2020 7:05 am
Quick terminological note: Usually, serial verbs are defined as sequences of verbs which (amongst other properties) cannot have a morphological marker of dependency inserted between them, so by most definitions these aren’t really serial verbs.

Are you referring to the -i in k'alisi? If so, that's not a marker of dependency, but it's still a fair point since serial verbs should have the same TAM markings (although this isn't even technically TAM!).

No, I’m talking about al-. The idea is that it should be impossible to insert a linking marker, which is what al- seems to be here. (A natlang example: English sentences like ‘I’ll go do it now’ look a lot like SVCs, but aren’t analysed that way due to the fact that you can insert ‘and’ with no change in meaning.) I believe constructions like yours are most commonly called ‘converbs’, though I’m not entirely sure that’s correct.

This is going to be interesting to work out due to the amount of inflection Jin will have. Since the -i singulative marker codes a verb to be episodic, both verbs in a serial construction should probably have that. Also, the tense marking that I haven't talked about yet will need to be the same. There's a slew of inflectional prefixes, which I might only mark on the 1st verb in the chain, with the understanding that it affects both. Voice is partly derivational, so that will be allowed to be different, but with restrictions. I don't think you'll be able to combine a direct voice with a passive or inverse voice, for example. It remains to be worked out, of course.

Yep, this all sounds pretty normal for SVCs. (Your system reminds me of Skou in particular, which has similar differences in concord between its various inflectional categories.)

bradrn wrote: ↑Sat Oct 17, 2020 7:05 am And, as usual, I have a question: are these ‘serial verbs’ used only for adverbial meanings, or do they have other uses as well?

Answer: yes, but I haven't thought through it much. … I think they will be used for verbs that are parts of a whole, connected event. That could be simultaneous actions or where one is a direct result of the other.

‘Direct result’ is very common in SVCs. Usually it’s accomplished through switching subjects, along the lines of ‘He hit die it = He hit it; it die’. ‘Simultaneous actions’ seems a bit less common, though still attested. ‘Whole, connected event’ is a very common constraint on SVCs — often the event is restricted to a culturally common one.

If you’re looking for more inspiration, here’s some other common categories:

• Motion preceding an event: ‘I go do it’
• Directionals: ‘I walk come = I walk to here’, ‘I take come it = I bring it here’
• Orientation: ‘I look rise = I look upwards’
• Case-marking: ‘I take knife cut bread = I cut the bread with a knife’, ‘I help you cook = I cook for you’
• TAM: ‘I finish eat = I’ve eaten’, ‘I begin walk = I start walking’

(Please don’t ask me how I know all this… I’ve spent far longer than is sane or healthy reading about this stuff, and now I feel duty-bound to inflict it upon other people )

My understanding is that - in some languages at least - serial verbs can have different objects. That's not the case here, so the verb will have to be very closely connected.

Syntactically, this is attested e.g. in Oceanic. Semantically, not so much: verbs very commonly have different objects even when objects are shared syntactically. This is particularly common in resultatives such as the aforementioned ‘He hit die it’ — ‘hit’ has an object, which isn’t shared with ‘die’.

bradrn wrote: ↑Sat Oct 17, 2020 9:03 am No, I’m talking about al-. The idea is that it should be impossible to insert a linking marker, which is what al- seems to be here. (A natlang example: English sentences like ‘I’ll go do it now’ look a lot like SVCs, but aren’t analysed that way due to the fact that you can insert ‘and’ with no change in meaning.) I believe constructions like yours are most commonly called ‘converbs’, though I’m not entirely sure that’s correct.

Ah, OK! Those are definitely not serial verbs. I see them being used for verbs where the subject and/or object are different than the main verb. This makes adjunct phrases (alwika jambi "using/with a knife") one of the primary uses.

The only example I meant to be a serial verb was k'alisi miba "cut quick". I may have confused the issue since one of my other typos was there. I originally translated miba as "quickly", but to be consistent it should just be the stative verb "quick". I intend that serial verbs in Jin appear at the start of a sentence (or clause) and are not separated by any other words. The examples using al- are what I was considering "coordinated sentences". Using "converbs" might be a good way to describe them as well. I'll look into that.

bradrn wrote: ↑Sat Oct 17, 2020 9:03 am Yep, this all sounds pretty normal for SVCs. (Your system reminds me of Skou in particular, which has similar differences in concord between its various inflectional categories.)

‘Direct result’ is very common in SVCs. Usually it’s accomplished through switching subjects, along the lines of ‘He hit die it = He hit it; it die’. ‘Simultaneous actions’ seems a bit less common, though still attested. ‘Whole, connected event’ is a very common constraint on SVCs — often the event is restricted to a culturally common one.

If you’re looking for more inspiration, here’s some other common categories:

I can always count on your research! I'll take at these, too, out of curiosity. That list will be helpful to wrap my head around types of serial constructions.

bradrn wrote: ↑Sat Oct 17, 2020 9:03 am (Please don’t ask me how I know all this… I’ve spent far longer than is sane or healthy reading about this stuff, and now I feel duty-bound to inflict it upon other people )

Hey, watch where you're swinging that research, mister!

Vardelm wrote: ↑Sat Oct 17, 2020 9:21 am

bradrn wrote: ↑Sat Oct 17, 2020 9:03 am No, I’m talking about al-. The idea is that it should be impossible to insert a linking marker, which is what al- seems to be here. (A natlang example: English sentences like ‘I’ll go do it now’ look a lot like SVCs, but aren’t analysed that way due to the fact that you can insert ‘and’ with no change in meaning.) I believe constructions like yours are most commonly called ‘converbs’, though I’m not entirely sure that’s correct.

Ah, OK! Those are definitely not serial verbs. I see them being used for verbs where the subject and/or object are different than the main verb. This makes adjunct phrases (alwika jambi "using/with a knife") one of the primary uses.

The only example I meant to be a serial verb was k'alisi miba "cut quick". I may have confused the issue since one of my other typos was there. I originally translated miba as "quickly", but to be consistent it should just be the stative verb "quick". I intend that serial verbs in Jin appear at the start of a sentence (or clause) and are not separated by any other words. The examples using al- are what I was considering "coordinated sentences".

Wait, I’m confused now… can you insert al- into the phrase kʼalisi miba, or is that impossible? From your post, it sounded like al- is required when the phrase doesn’t appear at the beginning of the sentence, but maybe I’m misunderstanding something…

bradrn wrote: ↑Sat Oct 17, 2020 9:03 am Yep, this all sounds pretty normal for SVCs. (Your system reminds me of Skou in particular, which has similar differences in concord between its various inflectional categories.)

‘Direct result’ is very common in SVCs. Usually it’s accomplished through switching subjects, along the lines of ‘He hit die it = He hit it; it die’. ‘Simultaneous actions’ seems a bit less common, though still attested. ‘Whole, connected event’ is a very common constraint on SVCs — often the event is restricted to a culturally common one.

If you’re looking for more inspiration, here’s some other common categories:

I'll take at these, too, out of curiosity. That list will be helpful to wrap my head around types of serial constructions.

You actually want to investigate this stuff‽ OK then. (Though don’t say I didn’t warn you…) I recommend starting with Dixon and Aikhenvald’s Serial Verb Constructions: A Cross-linguistic Typology, if you can find it. It’s a bit long, but you really only need the first chapter, which summarises the basics of SVCs. You’ll probably be interested in Chapter 15 as well — it gives an overview of Wolaitta, which looks similar to Jin in that it also uses ‘converbs’ with similarities to SVCs.

bradrn wrote: ↑Sun Oct 18, 2020 1:00 am Wait, I’m confused now… can you insert al- into the phrase kʼalisi miba, or is that impossible? From your post, it sounded like al- is required when the phrase doesn’t appear at the beginning of the sentence, but maybe I’m misunderstanding something…

I don't think I explained this particularly well.

I stated that:

Vardelm wrote: ↑Fri Oct 16, 2020 11:51 pm Verbs can appear as serial verbs. They appear right after each other at the start of the sentence; otherwise, they need to be connected with the linker al-.

When I say "at the start of a sentence", I mean the entire cluster of verbs that make up the serial verb construction. Instead of saying "at the start of a sentence", maybe that should be "before the arguments". I haven't directly & consciously thought about this, but I think as soon as you introduce the arguments - using any pronoun like qam, ja, qasim, jida, etc. - you then need to use a the linker to "tie" a new phrase back to the main verb.

So, this is the "start" of a sentence:

k'alisi miba qam umbabiwi ja nsatimi

While this is not:

k'alisi qam umbabiwi ja nsatimi almiba

Even when I say "start of the sentence", that's not entirely accurate, since I allowed for adverbs to appear before the main verb, as per:

mibanun k'alisi qam umbabiwi ja nsatimi

So, to me, there's some "fuzziness" in how to explain it. The language is VSO, so do I just say "co-occurring in the main verbal slot", or just "pre-argument", or both, or something else....???? Maybe I should explain the sentence structure at a slightly higher level than VSO:

Fronted adverb

Main verb and/or SVC

Interjected adverb

Arguments (pronouns & their dependent phrases)

Adverbial/adjunct phrases

Does any of that make things more clear?

bradrn wrote: ↑Sun Oct 18, 2020 1:00 am You actually want to investigate this stuff‽ OK then. (Though don’t say I didn’t warn you…) I recommend starting with Dixon and Aikhenvald’s Serial Verb Constructions: A Cross-linguistic Typology, if you can find it. It’s a bit long, but you really only need the first chapter, which summarises the basics of SVCs. You’ll probably be interested in Chapter 15 as well — it gives an overview of Wolaitta, which looks similar to Jin in that it also uses ‘converbs’ with similarities to SVCs.

HAH!

OK, yes, I'm interested to investigate, but only to a certain point! Or, rather, I'd love to dive into the rabbit hole but full-time work + grad school have a say in the matter! Still, the recommendation of a specific book w/ specific chapters is super helpful in making what investigation I do efficient, so thank you!

Vardelm wrote: ↑Sun Oct 18, 2020 9:21 am

bradrn wrote: ↑Sun Oct 18, 2020 1:00 am Wait, I’m confused now… can you insert al- into the phrase kʼalisi miba, or is that impossible? From your post, it sounded like al- is required when the phrase doesn’t appear at the beginning of the sentence, but maybe I’m misunderstanding something…

I don't think I explained this particularly well.

…

Fronted adverb

Main verb and/or SVC

Interjected adverb

Arguments (pronouns & their dependent phrases)

Adverbial/adjunct phrases

Does any of that make things more clear?

OK, that’s a lot clearer now, thanks! (Turns out I’d completely misinterpreted what you were saying: somehow I managed to conclude that you could move the entire verb complex out of sentence-initial position, in which case you need to add al-. Which of course makes no sense!)

In that case… hmm, I’m not quite sure how to analyse this one, actually. There’s no particularly hard boundary between SVCs and related constructions, so this sort of thing certainly feels like a realistic possibility, but I haven’t ever seen it before. I’d say you could get away with calling this either an ‘SVC’ or a ‘converb’, though to me there’s something distinctly un-SVC-like about the ability to turn the second verb into an adverbial.

bradrn wrote: ↑Sun Oct 18, 2020 9:39 am OK, that’s a lot clearer now, thanks!

Excellent!

bradrn wrote: ↑Sun Oct 18, 2020 9:39 am I’d say you could get away with calling this either an ‘SVC’ or a ‘converb’...

Yeah, either would probably work. I'll just need to read up a little more on both and decide which term will describe the situation best. Right now, I lean towards calling the grouped verb that come before the arguments "SVC", and any that come after the arguments (using al-) would be "converbs".

bradrn wrote: ↑Sun Oct 18, 2020 9:39 am there’s something distinctly un-SVC-like about the ability to turn the second verb into an adverbial.

I'm not sure what you mean here. Could you explain?