Conlang Random Thread

[k1234567890y]

Another factor for fusional languages, I think, is phonological change, which can make originally separated affixes fuse together, becoming a single affix, but this is my own opinion, I can be wrong.

[mèþru]

phonological change, which can make originally separated affixes fuse together, becoming a single affix

I think that that plus levelling and expanding by analogy are how most fusional paradigms formed.

[k1234567890y]

phonological change, which can make originally separated affixes fuse together, becoming a single affix

I think that that plus levelling and expanding by analogy are how most fusional paradigms formed.

ok (:

you are right, these happen, the cause can be phonological(i.e. merge or elision of sounds), morphological, semantic(i.e. the reclassification of nouns, the word for "sun" have different genders across Indo-European languages), or mixed.

when I was looking at Proto-Germanic and Old Saxon nouns on Wiktionary, the paradigms of some words don't match, the ending of some nouns are different from what is expected, it could be an evidence of this.

[Xwtek]

Sorry for misogyny, but is it possible to derive diminutive into feminime gender? Proto Garaudawa initially only has 2 genders, and it is reflected in Rkouic and converted into Proximate-Absolutive in Asent'oan, but split into four genders in the rest of language, taking original diminutive. (No longer productive in both Rkou and Asent'o) The resulting gender is Masculine, Feminime I, Neuter, and Feminime II. Later, Feminime II collapses either into Feminime I or to Neuter.

[Pedant]

Sorry for misogyny, but is it possible to derive diminutive into feminime gender? Proto Garaudawa initially only has 2 genders, and it is reflected in Rkouic and converted into Proximate-Absolutive in Asent'oan, but split into four genders in the rest of language, taking original diminutive. (No longer productive in both Rkou and Asent'o) The resulting gender is Masculine, Feminime I, Neuter, and Feminime II. Later, Feminime II collapses either into Feminime I or to Neuter.

Let’s put it this way: you yourself may not be misogynistic. Your con-people are allowed to be, so long as you acknowledge it. But it’s normally better to have some other factor coming into play for the feminine form to become a diminutive, maybe a similarity (after sound change) to some affix that functions as the diminutive. Or maybe the Feminine II form is also used for children, and then later becomes associated with the smaller form of anything (while the Feminine I becomes the major, well, feminine form).

[linguistcat]

... But it’s normally better to have some other factor coming into play for the feminine form to become a diminutive, ...

Akangka was saying the reverse, diminutive > feminine. I'm assuming what is meant is that the language started with no masculine-feminine distinction grammatically, but did have a diminutive marker of some sort, that then became feminine while unmarked nouns defaulted to masculine.

[Pedant]

... But it’s normally better to have some other factor coming into play for the feminine form to become a diminutive, ...

Akangka was saying the reverse, diminutive > feminine. I'm assuming what is meant is that the language started with no masculine-feminine distinction grammatically, but did have a diminutive marker of some sort, that then became feminine while unmarked nouns defaulted to masculine.

Ah...
In that case, maybe use good old sexual dimorphism applied to females of other animals, and then by extension to humans.

[Xwtek]

Sorry for misogyny, but is it possible to derive diminutive into feminime gender? Proto Garaudawa initially only has 2 genders, and it is reflected in Rkouic and converted into Proximate-Absolutive in Asent'oan, but split into four genders in the rest of language, taking original diminutive. (No longer productive in both Rkou and Asent'o) The resulting gender is Masculine, Feminime I, Neuter, and Feminime II. Later, Feminime II collapses either into Feminime I or to Neuter.

Let’s put it this way: you yourself may not be misogynistic. Your con-people are allowed to be, so long as you acknowledge it. But it’s normally better to have some other factor coming into play for the feminine form to become a diminutive, maybe a similarity (after sound change) to some affix that functions as the diminutive. Or maybe the Feminine II form is also used for children, and then later becomes associated with the smaller form of anything (while the Feminine I becomes the major, well, feminine form).

The originally is the diminutive suffix is originally formed by -i- suffix for most athematic noun, but -si- for thematic noun and some athematic noun (The other one uses prefix). The other uses for -i- suffix is an ordinary thematic suffix, otherwise no relation (Used in *kuni- woman, *xukradi- baby, *tafari- chief). The diminutive were used as a way to cope with homophony in Asent'o, later it becomes meaningless morpheme. In Rkou, due to extensive final vowel loss and reduction, the diminutive vowel were also lost, so it ceases to be contrastive. In Kumbikan, the -si- suffix were generalized, but leaving the few words untouched. However, due to many feminime words has -i on them, it causes broadening in meaning, so the original athematic noun were feminized with -i-, and this spreads to thematic -si-. Later -Vs- were lost.

The diminutive suffix originally has a meaning like -tzin in Nahuatl.

(The words is not fixed yet)

[Xwtek]

Is have > semelfactive realistic?

I went to a town once

1SG.NOM PRES-have-1.SG.PRES.IND PST-go-1.SG.PST.RES town-SG.LOC

Where RES is a verb form that is used to either mark a finished action that is still relevant or denoting a result of an action.

[zompist]

Isn't that exactly what the English experiential perfect is? Cf. "I've been to a town."

[akam chinjir]

Isn't that exactly what the English experiential perfect is? Cf. "I've been to a town."

It's a bit tricky: "I've been to a town" is an experiential perfect but "I've gone to a town" is a perfect of resulting state. But maybe it's only with "go" that you get that sort of alternation? Anyway the experiential perfect isn't derived from the perfect of result, which I think is what Akangka is contemplating doing.

(WALS has data on have-perfects: https://wals.info/chapter/68. It seems to be mostly a European thing. But the squib about this in the World Atlas of Grammaticalization, p. 245, implies that this gets you a perfective of result.)

[Xwtek]

Isn't that exactly what the English experiential perfect is? Cf. "I've been to a town."

It's a bit tricky: "I've been to a town" is an experiential perfect but "I've gone to a town" is a perfect of resulting state. But maybe it's only with "go" that you get that sort of alternation? Anyway the experiential perfect isn't derived from the perfect of result, which I think is what Akangka is contemplating doing.

(WALS has data on have-perfects: https://wals.info/chapter/68. It seems to be mostly a European thing. But the squib about this in the World Atlas of Grammaticalization, p. 245, implies that this gets you a perfective of result.)

Not only for "go". This construction is productive to many forms of verbs.

1SG.NOM PRES-have-1.SG.PRES.IND PST-eat-1.SG.PST.RES meat-SG.LOC
I ate meat once

Also, to form the perfective of result, the translation is:

1SG.NOM PST-go-1.SG.PST.RES town-SG.LOC PRES-arrive-3.SG.PRES.RES

And that have construction have never been a perfect of result.

[Knit Tie]

Hello!

Still working on my last future English "precoglang" and I have two questions.

Firstly, how common is it to have a phonology with /x/, but no /h/? What about a [x~h] morpheme with realisation dependent on position?

Secondly, I want to have a system that independently marks for tense, aspect, mood, evidentiality and also for "a little," "a standard amount" or "a lot" obligatorily for each verb unless you have a series of verbs with the same T, A or M, then it's unmarked and assumed to be the same as the last marked verb. How can I develop this system out of modern English's mostly analytical morphology? Can I also stick subject and object marking on the verb on top of this TAM-evidentiality-amount marking system?

[akam chinjir]

Firstly, how common is it to have a phonology with /x/, but no /h/? What about a [x~h] morpheme with realisation dependent on position?

This is normal; a prominent example is Mandarin. (Admittedly you won't go far with the assumption that if it happens in Mandarin phonology then it must be normal.)

Secondly, I want to have a system that independently marks for tense, aspect, mood, evidentiality and also for "a little," "a standard amount" or "a lot" obligatorily for each verb unless you have a series of verbs with the same T, A or M, then it's unmarked and assumed to be the same as the last marked verb.

I'd guess it depends a bit on what sort of verb series you're talking about, though I'd also guess that if you're happy dropping some of English's current morphology you could pull it off. One thing you can do is build on English constructions that would be called serial verb constructions if English had less morphology, like the andative ("I'm going running"; imperative andatives are already bare: "go eat!").

How can I develop this system out of modern English's mostly analytical morphology?

As I hinted, I wouldn't be surprised if it's easier the more analytic your starting point is---that way the existing morphology won't get in the way.

In fact my amateur impression is that the way to get to a significantly agglutinative language is to start with something completely analytic, design your syntax so you have a fairly complicated verb complex (verb sequences should actually be helpful here), and then stop hitting the space bar when you type out your verb complex. (Well, for the last step what you really do is make everything into a single phonological word, first of all by making it so that nothing besides the main verb has independent stress.)

Can I also stick subject and object marking on the verb on top of this TAM-evidentiality-amount marking system?

I suspect it would be weird not to include some kind of cross-referencing, though maybe clitics would be enough. (Agreement confuses the hell out of me.)

[Whimemsz]

Firstly, how common is it to have a phonology with /x/, but no /h/? What about a [x~h] morpheme with realisation dependent on position?

This is normal; a prominent example is Mandarin. (Admittedly you won't go far with the assumption that if it happens in Mandarin phonology then it must be normal.)

Another prominent example would be most dialects of Spanish

[Salmoneus]

Hello!

Still working on my last future English "precoglang" and I have two questions.

Firstly, how common is it to have a phonology with /x/, but no /h/? What about a [x~h] morpheme with realisation dependent on position?

A lot. I do think that if you have /x/ but no /h/, it's probably quite likely than your /x/ will sometimes be realised as [h], but I'm not certain of that.

Secondly, I want to have a system that independently marks for tense, aspect, mood, evidentiality and also for "a little," "a standard amount" or "a lot" obligatorily for each verb unless you have a series of verbs with the same T, A or M, then it's unmarked and assumed to be the same as the last marked verb.

OK, so...
- generally true "marking" isn't optional, it's on all verbs or none (though that's not absolute)
- what you're describing may be a serial verb construction, in which case it is common for a verb on the 'outside' of the complex to bear the marking for all of them (that is, first verb takes prefixes, last verb takes suffixes, etc)
- alternatively, are you sure the verbs really are taking morphological marking, rather than this being indicated by clitics?

How can I develop this system out of modern English's mostly analytical morphology?

You stick words onto the verbs. For instance, you could easily have (in an American language) suffixes kir- ('slightly', from "kind-a") vs. tol- ('a great deal, fully' from "totally"), with the intermediate condition unmarked.

Can I also stick subject and object marking on the verb on top of this TAM-evidentiality-amount marking system?

Sure, why not?

[Knit Tie]

Yeah, serial verb construction is what I'd like to do, with the whole construction being treated as a single verb. However, I'd also like to be able to mark the verbs in the construction that are different from the rest, i.e. "John is going to go to the store, buy a few eggs and cook an omelette," would be, with noun incorporation, "John storego-littleeggbuy-omelettecook." Not sure what it'll be in the proper gloss yet, or if I'd end up using noun incorporation. Without it, it'd be like "I decieve-lie-sleep my brother," "I pretended to be asleep to fool my brother."

Anyway, with regards to having verb series, would you say that it's better to say that verbs cannot join into a series unless they agree in all the TAM+evidentiality+subject+object? Or say that they can, in which case the different verb is marked for whatever is different but still otherwise behaves as part of the sequence?

I'd also ask if it's possible to have an obligatory marker for the little-moderate-a lot thing if the moderate amount is unmarked? How it it not simply two clitics with meanings of "a little" and "a lot" then?

[akam chinjir]

Serial verb constructions are usually defined so that they must be monoclausal, and that's usually (sensibly) taken to mean that there can only be one TAM setting for the whole sequence. In some supposed serial verb constructions this TAM setting gets marked on all the verbs, but that's definitely not necessary.

Another consequence usually drawn is that the verbs in a serial verb construction can't be separately negated---so if you want to be able to say, maybe, "John storego-noteggbuy" (he did go to the store, but didn't buy eggs), then you're probably not talking about serial verbs.

Another possibility is clause chains, in which each verb gets its own clause. Here you can have independent TAM and negation and so on. But some languages let you mark all but one of the verbs in a chain with a suffix that indicates same-subject and same-TAM. Turkish -ip is an example:

market-e   gid-ip yumurta al -d   -ım 
store -DAT go -SS eggs    buy-PAST-1sS
"I went to the sore and bought eggs"

(My Turkish is really awful and I can't decide if the object should be marked explicitly as a plural, giving yumurtalar. That use of -ip is pretty normal, though, except that I don't know how real linguists gloss it.)

One more point if you decide that it's serial verbs, the order of the verbs in a serial construction has a very strong tendency to be iconic. (The tendency is so strong that I've seen it argued that a functional explanation can't work, on the grounds that functional explanations can only explain general tendencies, not exceptionless patterns.) I mention this because your "I deceive-lie-sleep my brother" looks wrong to me, at least if it's a serial verb: the deception of the brother is the purpose of the fake sleeping, so I think you'd expect something more like "I lie-sleep-deceived my brother."

I'd also ask if it's possible to have an obligatory marker for the little-moderate-a lot thing if the moderate amount is unmarked? How it it not simply two clitics with meanings of "a little" and "a lot" then?

Clitics might not be as integrated with their host, phonologically speaking, as affixes---this can affect things like stress assignment, and you might have phonological processes that are blocked by clitic boundaries but not by affix boundaries. Also, clitics are often flexible in what they attach to---maybe an a lot clitic could be separate from the verb by another adverb, for example (and then attach to that other adverb instead of the verb).

[Hallow XIII]

Like most converbal forms, -Ip tends to be glossed "converb", because figuring out what each converbal marker actually does in a given language is a non-trivial exercise, especially in light of [we've known converbs primarily have discourse-relevant functions for 20 years and still I dont think anybody's actually followed up on that].

[masako]

What is "obligatory"? A mood, tense, aspect?

halidu
go-OBL-3s
She is obliged to go.