Conlang Random Thread

[keenir]

Well, so much for that theory; still, its an interesting idea, so I may try to get it to work in a paper model language.

a conlang is not a communication system,
for me it's more an experiment in philosophy of language...
________________________________________________
c·c¨h¨¼e¬Pc¨r¨«zª¹\ªc¹­

If my memory serves, (and to be fair, it doesn't always), you told us at least once, that you use 3SDL to write journal entries and notes to yourself. Thus you essentially told us that 3SDL is a communication system: you write in it, for yourself.

Do you not expect to be able to read your own messages to yourself?

Indeed, not being able to read it
forces me to recompose the meaning,
each time...

Well, that answers that.

So you made something for your own use...and you can't use it? You can't look at a reminder you put on the fridge, and say "Yup, i need to pay the water bill" or "I nearly forgot I have guests coming over, I should tidy up" ?

These two are from ten days apart, and clearly don't use the same orthography, but it's notable that both at least have word spacing and aren't full of random special characters. The second even seems to be basically 1:1 with its gloss, though the first is less clear.

My overall impression is that xxx makes his language impenetrable on purpose, for I assume personal reasons, but it still seems to me to be an actual structured conlang.

I was hoping the answer to the latest question would confirm your statement...but unless I misunderstand words like "recompose", it seems not. Drat.

[keenir]

this is not the purpose...

Then what is the purpose of something that, like the proverbial doctor's handwriting, you can't even read your own message?

Hmm...so how do we know that your post with that edited pic you said was about the Breton language and its paralel(?) with your conlang...but now we can accuse you of writing in 3SDL there that you favor the genocide of {snip} or of unicorns and krakens and all partisans of them both - and you can't prove us wrong.

EDIT: had to remove political war before it catches fire again.


wait, is 3SDL supposed to be...I think the word is asemic? something where there is no meaning to what is written? (this was proposed for Voynich, i believe; and its the case for Codex Seriphinius(sp))

if there is only one path, each journey has its own way of following it...
so that I can't guess it without crossing it again...

Guess?
Why would you want to guess at a warning label like "Do Not Drink: Poison" or a fortune cookie paper that says "Tickling Babies Is Fun"?

[xxx]

I don't use it to write a diary
but only to compose meaning as I go along,
and in a way that's impossible to guess in the future......
______________________________________________
c·c¨e¹c·JJeª­.ª¨KPc¨.ª¼¨c·M

[keenir]

I don't use it to write a diary
but only to compose meaning as I go along,
and in a way that's impossible to guess in the future......

okay, how do you "compose meaning"? and what, to you, does "compose meaning" mean, if it has no meaning?

(ie, when I wrote myself a note saying "stomach hurts. see doctor"...thats not something I want to guess about later)

[bradrn]

its an interesting idea, so I may try to get it to work in a paper model language.

I’d be interested in seeing this, at least!

(Also, thanks to salem for explicating some of the sentences. I’d also noticed some repeating elements, but not enough to stop me doubting myself.)

[jal]

I'll leave this here, again, and hope you'll stop engaging with them. I've put his ramblings on ignore, but y'all keeping engaging forces me to see their nonsense anyway...

One should not try to engage with xxx and hope to get any meaningful answer. I won't say they're a troll, but it's a thin thread they walk on...

JAL

[Ahzoh]

I think I'll stick with marking persons with suffixes, because with my current conjugation system, it has a nice aesthetic with some verbs.

conjugatedallverbs.png (77.06 KiB) Viewed 1630 times

But this is also why I struggle, it just doesn't feel like there's any room for additional morphemes to change grammatical voice. On the other hand, I feel limited in what morphemes to come up with because not a lot of consonants that pair well with others (which is why I have not explored infixation), not to the degree that /n t s/ do. A lot of my morphemes I make are sonorous (like nasals, liquids, and glides) and "neutral" (like coronals).

Essentially, P(V)T, T(V)P, T(V)T, T(V)K, and K(V)T categorical combinations feel ok, but P(V)K and K(V)P categorical combinations do not. So, as the root PaRaḪ- is my conlang's Fa'aLa, everything I come up with should be harmonious with it. But it's difficult.

---

Part of me just doesn't want any valence-changing operations whatsoever, but there are apparently no such languages without valence-changing operations (decreasing or increasing) morphologically or paraphrastically.

The other part of me wants to be able to do some of the kind of things Akkadian can do:
https://universaldependencies.org/akk/f ... bStem.html

Then there is coming up with morphemes or nonconcatenative morphology for the deverbals derived from verbs. And that is also where deciding on what grammatical voices I have is important.

Most importantly I think it is useful to have a deverbal that fulfills the role of action nominal, cognate object, and infinitive. After that, I need a deverbal that can modify nouns but can also stand alone as a substantive, such as a participle.

So an action deverbal and adnominal deverbal. If I have a causative voice then there must also be causative counterparts to these deverbals (because it is a derivational voice, not an inflected one). If I have a passive voice then I must also have passive counterparts. If I have both a passive and a causative voice, then my work now becomes fourfold and more difficult.

---

For the longest time, I've had a "pattern" of R₁aR₂R₂aR₃- which denoted the verbal adjective, conveying either a passive, resultative, or attributive sense depending on the transitivity class of the verb (active, dynamic intrasitive, and stative intransitive, repsectively).

Though I've also thought to use it to denote the causative. But if I do then I don't know how to derive a deverbal adjective from the causative derived stem because I don't want a R₁aR₂R₂aR₂R₂aR₃- pattern.

Also, if I don't have a passive voice, it feels wrong to have a deverbal adnominal with a passive sense without also a deverbal adnominal with an active sense.

[keenir]

O.O

I think part of the problem may be that you're trying to do / fix too many things all at once. Perhaps pick one thing to fiddle with, and focus exclusively on that - ignore all the other things.

Or just start a new conlang/model lang, and work on that to take your mind off the wounded kaiju below?

I think I'll stick with marking persons with suffixes, because with my current conjugation system, it has a nice aesthetic with some verbs.
conjugatedallverbs.png

But this is also why I struggle, it just doesn't feel like there's any room for additional morphemes to change grammatical voice. On the other hand, I feel limited in what morphemes to come up with because not a lot of consonants that pair well with others (which is why I have not explored infixation), not to the degree that /n t s/ do. A lot of my morphemes I make are sonorous (like nasals, liquids, and glides) and "neutral" (like coronals). It is more acceptable to have words like tapraḫni but kapraḫni and papraḫni feels wrong. So, as the root PaRaḪ- is my conlang's Fa'aLa, everything I come up with should be harmonious with it. But it's difficult.

---

Part of me just doesn't want any valence-changing operations whatsoever, but there are apparently no such languages without valence-changing operations (decreasing or increasing) morphologically or paraphrastically.

The other part of me wants to be able to do some of the kind of things Akkadian can do:
https://universaldependencies.org/akk/f ... bStem.html

Then there is coming up with morphemes or nonconcatenative morphology for the deverbals derived from verbs. And that is also where deciding on what grammatical voices I have is important.

Most importantly I think it is useful to have a deverbal that fulfills the role of action nominal, cognate object, and infinitive. After that, I need a deverbal that can modify nouns but can also stand alone as a substantive, such as a participle.

So an action deverbal and adnominal deverbal. If I have a causative voice then there must also be causative counterparts to these deverbals (because it is a derivational voice, not an inflected one). If I have a passive voice then I must also have passive counterparts. If I have both a passive and a causative voice, then my work now becomes fourfold and more difficult.

---

For the longest time, I've had a "pattern" of R₁aR₂R₂aR₃- which denoted the verbal adjective, conveying either a passive, resultative, or attributive sense depending on the transitivity class of the verb (active, dynamic intrasitive, and stative intransitive, repsectively).

Though I've also thought to use it to denote the causative. But if I do then I don't know how to derive a deverbal adjective from the causative derived stem because I don't want a R₁aR₂R₂aR₂R₂aR₃- pattern.

Also, if I don't have a passive voice, it feels wrong to have a deverbal adnominal with a passive sense without also a deverbal adnominal with an active sense.

[Ahzoh]

I think part of the problem may be that you're trying to do / fix too many things all at once.

That's what happens when you have a specific aesthetic in mind but you're also shit and unimaginative at diachronics. Read all I can about diachronics and still I cannot take that knowledge and turn morphological shit into gold.

Or just start a new conlang/model lang, and work on that to take your mind off the wounded kaiju below?

I can't make a million conlangs like other people do, it inevitably leads to the same problem. Also, making forks of the same language leads to its own form of frustrating hell.

[keenir]

I think part of the problem may be that you're trying to do / fix too many things all at once.

That's what happens when you have a specific aesthetic in mind but you're also shit and unimaginative at diachronics. Read all I can about diachronics and still I cannot take that knowledge and turn morphological shit into gold.

no, you cannot turn five tonnes of morphological cowpatties into pure gold. but then, nobody who tried turning anything into gold, ever started with such huge quantities.

Or just start a new conlang/model lang, and work on that to take your mind off the wounded kaiju below?

I can't make a million conlangs like other people do, it inevitably leads to the same problem.

I know my memory is swiss cheese on a good day...but I'm pretty sure I never suggested you make a million conlangs.

you're a great conlanger and a good person; I don't want to suggest you retire from the craft, but if you think its the only way to solve this problem, I'll back your call.

[bradrn]

Part of me just doesn't want any valence-changing operations whatsoever, but there are apparently no such languages without valence-changing operations (decreasing or increasing) morphologically or paraphrastically.

Morphologically or periphrastically, yes… but you’re only worried about morphology here! It’s perfectly plausible to have no morphology for voice or valency-changing.

(If you want some practical examples of this, English doesn’t really mark voice morphologically — it just has a single past participle form, and builds the passive on top of that with syntax. French works the same way, plus it has a middle voice formed by changing the object clitic pronoun.)

[Ahzoh]

Morphologically or periphrastically, yes… but you’re only worried about morphology here! It’s perfectly plausible to have no morphology for voice or valency-changing.

(If you want some practical examples of this, English doesn’t really mark voice morphologically — it just has a single past participle form, and builds the passive on top of that with syntax. French works the same way, plus it has a middle voice formed by changing the object clitic pronoun.)

I don't want periphrastic voice either, then it's too English.

[bradrn]

Morphologically or periphrastically, yes… but you’re only worried about morphology here! It’s perfectly plausible to have no morphology for voice or valency-changing.

(If you want some practical examples of this, English doesn’t really mark voice morphologically — it just has a single past participle form, and builds the passive on top of that with syntax. French works the same way, plus it has a middle voice formed by changing the object clitic pronoun.)

I don't want periphrastic voice either, then it's too English.

Well, in that case, you’ll have to compromise on one of your requirements.

Anyway, a periphrastic voice isn’t necessarily ‘too English’ — there’s other ways of doing it. Tamil, for instance, uses an infinitive form together with an auxiliary paṭu ‘fall, happen’. Kristang uses toká ‘touch’ in a serial verb clause.

If you want a morphological voice but don’t feel like working out the nonconcatenative morphology, you could also use a simple affix. After all, this is the most common way of marking passive amongst natlangs. And no-one says your morphology has to be completely nonconcatenative. (Indeed, I don’t believe such a language exists — all attested languages with nonconcatenative morphology also contain concatenative morphology.)

[Ahzoh]

If you want a morphological voice but don’t feel like working out the nonconcatenative morphology, you could also use a simple affix.

Well yes, that's part of my problem, I've been having trouble figuring out where to put the voice affixes without shifting stress away from the root. I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker. I wanted to have applicative voices, but it's hard to come up with affixes that sound ok next to other consonants. And I've already used up affixes using /n/ and /t/ for some of the mood marking prefixes.

I like to look at a variety of Mesopotamian/Ancient Near East languages for inspiration, like seeing words like meḫušurrut and evrišše and ḫašimbur and lagaštakkaš and see what I can take apart and turn into a word or morpheme.

[keenir]

If you want a morphological voice but don’t feel like working out the nonconcatenative morphology, you could also use a simple affix.

Well yes, that's part of my problem, I've been having trouble figuring out where to put the voice affixes without shifting stress away from the root.

I'm probably misremembering, but didn't you already say that you had it so the stress won't shift?

I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker.

Why not both?

I wanted to have applicative voices, but it's hard to come up with affixes that sound ok next to other consonants. And I've already used up affixes using /n/ and /t/ for some of the mood marking prefixes.

I'm sure nobody would object to you using /s/ with /n/ and-or with /t/ as affixes.

[bradrn]

I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker.

You could make it both! Passive/causative polysemy is well-attested from several languages (e.g. Korean, Olutec, Sonrai).

[Ahzoh]

I'm probably misremembering, but didn't you already say that you had it so the stress won't shift?

Yeah, I never ever said that. My words stress placements are highly mobile. But that means if I want stress to only fall on the root, then I can only have CV, CVV, CVC, CVCV CVCVV, and CVCVC suffixes and those are already taken up by the person markers, the negative suffix that attaches to the person markers, and the 2nd and 3rd person plural marker that gets tacked on at the end.

Why not both?

No.

I wanted to have applicative voices, but it's hard to come up with affixes that sound ok next to other consonants. And I've already used up affixes using /n/ and /t/ for some of the mood marking prefixes.

I'm sure nobody would object to you using /s/ with /n/ and-or with /t/ as affixes.

Issues would arise having na- be the future prefix and then having a nV voice prefix. I already can produce words with three pairs of geminated n's because of assimilation rules. Same can happen with /t/

[Ahzoh]

I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker.

You could make it both! Passive/causative polysemy is well-attested from several languages (e.g. Korean, Olutec, Sonrai).

The causative in this language is a lexical voice, that means whatever conveys the causative voice is a derived stem. The passive would not be a derived stem but inflectional, like the TAM markers. So it won't work. Not well, anyways.

I don't really need a passive voice, nor do I really want one. But I feel that if I have a deverbal adnominal, it must have an active-passive binary. Though I know of some language, Hittite or Hurro-Urartian, that only has a passive participle. And Akkadian' deverbal adjective is passive if derived from a transitive verb, resultative if derived from a dynamic intransitive, and attributive if derived from a stative intransitive verb. And it is the Akkadian deverbal adjective that I base my conlang's deverbal adnominal. But it feels wrong to have without a deverbal adnominal in the active sense.

Moreover aside from that, I have no other proper means to derive true adjectives. Attributive functions only ever come from nouns inflected in the equative case or verbs in a relative clause.

[keenir]

I wanted to have applicative voices, but it's hard to come up with affixes that sound ok next to other consonants. And I've already used up affixes using /n/ and /t/ for some of the mood marking prefixes.

I'm sure nobody would object to you using /s/ with /n/ and-or with /t/ as affixes.

Issues would arise having na- be the future prefix and then having a nV voice prefix. I already can produce words with three pairs of geminated n's because of assimilation rules. Same can happen with /t/

I'm not seeing a problem.

(btw, who exactly would be having the issues that arise? I believe the expression is "please stop borrowing worries from next week, when you haven't even finished building today")

I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker.

You could make it both! Passive/causative polysemy is well-attested from several languages (e.g. Korean, Olutec, Sonrai).

The causative in this language is a lexical voice, that means whatever conveys the causative voice is a derived stem. The passive would not be a derived stem but inflectional, like the TAM markers. So it won't work. Not well, anyways.

well thats just aweful. no, not that, i mean the other aweful. is there a third sort of aweful?

I don't really need a passive voice, nor do I really want one. But I feel that if I have a deverbal adnominal, it must have an active-passive binary.

must? you are the lord of the four corners here, god of gods and king of kings, who opens the doors and knows the ways. if you say it doesn't have a passive, no problem.

Moreover I have no proper means to derive true adjectives, they only ever come from inflected nouns or verbs in a relative clause, which feels wordy.

didn't we just tell you do try doing just one thing at a time?

[bradrn]

I also have only come up with one voice affix (si-/ši-) that I haven't decided if it's a passive marker or a causative marker.

You could make it both! Passive/causative polysemy is well-attested from several languages (e.g. Korean, Olutec, Sonrai).

The causative in this language is a lexical voice, that means whatever conveys the causative voice is a derived stem. The passive would not be a derived stem but inflectional, like the TAM markers. So it won't work. Not well, anyways.

Firstly, the ‘inflectional’/‘derivational’ distinction is overrated. It makes especially little sense with categories like the passive, which have aspects of both. So I don’t consider this a valid objection.

[…] I feel that if I have a deverbal adnominal, it must have an active-passive binary.

Why?