Conlang Random Thread

[Travis B.]

Another numbers change to Rihalle Kaafi -- nouns which are marked with numeric clitics for one through nine, and their adjectives, are inflected as singular, even though this does not affect agreement by verbal agreement clitics with them (e.g. a number marked with taa- 'two' would be agreed with as dual and a number marked with ah- 'three' would be agreed with as plural).

[evmdbm]

Would you consider
"I made the man fight the soldier" to be object control?

The reason I ask is that I have decided (I think) that I've got into a muddle. In Vedreki a sentence like "I ordered him to go" is basically constructed as in English
A qhalas ona iraq
I-nom order-past-masc-sg he-acc go-inf

But I also have a morphological causative infix -ökhax-.. So
He died
On qalus
He-nom die-past-msc-sg

I killed the man
A qalökhaxus sirega (I caused him to die)
I-nom die-caus-past-masc-sg man-acc-sg

So far so far good, I think, but the problem arises when I cauativise a transitive verb. Here I put the causee in the genitive case
I made the man fight the soldier
A tarökhaxas siregem qalgona
I-nom fight-CAUS-past-masc-sg man-gen-sg soldier-acc-sg

or periphrastically implying a somewhat more indirect relationship, and also because I wondered how I would deal with several causative relationships in a row (I made the man make the slave.... sort of thing)

I made the man fight the soldier (but I wasn’t there)
A tabus taraq siregem qalgona
I-nom make-Aux-past-masc-sg fight-inf man-gen-sg soldier-acc-sg

Now I'm thinking is this not just object control, so "I ordered him to fight the man" should also use the genitive.
A qhalas oneb taraq sirega
I-nom order-past-masc-sg he-gen fight-INF man-acc-sg

Or should I have two accusatives? But if I have two accusatives and the causee is in the accusative should I not also have two accusatives if I use the causative infix? Is there actually a difference between my object control clauses and where I'm causativising a transitive verb. Perhaps I should have a rule that you can only morphologically causativise intransitives? Would that be a plausible rule? It would also avoid my other messy question about passivising causatives. Easy if the underlying verb is intransitive

He was killed
On qalökhaghaxus
He-nom die-caus-pass-past-msc-sg

But what about "The man was made to fight the soldier?" If I passivise the morphological causative, do I end up with a passive with a direct object? I'm not sure I want that (although English has passives with direct objects: "He was given the book"). Might it not be easier just to have the periphrastic construction, particularly if it's just object control.

It's all one big long problem, but I think basically one problem. Fingers crossed it makes sense, what I'm getting at.

[bradrn]

Would you consider
"I made the man fight the soldier" to be object control?

In English, I believe it’s some form of raising-to-object (it’s fine to say ‘God made there be light’), but am not sure — it might be something different from both of them. You might have better luck asking in the Syntax random thread.

But either way, I strongly suspect that this is something which would differ between languages depending on the precise details of their causative construction.

(For that matter, I’m not convinced these processes are universal in the first place — there are just too many languages without deranked subordinate clauses, for one thing. Although they strike me as being more likely to be universal than most syntactic terminology is.)

The reason I ask is that I have decided (I think) that I've got into a muddle. In Vedreki a sentence like "I ordered him to go" is basically constructed as in English
A qhalas ona iraq
I-nom order-past-masc-sg he-acc go-inf

[…]

or periphrastically implying a somewhat more indirect relationship, and also because I wondered how I would deal with several causative relationships in a row (I made the man make the slave.... sort of thing)

I made the man fight the soldier (but I wasn’t there)
A tabus taraq siregem qalgona
I-nom make-Aux-past-masc-sg fight-inf man-gen-sg soldier-acc-sg

These two constructions seem to use different verbs — is that intentional? (Of course it’s perfectly reasonable to have several different causative verbs.)

Now I'm thinking is this not just object control, so "I ordered him to fight the man" should also use the genitive.
A qhalas oneb taraq sirega
I-nom order-past-masc-sg he-gen fight-INF man-acc-sg

Or should I have two accusatives? But if I have two accusatives and the causee is in the accusative should I not also have two accusatives if I use the causative infix? Is there actually a difference between my object control clauses and where I'm causativising a transitive verb. Perhaps I should have a rule that you can only morphologically causativise intransitives? Would that be a plausible rule?

Responding to each of these in turn:

1. Choose whichever case-marking you want, as long as you can come up with a consistent language-internal explanation for it.
2. You could also have two accusatives if you want, yes.
3. Why should the causative infix behave the same way as the periphrastic causative in the first place?
4. I don’t understand the question.
5. That rule is attested in natlangs, yes.
6. Anything can be plausible if you have a consistent explanation for it.

But what about "The man was made to fight the soldier?" If I passivise the morphological causative, do I end up with a passive with a direct object?

Well, yes, this is what happens if you passivise ditransitive verbs. I’m not aware of any natlang with a constraint against doing so.

[AwfullyAmateur]

The Sodemeresh word for mushroom, asarkada, comes from asar (edible plant) and kadi (holy)...

[Travis B.]

The Sodemeresh word for mushroom, asarkada, comes from asar (edible plant) and kadi (holy)...

Do the Sodemeresh speakers use magic mushrooms for religious rites?

[AwfullyAmateur]

The Sodemeresh word for mushroom, asarkada, comes from asar (edible plant) and kadi (holy)...

Do the Sodemeresh speakers use magic mushrooms for religious rites?

Yes, they believe the hallucinations are divine visions, which is why they're sacred plants.

[rotting bones]

I'm trying to create a conlang that is spoken in one city: https://docs.google.com/spreadsheets/d/ ... sp=sharing

The effect I'm trying to achieve is that the high register of the language preserves fuller forms of ancient words. At the same time, it draws most of its vocabulary from an unrelated imperial language (Luquri, locally Laukor). (The "Ancient" column is derived from the same proto-language as the low register. It's not a direct ancestor.)

Rough ideas for an alphabet: https://drive.google.com/file/d/1OAE9Y6 ... sp=sharing

[AwfullyAmateur]

I like that one of the names for the thread goddess is tulle. Is that a reference to the fabric?

[rotting bones]

I like that one of the names for the thread goddess is tulle. Is that a reference to the fabric?

Yes, and silver is basically rupee.

I'm thinking of a story like Lobsang Rampa's focused on rediscovering the ancient language by visiting archives in the imperial capital.

I'm a total conlanging noob, so any advice would be appreciated. For example, I don't have a handle on how to suggest that an extremely shortened word is related to a longer one despite extensive sound changes.

[malloc]

I'm a total conlanging noob, so any advice would be appreciated. For example, I don't have a handle on how to suggest that an extremely shortened word is related to a longer one despite extensive sound changes.

Yeah, that can be rather tricky. The word "lord" for instance derives from a worn-down compound "loafward" or more precisely hlāfweard in Old English. Lenition and semantic drift can really mess up words in unexpected ways.

[bradrn]

I'm trying to create a conlang that is spoken in one city: https://docs.google.com/spreadsheets/d/ ... sp=sharing

How are we supposed to pronounce these words? I assume it’s not using IPA…

Rough ideas for an alphabet: https://drive.google.com/file/d/1OAE9Y6 ... sp=sharing

I’m pleasantly surprised to see that the letterforms here are actually ones which could plausibly be written with a broad-edged pen. This is something a lot of people get wrong.

[rotting bones]

How are we supposed to pronounce these words? I assume it’s not using IPA…

Most of them follow conlang stereotype rules:

X is like ch in Bach.
Q is like in Arabic.
Ts is like in Japanese.

I'm not so sure about these:

Pf is like in German.
C can be like ch in English. It becomes j if there's a voiced consonant before or after it (ignoring vowels).

In addition, some of the words are not pronounced exactly as written. E.g. There may be unwritten glottal stops. The words were sent through a sound change applier of sorts, but they were primarily chosen for their appearance on the page. This might change at some point.

[bradrn]

X is like ch in Bach.
Q is like in Arabic.
Ts is like in Japanese.

I'm not so sure about these:

Pf is like in German.
C can be like ch in English. It becomes j if there's a voiced consonant before or after it (ignoring vowels).

Aargh, aargh, I hate it… please give us (and yourself) a coherent phonology, for the love of God! (Or whichever deity you prefer!)

[rotting bones]

Aargh, aargh, I hate it… please give us (and yourself) a coherent phonology, for the love of God! (Or whichever deity you prefer!)

All my conlangs start out as isolated words mentioned in stories before I fix them. The fixing doesn't usually go as far as the stories.

[rotting bones]

Found my notes. I don't know how helpful this is. The changes are minimalistic right now. If I don't like a derived word, I add intermediate contextual changes until I do. The notes roughly say:

Proto-Great-Mountain:

Consonants:

p, t, k,
pʰ, tʰ, kʰ,
m, n, ŋ,
s, ʃ, tʃ,
r, l

Ritual language:

Changes that always happen:

pʰ -> pf
tʰ -> ts
kʰ -> kx
tʃ -> ts
ʃ -> s

Some changes that happen in some contexts:

r -> l
Various vowel reductions and additions.
s -> ts

Words borrowed from the Rumsa substrate have the consonants:

p, t, k
ph, th, kh
m, n
s, ts, h
r, l

Proto-Luquri:

k, q, t
d
m, n
s, h
l

High register:

Honestly, I'm not sure my script is up to date with these notes.

pʰ -> ph
tʰ -> th
kʰ -> kh
tʃ -> ch
ʃ -> sh
q -> k
d -> r

Contextual changes:

Final vowel drop
s -> th or sh
Vowel harmony effects (usually merges)
Vowel lengthenings
H before vowel
k -> kh

[bradrn]

pʰ -> ph
tʰ -> th
kʰ -> kh
tʃ -> ch
ʃ -> sh

These don’t make any sense to me, for various reasons. If they’re IPA, /ph th kh/ are just a different way of transcribing /pʰ tʰ kʰ/, while /tʃ/→/ch~cʰ/ and /ʃ/→/sh~sʰ/ would make no phonetic sense. Meanwhile, if they’re not IPA, these are just describing different romanisations rather than any linguistic change.

[rotting bones]

These don’t make any sense to me, for various reasons. If they’re IPA, /ph th kh/ are just a different way of transcribing /pʰ tʰ kʰ/, while /tʃ/→/ch~cʰ/ and /ʃ/→/sh~sʰ/ would make no phonetic sense. Meanwhile, if they’re not IPA, these are just describing different romanisations rather than any linguistic change.

Right, in the high register, h after a consonant makes it ejective. I didn't want to type a backslash to escape the single quote and forgot to fix it afterwards. I don't like the look of a quotation mark inside a word anyway, so maybe I should keep it.

PS. Ejective because of this: https://en.wikipedia.org/wiki/Ejective_ ... Hypothesis

[bradrn]

These don’t make any sense to me, for various reasons. If they’re IPA, /ph th kh/ are just a different way of transcribing /pʰ tʰ kʰ/, while /tʃ/→/ch~cʰ/ and /ʃ/→/sh~sʰ/ would make no phonetic sense. Meanwhile, if they’re not IPA, these are just describing different romanisations rather than any linguistic change.

Right, in the high register, h after a consonant makes it ejective. I didn't want to type a backslash to escape the single quote and forgot to fix it afterwards. I don't like the look of a quotation mark inside a word anyway, so maybe I should keep it.

Hmm… interesting convention. Apostrophes are indeed the most common notation, but elsewhere in ASCII, Na'vi has used ⟨px tx kx⟩. You can also get away with transcribing it as a voiced/voiceless distinction, if you don’t have one already.

[rotting bones]

Hmm… interesting convention. Apostrophes are indeed the most common notation, but elsewhere in ASCII, Na'vi has used ⟨px tx kx⟩. You can also get away with transcribing it as a voiced/voiceless distinction, if you don’t have one already.

IIRC the only voiced consonant in the high register is m. The low register and Luquri have them though. The low register is what I'm trying to fix right now. If I axe the voiced consonants, using them to represent the ejectives is a good idea. Thanks.

Edit: Actually, it has n, l, r too, but my point still stands.

[bradrn]

Hmm… interesting convention. Apostrophes are indeed the most common notation, but elsewhere in ASCII, Na'vi has used ⟨px tx kx⟩. You can also get away with transcribing it as a voiced/voiceless distinction, if you don’t have one already.

IIRC the only voiced consonant in the high register is m. The low register and Luquri have them though. The low register is what I'm trying to fix right now. If I axe the voiced consonants, using them to represent the ejectives is a good idea. Thanks.

You're welcome! Though if I recall correctly (and I may not be), the convention is usually the other way round: that is, ⟨g k⟩ for /k kʼ/.

Also, I just discovered that Chechen has used ⟨h⟩ for ejectives. So it may be justified after all…

(Oh, and yet another convention I forgot: doubling the letter, as in Hadza.)