(Conlangs) Q&A Thread - Quick questions go here

A forum for all topics related to constructed languages
Salmoneus
MVP
MVP
Posts: 2352
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

There's no answer to that, because "time travel" is not a well-defined concept.

The two big questions would be: how widespread is time travel?; and what are the consequences of time travel on causality? [is this an eternal loop scenario, or a rewriting/many-worlds scenario? And if it's an eternal loop, does it permit self-caused events?]


[let's assume massive time-migration.
With self-caused events in an eternal loop, you could have people at A speaking English, and then suddenly people arrive from the future speaking Blagblag, a language that has never previously existed, and they convert everyone at A to speaking Blagblag, so it ends up being the majority language at B (a later time), at which point people go back and introduce Blagblag back to A.

With an eternal loop without self-causation, it doesn't matter what language the time-travellers speak, since they cannot alter the speech at A: if they did, they would have influenced the development of B-language, and hence partially caused it, which would be impossible. So either time travel is impossible, or else there must be sinister laws of nature that act to 'randomly' (but predictably) prevent paradoxes through bizarre contrivances and coincidences.

With many-worlds/rewriting time-travel, the people from B1 influence the language at A2, which develops into the language at B2, which influences the language at A3, and so on. So essentially, from the point of view of someone at A-tenmillion, language change suddenly happens infinitely quickly and everyone's speaking Blagblag out of nowhere all of a sudden - but from a metatemporal or transdimensional perspective, Blagblag has evolved out of English perpendicular to the flow of time, propagating and developing itself across metatime instead of across time. On the other hand, in this worldview, everyone at every point in time is speaking every possible language, depending on your point of view, so it doesn't really matter...]
User avatar
Creyeditor
MVP
MVP
Posts: 4336
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

It's again a question with a definitive answer: it depends [:D]
I think I like the mass migration many worlds scenario best. Maybe it's just due to starting to read "the long utopia" a while ago.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Titus Flavius
hieroglyphic
hieroglyphic
Posts: 63
Joined: 13 Apr 2021 14:53

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Titus Flavius »

How can i romanize these vowels? I have no idea:
?
User avatar
Creyeditor
MVP
MVP
Posts: 4336
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

One possibility:

short lax:
<ì ù>
<è ò>
<æ̀ à>
long lax:
<î û>
<ê ô>
<æ̂ â>
long tense:
<i u>
<e o>
<æ a>
overlong tense:
<í ú>
<é ó>
<ǽ á>

This means that the long close open front vowel has two possible romanizations: either long lax mid front vowel <ê> or long tense low front vowel <æ>.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Titus Flavius
hieroglyphic
hieroglyphic
Posts: 63
Joined: 13 Apr 2021 14:53

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Titus Flavius »

Thanks!
Now I think I may use this:

short lax

ĭ ŭ
ĕ ŏ
ă ău

long lax

ı u
e o
a au

long tense

i u̇
ė ȯ
ȧ ȧu

overlong tense:

ī ū
ē ō
ā āu
User avatar
LinguoFranco
greek
greek
Posts: 505
Joined: 20 Jul 2016 17:49
Location: U.S.

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by LinguoFranco »

Is isochrony useful for setting up a conlang? I know some people try to categorize languages as syllable-timed, stress-timed and mora-timed, but with most things linguistic, most natlangs don't fall neatly into any of those categories.

Is it still worth thinking about when designing a conlang?
User avatar
Creyeditor
MVP
MVP
Posts: 4336
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

It might make sense to get some intuitive feeling for types that are different from your native language, but IMHO you should not base your conlang's phonology and phonetics on it, as far as naturalistic conlanging is concerned. Like most holistic typologies isochrony is often essentialistic and empricially wrong (as you already pointed out).
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Pabappa
greek
greek
Posts: 490
Joined: 18 Nov 2017 02:41
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Pabappa »

lately i have a thing for frontloading syllables .... e..g. treating clusters as onsets instead of dividing them between leading and trailing syllables. this can affect other things ..... e.g. in Poswa, at one point, a syllable could begin with /ps/ but not with /pt/, which led to different developments for preceding vowels. /ipt/ shifted to /ept/ (because it was a closed syllable), but /ips/ remained /ips/. later, the syllable structure reverted to the more classical SAE style and a second round of vowel changes happened with both the inherited and the new clusters .... /ips ipt/ shifted to /ups upt/.
I'll take the theses, and you can have the thoses.
User avatar
eldin raigmore
korean
korean
Posts: 5849
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by eldin raigmore »

Isn’t what you call “front-loading” also known as “the Maximum Onset Principle”?
User avatar
Pabappa
greek
greek
Posts: 490
Joined: 18 Nov 2017 02:41
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Pabappa »

actually, thats a good point i hadnt thought of ..... i dislike clusters such as /sp/ in any position, so they tend not to occur in my languages and therefore the syllable onsets typically obey the sonority hierarchy. even so, a lot of these languages also have prenasals, such as in the name of Ntampamwa, a military leader, which is segmented as /nta-mpa-mwa/. so while i've never heard of it before, with the exception of the prenasals i do obey the maximum onset principle.

anyway my mentioning this was intended to show how syllable structure could interplay with sound shifts, and how /pt/ could trigger a shift that /ps/ did not, without just relying on random chance.
I'll take the theses, and you can have the thoses.
User avatar
Davush
greek
greek
Posts: 602
Joined: 10 Jan 2015 14:10

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Davush »

I have been considering the following sound change for a diachronic sketch I'm playing with, any ideas on how plausible it is?

j > z when it is the onset of a stressed syllable: *járan > záran.

Except ji > di, (mostly) unconditionally. Along with ri > di (/r/ does not change otherwise). *káji > kádi, *pasári > pasádi. This language does not otherwise have voiced stops, giving a stop-inventory of /p t d k/.

Probably assuming there was an intermediate but short-lived j > ɟ (or similar) stage. I think I have come across instances of both /j/ and /r/ behaving "oddly" before /i, but I'm not totally sure...
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4239
Joined: 31 Mar 2013 13:16

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by DesEsseintes »

Davush wrote: 21 May 2021 08:51 I have been considering the following sound change for a diachronic sketch I'm playing with, any ideas on how plausible it is?

j > z when it is the onset of a stressed syllable: *járan > záran.

Except ji > di, (mostly) unconditionally. Along with ri > di (/r/ does not change otherwise). *káji > kádi, *pasári > pasádi. This language does not otherwise have voiced stops, giving a stop-inventory of /p t d k/.

Probably assuming there was an intermediate but short-lived j > ɟ (or similar) stage. I think I have come across instances of both /j/ and /r/ behaving "oddly" before /i, but I'm not totally sure...
Abau, a language that lacks phonemic coronal stops, has [d] as an allophone of /r/ before /i/.

As for j → d, I think your own explanation using an intermediate stage is perfectly plausible. Thai has fortition of /j/ to [ɟ] before /i/ (personal experience, cannot provide a source), so that part at least is attested (by me).
Salmoneus
MVP
MVP
Posts: 2352
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

I think this sort of single-phoneme change is kind of generally weird - in contrast to more systematic changes - and is weirder the more phonemes you have (small inventories often have a lot of weirdness, with high levels of allophony and phoneme-specific changes, whereas large inventories seem to me to tend to behave more systematically).

However, weird things do happen - even in Romance, /g/ didn't behave completely parallel to either /k/ or /d/. And the details of your changes seem very plausible to me: closing before /i/, and fortition when stressed. And a drift from palatal to coronal. All three of those seem like common things to happen. So yes, this all seems reasonable to me, particularly in a language with a smaller inventory.

However, I don't think it gives you a /p t d k/ inventory/! I think it gives you /p t z r k/, with [d] as a neutralised allophone of /r/ and /z/ before /i/!
User avatar
Davush
greek
greek
Posts: 602
Joined: 10 Jan 2015 14:10

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Davush »

Thank you both - good to know it's not totally implausible. It is a small inventory, so also good to know that such changes happen more commonly in these (Austronesian and Papuan being inspiration).

@Salmoneus: I agree with your analysis of /p t z r k/ and [d] as allophone. I should have mentioned that [d] *might* become phonemic if the following /i/ is elided in things like [dié] > /de/ (or similar).
User avatar
Creyeditor
MVP
MVP
Posts: 4336
Joined: 14 Aug 2012 19:32

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Creyeditor »

I think I recently read of some Niger-Congo language that has a synchronic change of /l/ to [d] before /i/.
Edit: I found it here
An analogous property of liquid hardening is responsible for special realizations of /l/ before glides or high vowels. There are a number of languages where /l/ is realized as [d] before /i/, for example, Ciyao [Bantu; Mozambique] mil-a ‘swallow’ versus mid-isy-a ‘swallow a lot’; kul-a ‘grow big’ versus kud-isy-a ‘grow very big’ (Ngunga 2000:56). In Gbari [Nupoid; Nigeria], /l/ is realized as [ɗ] before /i/, for example, ɗí ‘eat’. In Gwari Kuta /l/ has further developed into a palatal stop with offglide before both /i/ and /e/, such as gyí ‘eat’, gyè ‘sharpen’ (Hyman & Magaji 1971:14, 15).
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Pabappa
greek
greek
Posts: 490
Joined: 18 Nov 2017 02:41
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Pabappa »

arguably it's a retention, though, since proto-Bantu is reconstructed as having [d] before the high vowels /i u/ and [l] elsewhere. this may have something to do with "tension" ... wrong term I know but it's whatever we'd use to describe why Japanese does something similar in merging /d z/ before high vowels and also affricating /t/ before high vowels.

r >d before /i/ therefore makes perfect sense.

I dont know how you'd get from /j/ > /d/, though, unless by /d/ you mean a palatalized stop similar to IPA [ɟ]. i wouldnt recommend depalatalization before /i/ and nowhere else.
I'll take the theses, and you can have the thoses.
Salmoneus
MVP
MVP
Posts: 2352
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

Salmoneus wrote: 07 Apr 2021 13:46
Creyeditor wrote: 07 Apr 2021 09:01
Salmoneus wrote: 06 Apr 2021 16:54 Note that although English is reluctant to put the relative clause between the dummy subject and the verb in the plain "it"-cleft, it does sometimes happen, and you can do it easily if you use "what" instead of "it that" - "what I saw was a dog". It may just be that English has in general become wary of pronoun+that sequences, which sound old-fashioned. "He that removes this sword from this stone shall be king of England" and so on.
I always that that 'what' relative clauses are different because they do not modify any noun phrase? So, 'what' would be part of the relative clause. But maybe this is the closest English equivalent.
[gratuitous Synge quotations]
I know nobody's that interested, but just to resurrect this from a few pages back, I came across a nice Irish sentence on duolingo I thought I'd share, re: clefting.

On the one hand: turns out you CAN front things before the verb sometimes. On the other hand: this still forces you to cleft.

So:
is bean mhaith í - she's a good woman [is woman good she]
emphatic form:
bean mhaith is ea í - it's a good woman she (is) [woman good is it she] - I think that basically the inverted form has been permitted with the neuter pronoun, 'ea', through weight of usage, but nowhere else, so even the feminine pronoun requires the neuter pronoun as a dummy if you want to invert. And it translates directly to that incredibly Irish "it's a good woman she is" construction in English. Except that English also requires an extra dummy copula. [the literal form, "it's a good woman her" is problematic because of the gender agreement!]

The actual duolingo sentence was: bean mhaith is ea í, nach ea?, adding a tag: it's a good woman she is, is it not?


And the additional weird thing about this cross-gender dummy agreement: the neuter pronoun, 'ea', is more or less ONLY used with the copula in inversions, and in tags like this (otherwise everything has to be masculine or feminine)...

[real gender-twisting: would it be possible to say ?cailín maith is ea í, I wonder (it's a good girl that she is)? I haven't gotten that far. The point here being that in Irish, 'girl' is masculine, so you'd have "it-neuter is a good-masc girl-masc, she-fem". However, I'm not sure yet whether Irish insists on grammatical (rather than semantic) gender in animate pronouns. I suspect not, but I don't know.]
Salmoneus
MVP
MVP
Posts: 2352
Joined: 19 Sep 2011 19:37

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Salmoneus »

Pabappa wrote: 22 May 2021 00:38
I dont know how you'd get from /j/ > /d/, though, unless by /d/ you mean a palatalized stop similar to IPA [ɟ]. i wouldnt recommend depalatalization before /i/ and nowhere else.
The point is that you'd be depalatalising everywhere, it's just that it only happens to appear before /i/. I don't think /j/ > /d/ is much more improbably than /j/ > /ddZ/ (Romance), for example.

That said, depalatalising before /i/ and nowhere else is certainly a thing - it's just dissimilation. A parallel is the boukolos rule in PIE: labiovelars are delabiovelarised adjacent to /u/.
User avatar
jimydog000
greek
greek
Posts: 482
Joined: 19 Mar 2016 04:14
Location: Australian Country

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by jimydog000 »

Gleb gave me an interesting sound change once, where Front vowels become centralised after a semivowel, /j/ and /w/.
User avatar
Pabappa
greek
greek
Posts: 490
Joined: 18 Nov 2017 02:41
Contact:

Re: (Conlangs) Q&A Thread - Quick questions go here

Post by Pabappa »

Salmoneus wrote: 22 May 2021 01:33
Pabappa wrote: 22 May 2021 00:38
I dont know how you'd get from /j/ > /d/, though, unless by /d/ you mean a palatalized stop similar to IPA [ɟ]. i wouldnt recommend depalatalization before /i/ and nowhere else.
The point is that you'd be depalatalising everywhere, it's just that it only happens to appear before /i/. I don't think /j/ > /d/ is much more improbably than /j/ > /ddZ/ (Romance), for example.

That said, depalatalising before /i/ and nowhere else is certainly a thing - it's just dissimilation. A parallel is the boukolos rule in PIE: labiovelars are delabiovelarised adjacent to /u/.
okay thanks, i misread .... i saw the line saying there was /ji/ > /di/ in isolation and didnt connect it with the other sound change of /j/ > /z/. so yes this looks fine.

i still dont think a sound shift involving depalatalization before /i/ and nowhere else is a good idea, though, should anyone else want to pursue it .... the boukolos thing is not a good precedent because that change involved losing one of two articulators while the other remains in place, whereas depalatalization before /i/ (and nowhere else) would involve shifting the consonant away from its vowel .... the very opposite of how most sound changes work. and you'd be left with a gap in the syllable inventory, where the only member missing is the simplest one. but oh well .... this doesnt really apply to the question at hand because it's about justifying /z~d/ allophony and not selective depalatalization.
I'll take the theses, and you can have the thoses.
Post Reply