Page 35 of 35

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

Posted: 19 Jul 2021 13:41
by sangi39
Omzinesý wrote: 19 Jul 2021 11:25 Maybe the hard question is still where such a consonant system could appear.
I mean, barring the lack of /b d g f/, it's basically got the same consonants as Dutch and various Frisian languages, the latter of which also saw (some) fronting of velars, and a merger of the dental fricatives into the alveolar plosives (I think). So you could do something like:

Code: Select all

Stage 0) m n p b t d k  ɡ  kʷ ɡʷ ɸ θ ð s z x xʷ j w l r (Proto-Germanic)
Stage 1) m n p b t d ts dz kʷ ɡʷ ɸ θ ð s z x xʷ j w l r (Fronting of Velars)
Stage 2) m n p b t d ts dz k  ɡ  ɸ θ ð s z x x  j w l r (delabialisation)
Stage 3) m n p b t d ts dz k  ɡ  ɸ θ ð s r x x  j w l r (rhoticization of *z)
Stage 4) m n p b t d ts dz k  ɡ  ɸ t d s r x x  j w l r (fortition of dental fricatives)
Stage 5) m n p b t d ts dz k  ɡ    t d s r x x  j w l r (loss of *ɸ ~ also becomes /x/ in certain environments)
Stage 6) m n p b t d ts z  k  ɡ    t d s r x x  j w l r (deaffrication of /dz/)
Stage 7) m n p β t ð ts z  k  ɣ    t d s r x x  j w l r (lenition of voiced plosives)
Stage 8) m n p ʋ t D ts z  k  ɣ    t d s r x x  j w l r (changes to fricatives)
Stage 9) m n p ʋ t D ts z  k  ɣ    t d s r x x  j ʋ l r (merger of *w into /ʋ/)
Stage 6 (if you ignore the missing /ɸ/) is, from what I can tell, a fairly standard consonant system for the various Frisian languages and Dutch, it's just the unconditional shift of the voiced plosives to voiced fricatives (barring /g/ > /ɣ/ which seems relatively common in the area) that stands out

That would make it fairly typical for a continental Germanic language near the North Sea, and then it just takes the shift of the voiced plosives to voiced fricatives a couple of steps further (Proto-Germanic already conditionally saw variation between the two sets, and this seems to have continued through to Old Dutch as well, so that "Stage 7" doesn't even have to be one step, it could just be the last step in a number of conditional changes that had been going on for centuries)

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

Posted: 19 Jul 2021 16:00
by All4Ɇn
In addition to the above, I think it'd be interesting to maybe add-in early devoicing of /b d ɡ/ in certain situations. If you're looking for Finnic influence you could have /p t k/ occur finally and alternate in grammatical forms with /ʋ z ɣ/.

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

Posted: 20 Jul 2021 11:32
by Omzinesý
VaptuantaDoi wrote: 14 Jul 2021 11:07 I've been reading Dixon's Ergativity, and it's got me wanting to make a (believable) ergative romlang. I've come up with two possible ways an ergative system could develop, namely:

1. A Gallo-Romance language in which the marked nominative incentivises an ergative-absolutive system, first only in masculine nouns but later spreading to feminine nouns, articles, adjectives and third-person pronouns. First- and second- person pronouns would remain nominative-accusative, which is a relatively common split for a split-S language to have. I already have an idea for where to place this an an althistory.

2. An Eastern Romance language in which the passive voice was used extensively until it becomes the only form for transitive sentences, which then had a subject in the absolutive and an object in the nominative. This would affect all nouns and pronouns, making it a fully ergative language; it would also preserve the Latin passive which could be interesting, and I also have a vague idea of where to set this one.

My question is, which one would be more realistic/interesting? Should I make a thread about one of them? Or about both?
I think this was a good one:
Esteval and Myhill 1988: Formal and theological aspects of the development from passive to ergative systems
This might also be interesting:
https://www.academia.edu/12126567/From_ ... ldai_2008_

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

Posted: 20 Jul 2021 11:48
by Omzinesý
Thank you. It actually doesn't seem that strange anymore. Northern Germany could be good, close enough to Dutch and Danish.
I find making anything based or affected by Finnish a bit hard.

Actually, Danish seems to have /t/ => /ts/ and *d realizing as /t/. I remembered it was only allophonical before /i/ or something.
I don't know how naturally it could be combined with /g/ > /ɣ/.

But probably
1) aspiration of/p t k/
2 /t/ => /ts/
3) /g/ > /ɣ/ (at any state)
4) because /ph/ and /b/ are the only phonemes where aspiration/voicing is distinctive, they merge word-initially. That coul coappear with /ɣ/ => /x/ word-initially as in Dutch.

Word-medially, Proto-Germanic representation of voiced fricatives could be preserved all the time.

I must try googling about Frisian. Wikipedia says quite little about them.

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

Posted: 21 Jul 2021 15:55
by teotlxixtli
I'm trying to develop a tense system with four past tenses and two future tenses, but I have no idea how to go about generating those endings organically. How does such a system evolve in the real world?

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

Posted: 21 Jul 2021 17:17
by eldin raigmore
teotlxixtli wrote: 21 Jul 2021 15:55 I'm trying to develop a tense system with four past tenses and two future tenses, but I have no idea how to go about generating those endings organically. How does such a system evolve in the real world?
In at least one natlang the four past tenses are accomplished with a stem-alternation and an auxiliary.
Like:
Immediate past : stem
Recent past: aux stem
Intermediate past: stam
Remote past: aux stam

Or something like that.

Is that understandable?

….

What semantics are you intending to communicate via your four past tenses and two future tenses?

More than one natlang with three pasts and two futures has:
Pre-hesternal past (before yesterday)
Hesternal past (yesterday)
Hodiernal past (earlier today)
Present
Hodiernal-and-crastinal future (later today or tomorrow)
Post-crastinal future (after tomorrow).

Are you intending something like that?

….

Are you doing degrees-of-remoteness? Like today vs yesterday vs before yesterday, or this year vs last year vs before last year?

Or, perhaps instead, stacking of relative tenses?
Like
Past;
Future within past;
Past within past;
Future within future within past;
Past within future within past;
Future within past within past;
Past within past within past?

Though simpler than that, I’d guess!

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

Posted: 21 Jul 2021 21:40
by teotlxixtli
eldin raigmore wrote: 21 Jul 2021 17:17
teotlxixtli wrote: 21 Jul 2021 15:55 I'm trying to develop a tense system with four past tenses and two future tenses, but I have no idea how to go about generating those endings organically. How does such a system evolve in the real world?
In at least one natlang the four past tenses are accomplished with a stem-alternation and an auxiliary.
Like:
Immediate past : stem
Recent past: aux stem
Intermediate past: stam
Remote past: aux stam

Or something like that.

Is that understandable?

….

What semantics are you intending to communicate via your four past tenses and two future tenses?

More than one natlang with three pasts and two futures has:
Pre-hesternal past (before yesterday)
Hesternal past (yesterday)
Hodiernal past (earlier today)
Present
Hodiernal-and-crastinal future (later today or tomorrow)
Post-crastinal future (after tomorrow).

Are you intending something like that?
This exactly! I remember watching an Artifexian video that talked about Yemba (I think) and I wanted to have something like that. What I don't know, however, is what auxiliaries to use for different degrees of remoteness. At least one will have to be used to establish one degree, then to create the stem change you described I'd need to generate an umlaut-type vowel change with an affix that's lost over time, so that's another auxiliary to combine to the root...

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

Posted: 23 Jul 2021 00:52
by VaptuantaDoi
Davush wrote: 15 Jul 2021 11:18 I think some Iranic and Neo-Aramaic(s) use the ergative only in the past/perfective forms, but I think it would be reasonable for the ergative-marked forms to spread into non-past verbs, especially if previous person-marking on non-past verbs had become very eroded or obscured. Some Neo-Aramaic actually developed ergativity, but then gradually generalized the ergative forms for all subjects of both transitive and intransitive verbs, leading back to a nom-acc alignment. So essentially you'd just need to decide how intransitive verbs don't end up with the same subject marking (either via retention or innovation)...
That's good enough for me [xD] I think I should be able to keep subject marking separate by innovating something for imperfect transitive verbs, probably based off auxiliaries.
Omzinesý wrote: 20 Jul 2021 11:32 I think this was a good one:
Esteval and Myhill 1988: Formal and theological aspects of the development from passive to ergative systems
This might also be interesting:
https://www.academia.edu/12126567/From_ ... ldai_2008_
Thanks! Although I couldn't find the first one online, I found a bunch of other papers which reference it and sound like they'll be useful.
shimobaatar wrote: 18 Jul 2021 13:38 To address the last point here, I'd be interested in seeing a thread about both. [:)]
I think I'll work on them offline until I'm relatively happy with them, then I'll make a thread comparing the two of them.

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

Posted: 24 Jul 2021 14:04
by eldin raigmore
In those split-ergative languages in which the split is based on something about the verb rather than (or in addition to?) something about the noun,
being past and/or perfective and/or realis increases the odds the patient will be unmarked absolutive while the agent will be marked Ergative;
while being future and/or imperfective and/or irrealis increases the odds the agent will be unmarked nominative while the patient will be marked accusative.

….

The OP of this thread is about how the actual marking arises diachronically, rather than how it correlates with the semantics.
Any information about that question is news to me! And I’ve enjoyed reading it and look forward to more!

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

Posted: 26 Jul 2021 18:32
by Shemtov
Is it possible to have noun-like adjectives that don't decline for case, even though the nouns do?

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

Posted: 26 Jul 2021 19:02
by WeepingElf
Shemtov wrote: 26 Jul 2021 18:32 Is it possible to have noun-like adjectives that don't decline for case, even though the nouns do?
Wouldn't they be no longer noun-like then?

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

Posted: 26 Jul 2021 19:06
by Shemtov
I'm asking if it's possible for the adjectives to mark gender and number and not case, even if nouns decline for all three.

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

Posted: 26 Jul 2021 21:22
by LinguistCat
Shemtov wrote: 26 Jul 2021 19:06 I'm asking if it's possible for the adjectives to mark gender and number and not case, even if nouns decline for all three.
I don't see why not, especially if they tend to stay near the nouns they modify. But I don't know if it's attested anywhere.

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

Posted: 26 Jul 2021 21:30
by CarsonDaConlanger
Shemtov wrote: 26 Jul 2021 19:06 I'm asking if it's possible for the adjectives to mark gender and number and not case, even if nouns decline for all three.
I'd expect it in a language where adjectives tend to be closer to verbs than nouns. Since verbs are very commonly marked for number and gender but not case, you'd expect the same for adjectives. The question is would they be in the same spot in relation to the root as the verbal affixes?

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

Posted: 27 Jul 2021 20:56
by Omzinesý
Shemtov wrote: 26 Jul 2021 18:32 Is it possible to have noun-like adjectives that don't decline for case, even though the nouns do?
Theoretically interesting question!

What is a noun-like adjective?

Noun as a comparative concept: say, words that typically express things and words that are morpho-syntactically part of the same category
Noun as a language-internal descriptive category: say, English nouns can typically be preceded be an article and have two numbers.

I think, noun-like adjectives must be noun-like in the sense of descriptive category.
If nouns in that language have case inflection, adjectives can be noun-like in other respects but not in that if they lack it. So, noun-likeness is quantitative rather than qualitative. In the least extreme situation, it's enough that adjectives are not verb-like.

My conlangs often express characteristics as genitives of abstract nouns "man of honour". There could still be something making adjectives noun-like instead of nouns, say, they form comparatives.

Latin, famously, inflects nouns and adjective similarly:
silva densa
silvae densae
silvas densas
silbibus densibus
...


Of course, it is also possible that adjectives have a case paradigm but it does not appear in all contexts.
Hungarian for example does not (usually) inflect adjectives when they modify nouns but inflects them when they are predicates. (Source: http://www.hungarianreference.com/Adjec ... tives.aspx )

a boldog fiú 'the happy boy'
a boldog fiúk 'the happy boys'

But
A fiú boldog. 'The boy is happy.'
A fiúk boldogok. 'The boys are happy.'
Shemtov wrote: 26 Jul 2021 19:06 I'm asking if it's possible for the adjectives to mark gender and number and not case, even if nouns decline for all three.
It's funny that no language comes to my mind. But I'm sure there are some. Adjective agreement is such a redundant phenomenon that it can basically choose any category to agree with without increasing/decreasing info.

Hungarian comes close (but its adjectives modifying nouns do sometimes inflect).

There could actually be syntactic reasons for adjectives not to have case inflection. If they never appear as heads and don't agree as modifyers, they don't appear in syntactic positions where cases are needed, if we don't interpret the unmarked nominative in the predicate form a case. Even if, adjectives could sometimes appear as heads and then have a full case paradigm, we can interpret them zero-derived nouns. (Finnish plural adjectives can appear in Nominative or Partitive as predicatives (with subtle semantic differences), so this was not to say adjectives couldn't have case inflection in those syntactic positions.)


Probably, the question was if adjectives as modifiers of nouns can have number/gender agreement but lack case agreement.
No language come to mu mind, but surely it happens in some lang.

If cases are a new innovation, it could be very likely. Estonian has some cases (actually it's questionable if they are syntactically cases and not postpositions because they don't cause agreement, but anyways) that don't cause agreement:

suure auto-ga
big car-COM
'with a big car'

suur-te auto-de-ga
big-PL car-PL-COM
'with big cars'
(I'm not sure if the examples are corrects (my skills are limited) but they should convey the idea.)

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

Posted: 28 Jul 2021 07:41
by Shemtov
In the end, I decided that since in the conlang I was talking about the genitive and dative are based on the "oblique" stem, which on its own is the accusative, I could have the adjective inflect only for nominative, "oblique" and vocative.