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

A forum for all topics related to constructed languages
User avatar
Shemtov
runic
runic
Posts: 2909
Joined: 29 Apr 2013 03:06

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

Post by Shemtov » 03 Jun 2018 07:56

I know it's well attested for the dual to be restricted to natural pairs from a general dual, but is it possible for the numeral two to be grammaticalized as a dual for natural pairs?
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
LinguoFranco
sinic
sinic
Posts: 413
Joined: 20 Jul 2016 16:49

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

Post by LinguoFranco » 03 Jun 2018 19:57

I have a couple of questions this time.

1. What are the universals and tendencies of dependent-marking languages? I've read that they are always nominative-accusative and usually place the adjective before the noun, but I think Chechen is a dependent-marking ergative language, according to Wikipedia, at least. Yeah, I know, it's not the most reliable source.

2. I'm trying to settle on my phonology. I really like palatal consonants such as /c/, but I don't know if I want palatalization to play a major part in the phonology via sound changes such as palatalization and assimilation. I'm torn between going to the Abkhaz route using a vertical vowel system with allophony based on palatalization and labialization, or a six vowel system with /æ ɑ e̞ i o̞ u/. I guess I don't really have question for this second part other than whether I should do palatal based allophony and go as far as Ubykh or Abkhaz?

EDIT: One more question. Are there any natlang examples of a paucaul number but no dual? Some of my older projects would have singular, paucal, and plural, but every natlang that I know of that has paucal number also has dual number.

User avatar
Pabappa
sinic
sinic
Posts: 214
Joined: 18 Nov 2017 02:41
Contact:

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

Post by Pabappa » 03 Jun 2018 20:44

LinguoFranco wrote:
03 Jun 2018 19:57

2. I'm trying to settle on my phonology. I really like palatal consonants such as /c/, but I don't know if I want palatalization to play a major part in the phonology via sound changes such as palatalization and assimilation. I'm torn between going to the Abkhaz route using a vertical vowel system with allophony based on palatalization and labialization, or a six vowel system with /æ ɑ e̞ i o̞ u/. I guess I don't really have question for this second part other than whether I should do palatal based allophony and go as far as Ubykh or Abkhaz?
I like vertical vowel systems ... i might be able to offer advice if you decide to go that route and come up with a question for which you cant find answers in a natlang. that said, Im not well educated ... most of what I know I've acquired through work with conlangs.
EDIT: One more question. Are there any natlang examples of a paucaul number but no dual? Some of my older projects would have singular, paucal, and plural, but every natlang that I know of that has paucal number also has dual number.
Russian has an archaic paucal number, without the dual ... in fact, it derived from what was earlier a dual. so that would seem to indicate that it is possible.
Image

User avatar
sangi39
moderator
moderator
Posts: 3185
Joined: 12 Aug 2010 00:53
Location: North Yorkshire, UK

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

Post by sangi39 » 04 Jun 2018 00:13

Shemtov wrote:
03 Jun 2018 07:56
I know it's well attested for the dual to be restricted to natural pairs from a general dual, but is it possible for the numeral two to be grammaticalized as a dual for natural pairs?
So, like "two-DUAL eye-DUAL" for "two eyes", rather than, say "two eye-DUAL" where the number "two" is just left unmarked? Or just the number "two" declining as if it were a dual noun?
You can tell the same lie a thousand times,
But it never gets any more true,
So close your eyes once more and once more believe
That they all still believe in you.
Just one time.

User avatar
Shemtov
runic
runic
Posts: 2909
Joined: 29 Apr 2013 03:06

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

Post by Shemtov » 04 Jun 2018 00:34

sangi39 wrote:
04 Jun 2018 00:13
Shemtov wrote:
03 Jun 2018 07:56
I know it's well attested for the dual to be restricted to natural pairs from a general dual, but is it possible for the numeral two to be grammaticalized as a dual for natural pairs?
So, like "two-DUAL eye-DUAL" for "two eyes", rather than, say "two eye-DUAL" where the number "two" is just left unmarked? Or just the number "two" declining as if it were a dual noun?
No. PLR-eye-DUAL or SING-eye-DUAL
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
DesEsseintes
cleardarkness
cleardarkness
Posts: 4681
Joined: 31 Mar 2013 12:16

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

Post by DesEsseintes » 04 Jun 2018 07:37

LinguoFranco wrote:
03 Jun 2018 19:57
I have a couple of questions this time.

1. What are the universals and tendencies of dependent-marking languages? I've read that they are always nominative-accusative and usually place the adjective before the noun, but I think Chechen is a dependent-marking ergative language, according to Wikipedia, at least. Yeah, I know, it's not the most reliable source.
I wonder where you read that. It doesn’t seem to hold up, as Kalaallisut is an erg-abs language with case marking on dependents. Or is it perhaps stating that languages that are strictly dependent-marking (ie no marking of arguments on the verb) are always nom-acc? I would like to know more about this.

User avatar
eldin raigmore
fire
fire
Posts: 6281
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » 04 Jun 2018 14:57

LinguoFranco wrote:
03 Jun 2018 19:57
I have a couple of questions this time.

1. What are the universals and tendencies of dependent-marking languages? I've read that they are always nominative-accusative and usually place the adjective before the noun, but I think Chechen is a dependent-marking ergative language, according to Wikipedia, at least. Yeah, I know, it's not the most reliable source.
Two ideas:

1. Go to the Universals Archive and look up all the (proposed) universals having the keyword “dependent”. That will tell you anything any linguists ever were sure enough to publish. It will also list counterexamples if Plank ever heard of any.

2. Go to WALS.info and compare a “locus of marking” feature to each other feature.

—————

BTW the Archive says ergative ‘langs tend to be dependent-marking; it doesn’t say the converse statistical implication.
And it says active-stative ‘langs and hierarchical ‘langs tend to be head-marking; it doesn’t say the converse.
And what it does say is statistical, not absolute nor universal.

It also says: verb-medial word-order and verb-final word-order favor dependent-marking, while verb-initial word-order and lack of any dominant word-order favor head-marking. But note the (statistical!) implication is from word-order type to marking-locus type, not the other way around.

And it says, dependent-marking languages tend to have adjectives similar to nouns.

And; inalienable-vs-alienable possession is almost never shown by dependent-marking morphology.

Also; clause-affecting relativization* favors dependent-marking and disfavors head-marking. There are several counterexamples but this is a statistical implication so they may not be enough to invalidate it..
*(An example of clause-affecting relativization would be having the verb of the relative clause be a participle instead of a finite verb. There are other, maybe better examples.)

Finally: if a relativization strategy involves pronominalization, it will also probably involve dependent-marking.

User avatar
LinguistCat
cuneiform
cuneiform
Posts: 161
Joined: 06 May 2017 06:48

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

Post by LinguistCat » 04 Jun 2018 17:50

Shemtov wrote:
04 Jun 2018 00:34
sangi39 wrote:
04 Jun 2018 00:13
Shemtov wrote:
03 Jun 2018 07:56
I know it's well attested for the dual to be restricted to natural pairs from a general dual, but is it possible for the numeral two to be grammaticalized as a dual for natural pairs?
So, like "two-DUAL eye-DUAL" for "two eyes", rather than, say "two eye-DUAL" where the number "two" is just left unmarked? Or just the number "two" declining as if it were a dual noun?
No. PLR-eye-DUAL or SING-eye-DUAL
I might be wrong, but I don't think most languages that have dual also mark them as singular or plural, let alone both in different situations. All nouns either get marked singular, dual OR plural. I'd be happy to see evidence from a natlang if I'm wrong, or you might not be going for naturalism. But that's just what I've noticed. It's like saying you'll be making a masc/fem/neut gender system and then also marking all neuter nouns additionally as masculine or feminine.

User avatar
Shemtov
runic
runic
Posts: 2909
Joined: 29 Apr 2013 03:06

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

Post by Shemtov » 04 Jun 2018 18:21

LinguistCat wrote:
04 Jun 2018 17:50
Shemtov wrote:
04 Jun 2018 00:34
sangi39 wrote:
04 Jun 2018 00:13
Shemtov wrote:
03 Jun 2018 07:56
I know it's well attested for the dual to be restricted to natural pairs from a general dual, but is it possible for the numeral two to be grammaticalized as a dual for natural pairs?
So, like "two-DUAL eye-DUAL" for "two eyes", rather than, say "two eye-DUAL" where the number "two" is just left unmarked? Or just the number "two" declining as if it were a dual noun?
No. PLR-eye-DUAL or SING-eye-DUAL
I might be wrong, but I don't think most languages that have dual also mark them as singular or plural, let alone both in different situations. All nouns either get marked singular, dual OR plural. I'd be happy to see evidence from a natlang if I'm wrong, or you might not be going for naturalism. But that's just what I've noticed. It's like saying you'll be making a masc/fem/neut gender system and then also marking all neuter nouns additionally as masculine or feminine.
It's not really a SING marker, as the language has a Bantu-based noun class system. My idea is that younger speakers will start to analyze the singular class marker as unmarked for number, with the plural class marker being a plural marker, but the old singular class marker can be used with a dual suffix.
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
eldin raigmore
fire
fire
Posts: 6281
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » 04 Jun 2018 23:39

LinguistCat wrote:
04 Jun 2018 17:50
I might be wrong, but I don't think most languages that have dual also mark them as singular or plural, let alone both in different situations. All nouns either get marked singular, dual OR plural. I'd be happy to see evidence from a natlang if I'm wrong, or you might not be going for naturalism. But that's just what I've noticed. It's like saying you'll be making a masc/fem/neut gender system and then also marking all neuter nouns additionally as masculine or feminine.
I believe most languages with dual number do mark most dual nouns in a different way from singular nouns and from plural nouns.

But according to “Number” by Corbett. (p. 169 and others), some languages use “constructed duals” for most nouns.
The noun is marked as plural but other words (e.g. the verb) agree with it as if it were singular; or vice-versa.

User avatar
Shemtov
runic
runic
Posts: 2909
Joined: 29 Apr 2013 03:06

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

Post by Shemtov » 05 Jun 2018 02:19

eldin raigmore wrote:
04 Jun 2018 23:39
LinguistCat wrote:
04 Jun 2018 17:50
I might be wrong, but I don't think most languages that have dual also mark them as singular or plural, let alone both in different situations. All nouns either get marked singular, dual OR plural. I'd be happy to see evidence from a natlang if I'm wrong, or you might not be going for naturalism. But that's just what I've noticed. It's like saying you'll be making a masc/fem/neut gender system and then also marking all neuter nouns additionally as masculine or feminine.
I believe most languages with dual number do mark most dual nouns in a different way from singular nouns and from plural nouns.

But according to “Number” by Corbett. (p. 169 and others), some languages use “constructed duals” for most nouns.
The noun is marked as plural but other words (e.g. the verb) agree with it as if it were singular; or vice-versa.
What I'm proposing is for a Bantu-based language which has SING.CLASS and PLR.CLASS prefixes, and younger speakers reanalyze the SING.CLASS prefixes as just a CLASS marker, unmarked for Number (i,e, a NONPLR marker) and thus can accept duals for natural pairs. Is this possible?
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
eldin raigmore
fire
fire
Posts: 6281
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » 05 Jun 2018 07:08

Shemtov wrote:
05 Jun 2018 02:19
What I'm proposing is for a Bantu-based language which has SING.CLASS and PLR.CLASS prefixes, and younger speakers reanalyze the SING.CLASS prefixes as just a CLASS marker, unmarked for Number (i,e, a NONPLR marker) and thus can accept duals for natural pairs. Is this possible?
It’s hard for me to think straight about it.
Corbett says that in some languages some nouns that come in natural pairs do have unmarked duals and marked singulars (or singulatives, maybe?) , as well as marked plurals. Or at least I think I remember his saying something like that.
(See page 96, “4.2.2 The dual in Maltese”, for what I’m hoping is relevant to this. )

So, as far as the diachronics go, I’m guessing my best advice would be “try it and find out!”
I can’t think of an objection.
Do you need your diachronic process to actually be attested in real life, or just sort of similar, as far as anyone can tell, to some things that probably actually have happened?
Assuming the latter; if there’s anything wrong with your idea, I’ll have to leave it up to someone else to tell you!

User avatar
Shemtov
runic
runic
Posts: 2909
Joined: 29 Apr 2013 03:06

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

Post by Shemtov » 05 Jun 2018 07:48

eldin raigmore wrote:
05 Jun 2018 07:08
Shemtov wrote:
05 Jun 2018 02:19
What I'm proposing is for a Bantu-based language which has SING.CLASS and PLR.CLASS prefixes, and younger speakers reanalyze the SING.CLASS prefixes as just a CLASS marker, unmarked for Number (i,e, a NONPLR marker) and thus can accept duals for natural pairs. Is this possible?
It’s hard for me to think straight about it.
Corbett says that in some languages some nouns that come in natural pairs do have unmarked duals and marked singulars (or singulatives, maybe?) , as well as marked plurals. Or at least I think I remember his saying something like that.
(See page 96, “4.2.2 The dual in Maltese”, for what I’m hoping is relevant to this. )

So, as far as the diachronics go, I’m guessing my best advice would be “try it and find out!”
I can’t think of an objection.
Do you need your diachronic process to actually be attested in real life, or just sort of similar, as far as anyone can tell, to some things that probably actually have happened?
Assuming the latter; if there’s anything wrong with your idea, I’ll have to leave it up to someone else to tell you!
It's actually synchronic, as the language is in the middle of a shift in the numeral system.
Many children make up, or begin to make up, imaginary languages. I have been at it since I could write.
-JRR Tolkien

User avatar
eldin raigmore
fire
fire
Posts: 6281
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » 05 Jun 2018 13:08

Shemtov wrote:
05 Jun 2018 07:48
It's actually synchronic, as the language is in the middle of a shift in the numeral system.
“Numeral system” and “grammatical number” are not the same thing. In fact they’re probably nearly independent of each other. (in my experience).

The shift you’re talking about is a step in the evolution of your conlang. It so happens that your current generations of speakers are the ones between whom the change is taking place. Might “contemporary diachronic” be an appropriate way of saying that? Or maybe “synchronic evolution”? —— Anyway, that’s what I meant.

———

What shift in the numeral system are you talking about? Have you already described/explained it and I just missed it?

holbuzvala
cuneiform
cuneiform
Posts: 170
Joined: 01 Jan 2017 14:03

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

Post by holbuzvala » 05 Jun 2018 20:38

I'm doing a smidge of lexicon building, and I was considering the verbs (in English) of 'to choose' and 'to decide'. I'm tempted to fuse these into a single verb in my language, but before that I want to tease out any semantic differences between them to be sure I think they're fusable.

One friend of mine has suggested the main difference is of discreteness and the origin of the sentiment, namely: 1, 'to choose' is when the agent to picks from a discrete number of items presented to him/her/them; while 2, 'to decide' is when the agent desires to do one of possibly numberless potential actions, and that this sentiment arises from within.

I'm not totally convinced by this, so I thought I'd ask here to see what y'all thought.

User avatar
eldin raigmore
fire
fire
Posts: 6281
Joined: 14 Aug 2010 18:38
Location: SouthEast Michigan

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

Post by eldin raigmore » 05 Jun 2018 22:06

One also decides matters of fact.

——

I’m not sure about the discrete number of whatevers for either verb. It seems to me one could choose from among infinite choices; I feel I have done so. And the group of possible choices and/or decisions may be a mass rather than a countable discrete collection; when you decide how big a piece of cake to slice for yourself, for instance.

——

One possible nuance; when you decide something is a fact, that tends to be based on evidence. Often, or maybe even ordinarily, it is a preponderance of credible evidence, or even beyond reasonable doubt. So it subjectively feels like there’s little or no choice in the matter.

But when you choose what to believe, that often involves a noticeable portion of faith. You make your choice to believe in the felt absence of a sufficiency of evidence or reason. It feels very much as if the choice is yours what to believe.


I could be wrong.

Birdlang
greek
greek
Posts: 680
Joined: 25 Dec 2014 20:17
Location: Virginia

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

Post by Birdlang » 08 Jun 2018 23:30

Nachtuil wrote:
22 May 2018 03:05
Birdlang: You could consider just not indicating them with orthography. It really does depend on the rest of your phonology and the phonotactics of the language. Like if you have no /x/ or /h/ using <h> for /ħ/. For the retroflexes, I have seen combinations of regular stops and <r>s done. Like /ʈ/ could be <tr>. Using combinations with <l> might also be logical. You could also consider doubling letters for the retroflexes. /ɖ ʈ ʂ ʐ ʈʂ ɖʐ/ becoming <dd tt ss zz tts ddz> but it does depend on your phonotactics too as you may have confusion at syllable boundaries if say, your language allowed d in both coda and onset. /tod-da/ and /toɖ-a/ might result in the same letters. Diacritics is always an option too. If you give an example of a full phonology it will be easier for people to help you.

...

Here’s the phonology
/m ɱ n ɳ ɲ ŋ ɴ/
/p b p̪ b̪ t d ʈ ɖ k g ʡ ʔ/
/ts dz ʧ ʤ ʈʂ ɖʐ/
/ɸ β f v s z ʂ ʐ ʃ ʒ x ɣ ħ ʕ ʜ ʢ h ɦ/
/ʋ ʋ̪ w l ɾ ɭ ɽ j ɰ ɥ ʕ̞/
/ɬ ɮ/

/i y ʉ ɯ u ʊ e ø o æ ɑ ɒ/ plus length
I’m thinking this
m ɱ n ṇ ɲ ŋ ṅ
p b *ꝕ (p with squirrel tail) *ꞗ (b with flourish tail) t d ṭ ḍ k g *ꝗ (q with stroke) q
c *ꝣ (visigothic z) ƈ ʒ ċ ż
ᵽ ƀ f v s z ṣ ẓ š ž *ꝁ (k with stroke) ǥ ḥ *ꜧ (heng) ɦ *ꝛ (r rotunda) h ḫ
v *ⱳ (w with hook) w l r ḷ ṛ j ÿ ẅ *ꞓ (c with bar)
ɬ ɫ
i ü ŭ ů u û e ö o æ a ḁ plus macron for length.
Ꭓꭓ Ʝʝ Ɬɬ Ɦɦ Ɡɡ Ɥɥ Ɫɫ Ɽɽ Ɑɑ Ɱɱ Ɐɐ Ɒɒ Ɓɓ Ɔɔ Ɖɖ Ɗɗ Əə Ɛɛ Ɠɠ Ɣɣ Ɯɯ Ɲɲ Ɵɵ Ʀʀ Ʃʃ Ʈʈ Ʊʊ Ʋʋ Ʒʒ Ꞵꞵ Ʉʉ Ʌʌ Ŋŋ Ɂɂ Ɪɪ Ææ Øø Ð𠌜 Ɜɜ Ǝɘ

User avatar
LinguoFranco
sinic
sinic
Posts: 413
Joined: 20 Jul 2016 16:49

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

Post by LinguoFranco » 11 Jun 2018 17:34

I'm trying to understand how morae work. If I'm understanding it correctly, then a basic syllable such as V or CV are 1 mora, or a light syllable, while something like VV or CVV or even CVC in some languages are 2 morae, or a heavy syllable. I'm guessing this is the basis for stress in a word? What if a word had, say, a total of 4 morae (like CVV.CVV) or something, would they both be stressed?

Also, why do something languages treat a CVC syllable with a coda sonorant as heavy, while those with an obstruent coda are light?

User avatar
LinguistCat
cuneiform
cuneiform
Posts: 161
Joined: 06 May 2017 06:48

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

Post by LinguistCat » 11 Jun 2018 18:22

LinguoFranco wrote:
11 Jun 2018 17:34
I'm trying to understand how morae work. If I'm understanding it correctly, then a basic syllable such as V or CV are 1 mora, or a light syllable, while something like VV or CVV or even CVC in some languages are 2 morae, or a heavy syllable. I'm guessing this is the basis for stress in a word? What if a word had, say, a total of 4 morae (like CVV.CVV) or something, would they both be stressed?

Also, why do something languages treat a CVC syllable with a coda sonorant as heavy, while those with an obstruent coda are light?
I only really know Japanese but I was just recently looking at pitch accent stuff. Aside from a few dialects that have word level pitch accent, most dialects have specific rules about how pitch is treated based on the pitch accent certain sets of words had at an earlier stage of the language. But to simplify a lot and not bring in recent borrowings or anything like that...

The second part of a 2 mora syllable is considered unable to accept the tone drop - whether it is a second vowel, the syllabic nasal, or the doubling of the consonant of the next mora. So, if for other reasons the tone drop would normally be on that "defective" mora, it moves back to the main mora in the same syllable. For your example of two heavy syllables in a row, it's a little harder because it's likely it would be a compound word, which have their own rules, but tend to leave the accent to the second part of the compound. The most likely pitch accent would be either unaccented or CVV.CV'V with the apostrophe marking where the downstep is.

User avatar
Omzinesý
runic
runic
Posts: 2509
Joined: 27 Aug 2010 07:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý » 11 Jun 2018 18:33

LinguoFranco wrote:
11 Jun 2018 17:34
I'm trying to understand how morae work. If I'm understanding it correctly, then a basic syllable such as V or CV are 1 mora, or a light syllable, while something like VV or CVV or even CVC in some languages are 2 morae, or a heavy syllable.
You are right.

Languages are called syllable-timed or mora-timed.
In a syllable timed language, you try to pronounce both CV and CVC equally heavy, i.e. spend equally long time pronouncing them (I guess that's not acoustically correct but a model of understanding it).
In a mora-timed language, you pronounce the two-moraic syllable more heavy than the one-moraic syllable.
LinguoFranco wrote:
11 Jun 2018 17:34
I'm guessing this is the basis for stress in a word? What if a word had, say, a total of 4 morae (like CVV.CVV) or something, would they both be stressed?
It can be the basis for stressing in a language, it doesn't have to be. There are different stressing rules in different languages. They can be based on morae or not. Japanese, for example, is called a mora-timed language but the concept of stress cannot even be utilized in describing in.
LinguoFranco wrote:
11 Jun 2018 17:34
Also, why do something languages treat a CVC syllable with a coda sonorant as heavy, while those with an obstruent coda are light?
Mora is a theoretical, instrumental concept of describing a language. So the answer to your last question is just that languages do what they do and descriptions have to accept it.

Post Reply