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

A forum for all topics related to constructed languages
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

Oh, I see [:)]
Then yes, both are attested in natlangs. Some natlangs only allow contours in heavy syllables, other natlangs (also) allow contour tones on light syllables.
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]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

One thing that may be worth saying is to bear in mind that tones don't really need to be anchored to anything at all. Once a tone, or tonal element, is generated, it can remain fixed to its original carrier, but it can also in many language wander throughout the word, or even into adjacent words. This can easily happen when sound changes destroy the original carrier but can also just... happen. It seems to happen less the more complicated the tonal system is - it seems very common in two-tone systems.

EDIT: meant to add, this is probably tied to the "register tone" vs "contour tone" typology. Languages in which tones are fixed to a given vowel (syllable, mora...) are more likely to interpret any complex tones as discrete 'contour' tones; those in which tone easily wanders are more likely to intepret complex surface tones as the result of wandering simple register tones. [that is, if the tone on a vowel rises and falls, you can see this either as a "rising-falling" contour tones, or as a sequence of three low-high-low register tones]. Conversely, languages that accept that they have complex contour tones are less likely to have tone movement (though they may still have sandhi) and hence regard tones as fixed to a carrier, while languages that only have simple register tones are more likely to allow tones to drift. I don't know which direction the causality goes in this relationship!
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

Some Grassfields Bantu languages (e.g. Oku) might be the obligatory exception to this tendency. Lots of monosyllabic words with contour tones but (especially in assocative constructions) the contours (or parts of them) tend to wander around a lot. Non-Sinitic Sino-Tibetan languages like Chungli Ao might be similar.
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
sangi39
moderator
moderator
Posts: 3024
Joined: 12 Aug 2010 01:53
Location: North Yorkshire, UK

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

Post by sangi39 »

LinguoFranco wrote: 16 Jun 2022 16:36
Creyeditor wrote: 16 Jun 2022 08:32 There are natlangs that have word-tone melodies like HL oor HLH (where H is a high tone and L is a low . In some languages these yield contour tones like /kânà/ HL.L, /kâná/ HL.H. In other languages contour tones might be more restricted, yielding /kánà/ H.L and /kâná/ for these melodies. Yet other languages might disallow all contour tones and only allow /kánà/ for bisyllabic words here.

To be fully clear, I did not get what you mean by:
LinguoFranco wrote: 15 Jun 2022 17:15 Rising and falling tones seem to occur in contours,
because rising and falling tones are contour tones.
Feel free to ask more questions or the same question again of course.
I meant about rising and falling tones occurring only in heavy syllables such as long vowels.
I think that tends to be where they can arise, since, sort of, you'd have, say, /ka˥.a˩.na˥/ where "long vowels" are probably viewed better as two consecutive short vowels that happen to be the same (I think there are similar views for languages like Japanese). Then what ends up happening is that the tones on the two consecutive vowels end up becoming a contour tone so you get something like /ka:˥˩.na˥/. But that doesn't have to be the end of the process, the vowel could shorten, and retain the falling tone, for example.

Going even further back, the long vowels could come from older diphthongs, e.g. /ai/ > /a:/, or through the loss of some intervocalic consonant, e.g. /aha/ > /a.a/ > /a:/ (this could also lead to other vowel clusters, like /a.e/, that merge into long vowels), and if you wanted to be a little different, it could result from something similar to haplology, e.g. /ba.ba.ka/ > /ba.a.ka/ (the second instance of two identical consonants in adjacent syllables is dropped)

So you could have, for example, ka˥ka˩na˥ > ka˥a˩na˥ > ka:˥˩na˥ > ka˥˩na˥, developing rising and falling tones on short vowels without employing things like closed syllables, loss of coda consonants, etc.
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.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

The vowels need not even be contiguous: once the tone is generated, it can escape the loss of its bearer by shifting to another vowel. So /kánùta/ could become /kânta/.

Contours can also be created through sandhi with adjacent tones, or from the phonation of adjacent consonants. Or from the presence or absence of word boundaries, which often have, as it were, their own inherent tone due to their association with tonal resets. Or from neutralisation of vowel length even in the absence of previous tone (as long vowels will often have a less constant tone than short vowels to begin with, which is easily reinterpeted as an intentional contour).
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

I'm currently trying to develop a phonology where clusters don't occur except in very few cases, but there's also has a large number of a specific type of geminate - one where the first in the pair is effectively a standalone, syllabic consonant.

E.g. /axxa/ is realized as [a.x.xa] (or probably more accurately [aʔxəxa]). The naturalistic plausibility of that is kind of a question in itself, so feel free to comment on that.

My main concern is how this would develop diachronically - specifically, what would be the "main" type of sound change that could lead to this? The most obvious ones would be various forms of assimilation, but they would have to occur for almost every consonant in the inventory, which feels a bit too "laser-guided" to be natural to me - though I intend to have tones, a few diphthongs and vowel length to cope1 with homophones, so maybe that could help drive it along.

The most sensible idea I could come up with is this: first there's a broad loss of clusters except for /ʔC/ or /ʔəC/ sequences, which would then reduce into a duplicate of the following consonant. (e.g. /aʔxa/ > /axxa/, /aʔəma/ > /amma/), sort of as a an anticipation of the next syllable while trying to preserve the time-structure of the word. Having this as a pattern that repeats frequently across the lexicon feels, on the face of it, more natural to me - though I'd mix it in with other changes that would similarly lead to a geminate.

But my knowledge of phonotactics and sound changes in natural languages is pretty poor, and I don't even know where or how I could search for this in targeted ways - so I can only speculate. What are your thoughts? Any ideas?

1 I use "cope" in the sense that I'm intentionally implementing this to reduce homophones and increase the viability of the geminates, not in the sense that the language intentionally develops them as a reaction. These features could develop alongside the loss of clusters, and probably in part due to them. This could make the changes feel less "laser-guided" in turn, I suppose.

EDIT: Note that I'm not suggesting that the proto-language needs to have clusters. I'm only working off of the assumption that, unless the geminates are purely excrecent, there needs to be some kind of reduction process to produce them, at one stage of which there would probably be some kind of cluster, unless complete assimilation has already occured across the board in the relevant places.

EDIT 24.06.2022: After some additional thought - maybe reduplication could be a source? The proto-language would have few to no clusters, but a system of grammatical reduplication. It would then be lost, with the reduplications baked into words in the form of the geminates.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
Aseca
sinic
sinic
Posts: 359
Joined: 04 Jun 2012 14:34

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

Post by Aseca »

Am wondering if pre verb modals or post verbal inflections are better for situationals eg You can eat vs eatable or you may eat vs if you eat?
I noticed that if I stick to post verbal inflections the order of verbs become very regimented (like Japanese) and you just end up sticking to grammatical rules for the person inflections after stacking onto inflected verb forms so I added prefix forms to break up the rules and allow for more flexibility.
Not sure if that's the best thing to do.
Eg
hebr.an -> iza hebriri (eat-able thing) or izár hebran ([he] can.2sg eat.INF)
gi.an -> iz gikvár ki (it happen[wk]-if.3sg to him/her) or iz vár gian ki (it if.3sg happen.INF to him/her)
Sikatāyām kaṇam lokasya darśasi, svargam phale vanye ca.
See a world in a grain of sand, and a heaven in a wild flower.
Ānantam tava karatalena darasi, nityatām ghaṇṭabhyantare ca.
Hold infinity in the palm of your hand, and eternity in an hour.
User avatar
Omzinesý
mongolian
mongolian
Posts: 4079
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý »

Nel Fie wrote: 23 Jun 2022 15:12 I'm currently trying to develop a phonology where clusters don't occur except in very few cases, but there's also has a large number of a specific type of geminate - one where the first in the pair is effectively a standalone, syllabic consonant.

E.g. /axxa/ is realized as [a.x.xa] (or probably more accurately [aʔxəxa]). The naturalistic plausibility of that is kind of a question in itself, so feel free to comment on that.

My main concern is how this would develop diachronically - specifically, what would be the "main" type of sound change that could lead to this? The most obvious ones would be various forms of assimilation, but they would have to occur for almost every consonant in the inventory, which feels a bit too "laser-guided" to be natural to me - though I intend to have tones, a few diphthongs and vowel length to cope1 with homophones, so maybe that could help drive it along.

The most sensible idea I could come up with is this: first there's a broad loss of clusters except for /ʔC/ or /ʔəC/ sequences, which would then reduce into a duplicate of the following consonant. (e.g. /aʔxa/ > /axxa/, /aʔəma/ > /amma/), sort of as a an anticipation of the next syllable while trying to preserve the time-structure of the word. Having this as a pattern that repeats frequently across the lexicon feels, on the face of it, more natural to me - though I'd mix it in with other changes that would similarly lead to a geminate.

But my knowledge of phonotactics and sound changes in natural languages is pretty poor, and I don't even know where or how I could search for this in targeted ways - so I can only speculate. What are your thoughts? Any ideas?

1 I use "cope" in the sense that I'm intentionally implementing this to reduce homophones and increase the viability of the geminates, not in the sense that the language intentionally develops them as a reaction. These features could develop alongside the loss of clusters, and probably in part due to them. This could make the changes feel less "laser-guided" in turn, I suppose.

EDIT: Note that I'm not suggesting that the proto-language needs to have clusters. I'm only working off of the assumption that, unless the geminates are purely excrecent, there needs to be some kind of reduction process to produce them, at one stage of which there would probably be some kind of cluster, unless complete assimilation has already occured across the board in the relevant places.

EDIT 24.06.2022: After some additional thought - maybe reduplication could be a source? The proto-language would have few to no clusters, but a system of grammatical reduplication. It would then be lost, with the reduplications baked into words in the form of the geminates.
Usually geminates are pronounced just as long consonants.

The most obvious one is elision and assimilation: akata => akta => atta

Another source is stressing.
Estonian and some Finnish dialects have:
V́CV: => V́C:V: => V́C:V
When a short stressed vowel is followed by a single consonant that is followed by a long (unstressed) vowel, the stressed syllable is strengthened (because it is so much lighter than the unstressed one) by geminating the consonant. Then the long vowel has been shortened, making gemination phonemic.

Standard Finnish: talo 'house NOM' - taloon 'house ILL'
Some dialects: talo - tallo
Last edited by Omzinesý on 26 Jun 2022 13:23, edited 1 time in total.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
mongolian
mongolian
Posts: 4079
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý »

Aseca wrote: 26 Jun 2022 08:30 Am wondering if pre verb modals or post verbal inflections are better for situationals eg You can eat vs eatable or you may eat vs if you eat?
I noticed that if I stick to post verbal inflections the order of verbs become very regimented (like Japanese) and you just end up sticking to grammatical rules for the person inflections after stacking onto inflected verb forms so I added prefix forms to break up the rules and allow for more flexibility.
Not sure if that's the best thing to do.
Eg
hebr.an -> iza hebriri (eat-able thing) or izár hebran ([he] can.2sg eat.INF)
gi.an -> iz gikvár ki (it happen[wk]-if.3sg to him/her) or iz vár gian ki (it if.3sg happen.INF to him/her)
I don't really understand your question but surely languages can have prefixes and/or suffixes. It depends on your artistic goals.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Omzinesý wrote: 26 Jun 2022 13:18 Usually geminates are pronounced just as long consonants.

The most obvious one is elision and assimilation: akata => akta => atta

Another source is stressing.
Estonian and some Finnish dialects have:
V́CV: => V́C:V: => V́C:V
When a short stressed vowel is followed by a single consonant that is followed by a long (unstressed) vowel, the stressed syllable is strengthened (because it is so much lighter than the unstressed one) by geminating the consonant. Then the long vowel has been shortened, making gemination phonemic.

Standard Finnish: talo 'house NOM' - taloon 'house ILL'
Some dialects: talo - tallo
Thank you for the reply! Yes, based on what I could find, simply "holding" the consonant as opposed to repeating it is the usual approach in natural languages. Maybe it's purely a matter of me misusing the word (i.e. I should have called it a reduplication instead of a gemination).

The examples you give match more or less what I found, though as said, the elision and assimilation pathway seems somewhat unlikely to happen for all consonants (to me, and I'm not a good yardstick on knowing); unless it was transferred to all consonants by analogy as opposed to an unconscious sound change, or the proto-phonotactics were already highly skewed towards facilitating this kind of change.
The Finnish pathway of stress-and-vowel-length-based gemination you offer seems more effective in producing length, rather than repetition. Thus probably not useable in this instance, but I'm sure it'll come in handy for other projects, so thank you for providing it!

I asked the same question elsewhere, and someone offered Japanese as an example, which apparently produced this sort of "stuttered" geminate in some instances - Wikipedia has more details for it, as a starting point (in case anybody else is interested): Japanese Phonology - Geminates
(There are also some additional infos in the "Sound Change" section of the same page.)

Another possible pathway I thought about would be to introduce these sort of geminates as phonological "decoratives". Certain natlangs like Khmer have purely decorative morphology, I'm told. That said, unless anyone else has any other ideas, or I happen to stumble across something by myself, I'll probably just make a decision by fiat and hope for the best.
Omzinesý wrote: 26 Jun 2022 13:21
Aseca wrote: 26 Jun 2022 08:30 Am wondering if pre verb modals or post verbal inflections are better for situationals eg You can eat vs eatable or you may eat vs if you eat?
I noticed that if I stick to post verbal inflections the order of verbs become very regimented (like Japanese) and you just end up sticking to grammatical rules for the person inflections after stacking onto inflected verb forms so I added prefix forms to break up the rules and allow for more flexibility.
Not sure if that's the best thing to do.
Eg
hebr.an -> iza hebriri (eat-able thing) or izár hebran ([he] can.2sg eat.INF)
gi.an -> iz gikvár ki (it happen[wk]-if.3sg to him/her) or iz vár gian ki (it if.3sg happen.INF to him/her)
I don't really understand your question but surely languages can have prefixes and/or suffixes. It depends on your artistic goals.
Idem - I'm not entirely sure I understand but it sounds like the sort of thing you can decide based on your personal preferences.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
Omzinesý
mongolian
mongolian
Posts: 4079
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

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

Post by Omzinesý »

Nel Fie wrote: 26 Jun 2022 15:38 [...]
The examples you give match more or less what I found, though as said, the elision and assimilation pathway seems somewhat unlikely to happen for all consonants (to me, and I'm not a good yardstick on knowing); unless it was transferred to all consonants by analogy as opposed to an unconscious sound change, or the proto-phonotactics were already highly skewed towards facilitating this kind of change.
Not at all.
akata => atta
akana => anna
akava => avva
etc.
Assimilation is nearly always regressive, that is the new geminate takes the value of the second segment, so different second segments produce different geminates. (This is, of course, just the simplest solution, and more interesting things can happen.)
Nel Fie wrote: 26 Jun 2022 15:38 The Finnish pathway of stress-and-vowel-length-based gemination you offer seems more effective in producing length, rather than repetition. Thus probably not useable in this instance, but I'm sure it'll come in handy for other projects, so thank you for providing it!
Northern Saami (remotely related to Finnish and Estonian, but better handled as a contract language) has three lengths of consonants. The longest length is historically produced by a similar phenomenon than what I described for Finnish dialects. The longest grade of some of Northern Saami plosives is sometimes analysed to have an extra-short vowel between the consonants. With a quick Googling, I didn't find those analysis, but it is apparently possible. There are people who know Saami much better than me.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Is that so? Woah.
So, if we understand each other correctly, having a diachronic or synchronic sound change that amounts to:

C1C2 > C2C2

... is a thing that happens in natural languages? Because everything I found so far seemed to suggest this would be too broad to happen for all combinations of consonants, in all environments at a single point in time.
Omzinesý wrote: 26 Jun 2022 20:25 Northern Saami (remotely related to Finnish and Estonian, but better handled as a contract language) has three lengths of consonants. The longest length is historically produced by a similar phenomenon than what I described for Finnish dialects. The longest grade of some of Northern Saami plosives is sometimes analysed to have an extra-short vowel between the consonants. With a quick Googling, I didn't find those analysis, but it is apparently possible. There are people who know Saami much better than me.
Ooh, good suggestion! I'll try looking if I find anything that I can use.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
User avatar
cedh
MVP
MVP
Posts: 386
Joined: 07 Sep 2011 22:25
Location: Tübingen, Germany
Contact:

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

Post by cedh »

Nel Fie wrote: 23 Jun 2022 15:12 I'm currently trying to develop a phonology where clusters don't occur except in very few cases, but there's also has a large number of a specific type of geminate - one where the first in the pair is effectively a standalone, syllabic consonant.

E.g. /axxa/ is realized as [a.x.xa] (or probably more accurately [aʔxəxa]). The naturalistic plausibility of that is kind of a question in itself, so feel free to comment on that.

[...]

EDIT 24.06.2022: After some additional thought - maybe reduplication could be a source? The proto-language would have few to no clusters, but a system of grammatical reduplication. It would then be lost, with the reduplications baked into words in the form of the geminates.
Yes, reduplication is probably a very good source for this. Especially if you want to use it mainly for morphological purposes.

Step 1:
Certain morphological properties are expressed through partial CV- or VC-reduplication (probably in a specific position in the word, e.g. word-initially or adjacent to the stressed syllable): axa :: axaxa, nene :: nenene, tibu :: titibu/tibibu/tibubu

Step 2:
The vowel of the reduplicated syllable (and quite possibly all vowels in unstressed open syllables adjacent to the main stress?) gets reduced to schwa: axa :: axəxa, nene :: nenəne, tibu :: titəbu/tibəbu

Step 3:
/ə/ is reduced to zero (or pure length/syllabicity) between two identical consonants, probably first in quick colloquial speech between two fricatives or nasals, and then optionally in more environments and/or more formal versions of the language: axa :: axːxa, nene :: nenːne, tibu :: titəbu/tibːbu

Step 4:
If there are reduplicated syllables with different consonants on both sides, these can simply undergo assimilation by analogy if they are comparably rare and/or occur almost exclusively in a specific morphological situation, or they can undergo phonological assimilation by schwa reduction followed by Finnish-style creation of geminates: titəbu > titːbu (> tiʔːbu) > tibːbu, or a combination of both.
(Or they could remain distinct, if that seems more interesting to you.)
User avatar
Nel Fie
cuneiform
cuneiform
Posts: 139
Joined: 23 May 2022 15:18

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

Post by Nel Fie »

Thank you for the confirmation and the step-by-step! I'm probably going to include the reduplication path you describe - but as you say, it's an approach that may have greater implications in morphology, and - at least initially - I wanted these geminates to merely be a phonological feature, without any further implications.
But this might be a solvable problem. I'm thinking that if I place the reduplicates far enough in the past, and have the proto-lang already in a place roughly corresponding to your step 2, or shortly before 3, the "contemporary" version of the language might then come out after your step 4, which would only have to do minor work in the grand scheme of things. The geminates would have spread elsewhere through further assimilation and analogy, and other sounds changes - along with grammaticalization and semantic drift - could have erased most traces of "morphologically revealing" geminates while preserving them as a phonological feature.
:deu: Native (Swabian) | :fra: Native (Belgian) | :eng: Fluent | :rus: Beginner
DeviantArt | YouTube | Tumblr
tokibuni
rupestrian
rupestrian
Posts: 22
Joined: 29 Jun 2022 23:44

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

Post by tokibuni »

I'm working on my copula at the moment (the language is VSO). I'm planning to use a couplar verb that only inflects for tense for the predicate nominal and adjective constructions. Where I'm running into problems is planning out how I want the predicate locational, existential, and possessive to function. At the moment, I'm wanting to do a separate copula for locational constructions (be-at essentially). Would I be able to have Predicate Location and Possessive constructions use the same copula? If so, is there any reliable way to distinguish between them, or would context presumably take care of the problem? If I do go with the above, then I was thinking of just using a verb for "exists" for the existential.

I'm just trying to find out if this makes sense or might cause problems.
ara vu buni nnovki malasobuni - The moon is bright at night
- Kuima
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

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

Post by Salmoneus »

To be clear I understand you:

- you want to convey possession using a copula/existential/locational-type verb, in a construction sort of like "the box is with him", instead of with a possessive verb like "he has the box"?

- you want the verb (/copula) used in this construction to be different from the verb used in normal copular predication constructions like "he is a dog" or "the dog is blue"?

- you're concerned that doing this would leave constructions like "the box is with him" and "the box is in Italy" formally identical, so that ambiguity could arise whether the box is located alongside him or in his possession, and whether the box is inside Italy or belongs to it?


If that's what you mean...

- you probably know by now that this is actually the standard construction for old-fashioned European languages. Latin (Proto-Romance) and Proto-Germanic both worked like this, and Celtic languages still do. Outside Europe, it's very common still.

- there are indeed languages in which the existential used in possession is different from the true copula. Irish is an example of this. Well, sort of:
tá sé ann - it is here
tá sé sa teach - it is in the house
tá sé ag an bhfear - the man has it
is madra é - it is a dog

Modern Irish isn't a perfect example, because it actually uses the existential verb for most adjectival predication as well (and sometimes even for nominal predication). However, this is apparently a modern development, and previously the distinction was just as you want it.

- on the question of ambiguity, I'd just ask: what ambiguity? You're assuming that being at, with or for a person is different from them having it - but are they really different? Even if you can sometimes say "it's near him now, but it's actually mine", that could just be seen as a difference in verbal aspect, between something that is generally, habitually, with you, but that happens to currently, momentarily, be with him. I do think it's likely that most languages will have some way of saying explicitly "he has legal ownership rights over this" vs "right this moment it is located physically near to him", but in many languages these disambiguating constructions will not be the normal way of saying either thing.

That said, using an oblique construction for possession doesn't necessarily make it identical to location. For a start, these constructions often put the owner in the dative rather than in a locative. And for another thing, these constructions can use prepositions, and the choice of preposition can make the meaning more clear.

----


The thing I'd actually find LESS likely is having a third, distinct, existential verb only used when there's no oblique argument, because semantically and syntactically these existential possessive constructions ARE existentials. [It exists with me > I have it]. But this isn't impossible, I don't think - I could imagine these constructions forming with an existential verb, but then having 'pure' existential functions taken over by some new verb that doesn't get as far as replacing the old verb in these constructions.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

tokibuni wrote: 29 Jun 2022 23:54 I'm working on my copula at the moment (the language is VSO). I'm planning to use a couplar verb that only inflects for tense for the predicate nominal and adjective constructions. Where I'm running into problems is planning out how I want the predicate locational, existential, and possessive to function. At the moment, I'm wanting to do a separate copula for locational constructions (be-at essentially). Would I be able to have Predicate Location and Possessive constructions use the same copula? If so, is there any reliable way to distinguish between them, or would context presumably take care of the problem? If I do go with the above, then I was thinking of just using a verb for "exists" for the existential.

I'm just trying to find out if this makes sense or might cause problems.
I might have understood your question in a slightly different way from Sal and I will mostly focus on how to distinguish between predicative possessive and predicative locative constructions if both use the same copula. As with many questions, there are basically four options: flagging, indexing, word order, and intonation. I will focus on flagging and word order here.

If you use flagging, the answer is easy. You can distinguish between the two copular sentences based on the adpositions used. Compare the following two pseudo-conlang sentences. The first sentence uses with to introduce the possessed noun, the second one uses in.

Chi nies tim etsik.
I be with box
'I have a box.'

Etsik nies ni dlaw.
box be in forest
'The box is in the forest.'

The second option is to use word order. Locations could for example precede the verb, as they are not really arguments. Possessed noun could follow the possessor. This would mean that you do not need adpositions. Compare the two sentences from another pseudo-conlang. In the first sentence the location forest precedes the verb. In the second sentence the possessed noun box follows the possesor and the verb. This is at least theoretically possible, I don't know if this specific pattern is attested in any natlang.

Natuh ada ayas.
forest be I
'I am in the forest.'

Ada ayas skob.
be I box
'I have a box.'

I agree with Sal that you do not need a third copula for existential constructions, since existential can be easily differentiated in that they only ever take one argument.
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]
tokibuni
rupestrian
rupestrian
Posts: 22
Joined: 29 Jun 2022 23:44

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

Post by tokibuni »

Thank you Salmoneus and Creyeditor, your answers were exactly what I was looking for, so now I have an idea of where to go from here.
ara vu buni nnovki malasobuni - The moon is bright at night
- Kuima
User avatar
LinguoFranco
greek
greek
Posts: 613
Joined: 20 Jul 2016 17:49
Location: U.S.

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

Post by LinguoFranco »

One more question about tone:

So, a register tone language will have tone be carried over the whole word. Some allow contours in long vowels or heavy syllables. Do these contours affect the whole word to?

Say the word /ma.naː.ka/ has a falling tone on the second syllable /naː/, would the high tone spread to the first syllable /ma/, or could it have a low tone? This is assuming /ma.naː.ka/ is just the stem, no affixes or anything.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

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

Post by Creyeditor »

Both are possible, depending on the language. In some languages toneless syllables get their tone by spreading (high) tones. In other languages toneless syllables get a default low tone. This is somewhat simplified, since in many languages both can happen in different contexts. In some languages high tones only spread rightwards for example.
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]
Post Reply