(Conlangs) Q&A Thread - Quick questions go here
- KaiTheHomoSapien
- greek
- Posts: 641
- Joined: 15 Feb 2016 06:10
- Location: Northern California
Re: (Conlangs) Q&A Thread - Quick questions go here
Does it make any sense to have a negative particle only for irrealis verbs? It's my understanding that Ancient Greek has this with μή, but I wasn't sure if that was the exact situation.
Lihmelinyan has two forms of "not": te and ta
te is for indicative
ta is for subjunctive/optative
ánkei te héti = he is not in the house (indicative)
ánkei ta hēti = he may not be in the house (subjunctive)
Any natlangs or conlangs that do something like this?
Lihmelinyan has two forms of "not": te and ta
te is for indicative
ta is for subjunctive/optative
ánkei te héti = he is not in the house (indicative)
ánkei ta hēti = he may not be in the house (subjunctive)
Any natlangs or conlangs that do something like this?
- Creyeditor
- MVP
- Posts: 5091
- Joined: 14 Aug 2012 19:32
Re: (Conlangs) Q&A Thread - Quick questions go here
I have seen something similar, yet different in Oceanic languages. Any negated verb will be in the irrealis form. This kind of makes sense, since we talk about an event that does not take place, even though it probably could. I have never seen mood marked on the negation particle, but it doesn'tstrike me as unnaturalistic.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
Re: (Conlangs) Q&A Thread - Quick questions go here
I think it's fine, even if I don't know of a natlang example.KaiTheHomoSapien wrote: ↑26 Sep 2020 06:31Does it make any sense to have a negative particle only for irrealis verbs?
For what it's worth, Standard Arabic has four words for 'not':
لا laa - used with indicative present, subjunctive, jussive and imperative
ما maa - used with indicative past
لم lam - used with jussive to negate a verb with past semantics (a more formal alternative to maa)
لن lan - used with indicative present to negate a verb with future semantics
I'd like to direct your attention towards 'lan', to the extent one can say verbs in a future tense are 'irrealis'. 'Irrealis' marking is not limited to subjunctive/optative kinds of things: some languages effectively mark the expression of the future in verbs as irrealis, and/or also negated verbs in general.
hīc sunt linguificēs. hēr bēoþ tungemakeras.
- DesEsseintes
- mongolian
- Posts: 4331
- Joined: 31 Mar 2013 13:16
Re: (Conlangs) Q&A Thread - Quick questions go here
لن lan is used with the subjunctive, not the indicative.
Re: (Conlangs) Q&A Thread - Quick questions go here
I'm almost certain that there are many languages that have different negation marking on irrealis verbs. And also that there are language with different modal marking on negative verbs.
EDIT: three more specific things come to mind.
First, because many languages (~10% according to WALS) mark an irrealis mood on negatives, some of those languages avoiding neutralising mood in negatives by creating a separate negation construction for irrealis verbs. Unfortunately, the only example WALS gives is Quechua, which marks all negatives as questions, and then marks real negative interrogatives differently. But I'm sure it happens with broader irrealis negatives too. [and if you don't want the irrealis marking on the verb, you could always just have it drop diachronically]
Second, it occurs to me that Proto-Indo-European is an example of this. Or Indo-European, at least. PIE is reconstructed as having two different negating particles, one specifically for imperatives. [But Tocharian uses the imperative-negator as a plain negator, so it's possible it was plain all along and became limited to imperatives in certain continguous daughters]. Again, this is an imperative specifically, but the same principle could apply to other irrealis forms, surely. [wait, this is the particle in Greek that you cite]
Third, there are two obvious ways to get your ta/te alternation. First, you could just have a negative particle fuse with a following irrealis particle (that, being itself rather light, was dropped in non-negative irrealis verbs once a more distinctive form of marking was innovated). Second, in some language the 'negative particle' is actually an auxiliary verb, and ta/te could indicate the negative verb being inflected for the irrealis mood (it could then go from being a full verb to being just a plain particle, if you wanted).
EDIT: three more specific things come to mind.
First, because many languages (~10% according to WALS) mark an irrealis mood on negatives, some of those languages avoiding neutralising mood in negatives by creating a separate negation construction for irrealis verbs. Unfortunately, the only example WALS gives is Quechua, which marks all negatives as questions, and then marks real negative interrogatives differently. But I'm sure it happens with broader irrealis negatives too. [and if you don't want the irrealis marking on the verb, you could always just have it drop diachronically]
Second, it occurs to me that Proto-Indo-European is an example of this. Or Indo-European, at least. PIE is reconstructed as having two different negating particles, one specifically for imperatives. [But Tocharian uses the imperative-negator as a plain negator, so it's possible it was plain all along and became limited to imperatives in certain continguous daughters]. Again, this is an imperative specifically, but the same principle could apply to other irrealis forms, surely. [wait, this is the particle in Greek that you cite]
Third, there are two obvious ways to get your ta/te alternation. First, you could just have a negative particle fuse with a following irrealis particle (that, being itself rather light, was dropped in non-negative irrealis verbs once a more distinctive form of marking was innovated). Second, in some language the 'negative particle' is actually an auxiliary verb, and ta/te could indicate the negative verb being inflected for the irrealis mood (it could then go from being a full verb to being just a plain particle, if you wanted).
- DesEsseintes
- mongolian
- Posts: 4331
- Joined: 31 Mar 2013 13:16
Re: (Conlangs) Q&A Thread - Quick questions go here
I have an example of a language with different negatives used according to mood: Blackfoot. These are prefixes within the verb complex so they’re not particles but it might nevertheless be of use as a reference.
máát- is used for indicative verbs.
miin- is used in the imperative.
kátá- is used for irrealis/counterfactual.
sta- is used for ‘in order not to’.
Hope that helps.
máát- is used for indicative verbs.
miin- is used in the imperative.
kátá- is used for irrealis/counterfactual.
sta- is used for ‘in order not to’.
Hope that helps.
- KaiTheHomoSapien
- greek
- Posts: 641
- Joined: 15 Feb 2016 06:10
- Location: Northern California
Re: (Conlangs) Q&A Thread - Quick questions go here
Thank you for the answers. It's been helpful
It's good to know that it isn't unheard of. My intent with making the particles similar in form is to suggest that they are etymologically related and were presumably at one point a single form. Lihmelinyan also has an etymologically unrelated prohibitive particle used with the so-called "injunctive mood" (really just a tenseless form of the verb with multiple miscellaneous functions) to form prohibitions.
It's good to know that it isn't unheard of. My intent with making the particles similar in form is to suggest that they are etymologically related and were presumably at one point a single form. Lihmelinyan also has an etymologically unrelated prohibitive particle used with the so-called "injunctive mood" (really just a tenseless form of the verb with multiple miscellaneous functions) to form prohibitions.
Re: (Conlangs) Q&A Thread - Quick questions go here
How common is /ç/?
Could it be reasonably contrasted with /x/ and/or /ʝ/?
Could it be reasonably contrasted with /x/ and/or /ʝ/?
- DesEsseintes
- mongolian
- Posts: 4331
- Joined: 31 Mar 2013 13:16
Re: (Conlangs) Q&A Thread - Quick questions go here
An /x ç/ seems perfectly plausible to me; they are very distinct sounds. In South America, Taushiro, Jaqaru, and Trinitario are listed as having both phonemes. None of those languages have a voicing contrast in fricatives, but it would be easy to envisage a language where voicing would result in /x ç/ vs /ɣ ʝ/ where /ɣ/ might then go in any of a number of directions (zero, /w/, etc.).
Re: (Conlangs) Q&A Thread - Quick questions go here
I think the Goidelic languages all contrast palatal and velar fricatives (both voiced and voiceless), but that's part of the wider palatalised vs. velarised consonant distinction found in the languages. There are some people who transcribe this as /xʲ/ (and likewise /c/ as /kʲ/, for example) presumably for this reason.
Kabyle apparently has the palatal fricatives contrasting with uvular fricatives (both voiced and voiceless), but it only contrasts velar and uvular plosives when they're labialised, i.e. there's /kʷ q qʷ/ but not /k/ (Adyghe, over in the Caucasus, coincidentally does the same thing with its plosives).
On the note of the Goidelic languages, some languages have a post-velar fricative that seems to be alternately transcribed as /ç/ or /xʲ/, but they all also seem to have a palatalisation distinction that applies to (more or less) all consonant phonemes. I wonder then, if a phonemic /ç/ is more "stable" if it's sort of "reinforced" by the rest of the phoneme inventory.
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.
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.
Re: (Conlangs) Q&A Thread - Quick questions go here
In Norwegian and German you have phonemic /ç/ vs /ʃ~ʂ/, which seems to me like an even narrower contrast than with /x/. Swedish has /ɕ/ vs /ɧ/ (which is basically /x/). In Swedish /j/ is commonly [j~ʝ~ʑ] which doesn't match entirely with /ɕ/ but comes pretty close.
- Creyeditor
- MVP
- Posts: 5091
- Joined: 14 Aug 2012 19:32
Re: (Conlangs) Q&A Thread - Quick questions go here
Some people argue that German is developing a phonemic contrast between /ç/ and /x/ based on the mininal pair Kuchen 'little cow' vs, Kuchen 'cake'. I have also seen German /j/ described as /ʝ/ but I can't recall if it was based on phonetic or phonological evidence.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
Re: (Conlangs) Q&A Thread - Quick questions go here
Yup; when it doubt, assume Irish already does it. [it also has the voiced version of this contrast]sangi39 wrote: ↑27 Sep 2020 12:52I think the Goidelic languages all contrast palatal and velar fricatives (both voiced and voiceless), but that's part of the wider palatalised vs. velarised consonant distinction found in the languages. There are some people who transcribe this as /xʲ/ (and likewise /c/ as /kʲ/, for example) presumably for this reason.
Yes, very much so.On the note of the Goidelic languages, some languages have a post-velar fricative that seems to be alternately transcribed as /ç/ or /xʲ/, but they all also seem to have a palatalisation distinction that applies to (more or less) all consonant phonemes. I wonder then, if a phonemic /ç/ is more "stable" if it's sort of "reinforced" by the rest of the phoneme inventory.
I think there's a tendency for anything between alveolar and velar to be easily interpreted as a secondary palatalisation on either a velar or an alveolar; so they tend to more stable if there's palatalisation elsewhere too.
Re: (Conlangs) Q&A Thread - Quick questions go here
Excellent, thanks for all the information guys n' gals! I was a bit worried as Wikipedia (fantastic resource I know) tells me that /ç/ is a relatively rare phoneme and contrasts between /ç/, /x/, or /ʝ/ and /ɣ/ even rarer.
Nonetheless, this exact contrast is what my Crimean Gothic reconstruct has, i.e. lachjen [ˈla.çən] (< *hlahhjan) contrasted with the first person past tense form luch [ˈluːx] (< *hlōh), contrasted with the 3rd person past plural lugen [ˈluː.ɣən]. The /ç/ phoneme here in fact might be better described as "post-palatal" perhaps.
Nonetheless, this exact contrast is what my Crimean Gothic reconstruct has, i.e. lachjen [ˈla.çən] (< *hlahhjan) contrasted with the first person past tense form luch [ˈluːx] (< *hlōh), contrasted with the 3rd person past plural lugen [ˈluː.ɣən]. The /ç/ phoneme here in fact might be better described as "post-palatal" perhaps.
- KaiTheHomoSapien
- greek
- Posts: 641
- Joined: 15 Feb 2016 06:10
- Location: Northern California
Re: (Conlangs) Q&A Thread - Quick questions go here
Yeah, /ç/ seems to be one of those that's almost always an allophone (of /h/ or /x/ usually).
Re: (Conlangs) Q&A Thread - Quick questions go here
So far as I'm aware, both these things are true. The palatal area seems particularly unstable; not only are there usually few phonemes there, but there are often breaks in series, anomalies (eg the Irish example is actually /x/ vs /ç/ vs /ɣ/ vs... /j/. Although it does go back to being a fricative next to consonants. [and apparently there's a third allophone halfway between an approximant and a fricative!?]), and even sometimes additional phonemes in the area that 'shouldn't be there (like an anomalous voiced fricative without a voiceless counterpart).
However, just because something is rare doesn't mean it can't happen! I mean, having over a dozen different vowel qualities, and phonemic triphthongs, and interdental fricatives, and more fricatives than stops is very rare indeed... but look at English!
This seems like a sensible way to derive such a contrast. However, I wouldn't expect it to be very stable in the long-run. In particular, /hj/ isn't that common in Proto-Germanic anyway. If that's the only source of the phoneme, it would appear in, what, two, three verbs only? And only in certain forms of the verb? If there's not some other source of the phoneme, I'd expect it to quickly level to the velar fricative by analogy to the non-geminating forms of the verb, or just through merger.Nonetheless, this exact contrast is what my Crimean Gothic reconstruct has, i.e. lachjen [ˈla.çən] (< *hlahhjan) contrasted with the first person past tense form luch [ˈluːx] (< *hlōh), contrasted with the 3rd person past plural lugen [ˈluː.ɣən]. The /ç/ phoneme here in fact might be better described as "post-palatal" perhaps.
[full disclosure: my own Germanic language also has this distinction, although I call [x] /h/ in theory. I mostly derive it from /h/ adjacent to /i/, including following original /ai/ (which merges with /au/ as /a:/; this and the loss of many unstressed /i/s make it phonemic). ]
Re: (Conlangs) Q&A Thread - Quick questions go here
That is very true. It seems like mergers tend to create irregularity.Salmoneus wrote: ↑27 Sep 2020 20:57So far as I'm aware, both these things are true. The palatal area seems particularly unstable; not only are there usually few phonemes there, but there are often breaks in series, anomalies (eg the Irish example is actually /x/ vs /ç/ vs /ɣ/ vs... /j/. Although it does go back to being a fricative next to consonants. [and apparently there's a third allophone halfway between an approximant and a fricative!?]), and even sometimes additional phonemes in the area that 'shouldn't be there (like an anomalous voiced fricative without a voiceless counterpart).
However, just because something is rare doesn't mean it can't happen! I mean, having over a dozen different vowel qualities, and phonemic triphthongs, and interdental fricatives, and more fricatives than stops is very rare indeed... but look at English!
You are probably correct. /h/ seems to have been lost in the actual language, with CGo. lachen being its only reflex in the corpus. One scholar theorized that the voiced velar fricative was extended to the infinitive here, i.e. [ˈla.ɣən] (perhaps initially being [ˈla.ən]) although it is also possible the language had a West Germanic style gemination before /j/ to /hh/ > /xx/. Though a phoneme from palatalization had also been suggested here.This seems like a sensible way to derive such a contrast. However, I wouldn't expect it to be very stable in the long-run. In particular, /hj/ isn't that common in Proto-Germanic anyway. If that's the only source of the phoneme, it would appear in, what, two, three verbs only? And only in certain forms of the verb? If there's not some other source of the phoneme, I'd expect it to quickly level to the velar fricative by analogy to the non-geminating forms of the verb, or just through merger.
[full disclosure: my own Germanic language also has this distinction, although I call [x] /h/ in theory. I mostly derive it from /h/ adjacent to /i/, including following original /ai/ (which merges with /au/ as /a:/; this and the loss of many unstressed /i/s make it phonemic). ]
That's fascinating!
Re: (Conlangs) Q&A Thread - Quick questions go here
What could be the origin of retroflex affricates?
In Slavic languages, they usually derive from palatals/postalveolars after new palatalization has pushed them to retroflexes.
In Swedish, retroflexes derive from /r/ + a velar, but they are not affricates.
I think English <tr> is pronounced as an affricate in some dialects.
What I'm thinking is having them without palatals contrasting with them.
In Slavic languages, they usually derive from palatals/postalveolars after new palatalization has pushed them to retroflexes.
In Swedish, retroflexes derive from /r/ + a velar, but they are not affricates.
I think English <tr> is pronounced as an affricate in some dialects.
What I'm thinking is having them without palatals contrasting with them.
My meta-thread: viewtopic.php?f=6&t=5760
- DesEsseintes
- mongolian
- Posts: 4331
- Joined: 31 Mar 2013 13:16
Re: (Conlangs) Q&A Thread - Quick questions go here
Several Athabaskan languages, including Gwich’in and Hän, had the following shift:
tʃ tʃʷ → ts tʂ
I think all of them still contrast these with tʃ from earlier *k.
- Creyeditor
- MVP
- Posts: 5091
- Joined: 14 Aug 2012 19:32
Re: (Conlangs) Q&A Thread - Quick questions go here
I think rhotics and labialization are a possible source, because all three have similar acoustic effects on the third formant.Omzinesý wrote: ↑28 Sep 2020 14:10 What could be the origin of retroflex affricates?
In Slavic languages, they usually derive from palatals/postalveolars after new palatalization has pushed them to retroflexes.
In Swedish, retroflexes derive from /r/ + a velar, but they are not affricates.
I think English <tr> is pronounced as an affricate in some dialects.
What I'm thinking is having them without palatals contrasting with them.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics
"Thoughts are free."
Produce, Analyze, Manipulate
1 2 3 4 4
Ook & Omlűt & Nautli languages & Sperenjas
Papuan languages, Morphophonology, Lexical Semantics