Languages with interesting phonotactics

A forum for discussing linguistics or just languages in general.
User avatar
k1234567890y
mayan
mayan
Posts: 2368
Joined: 04 Jan 2014 04:47
Contact:

Re: Languages with interesting phonotactics

Post by k1234567890y »

Zekoslav wrote: 05 Oct 2019 10:01
Allowing /s/ + stop clusters to begin a syllable is actually one of Indo-European peculiarities, since these clusters also go against the sonority hierarchy. In that context Western Romance languages which put a vowel before these clusters are actually making things more normal!
lol sounds interesting, thanks for telling (:
Ser wrote: 04 Oct 2019 23:46
k1234567890y wrote: 04 Oct 2019 17:36In Old Church Slavonic, all syllables must end in a vowel, while there are consonant clusters. So words like kostь "bone(nom.sg)" is rendered as ko-stь instead of kos-tь
Well, lots of languages are also commonly analyzed with their [st] medial clusters analyzed as onsets, because they allow initial st-, e.g. this is how Italian is usually analyzed. The fun thing about Old Church Slavonic is that even clusters with consonants of the same MOA like [gd] or that go against the sonority hierarchy like [ʒd] as in къгда kŭgda 'when' and дъждь dŭždĭ 'rain' are best analyzed as [kʊ.gda] and [dʊ.ʒdɪ].

This gives Old Church Slavonic a very odd character with mostly CCV and CV syllables, plus the occasional CCCV syllable (where the last C is a [v] or [j] glide):

нѣколико [næ.ko.li.ko] 'some'
срьдьце [srɪ.dɪ.tse] 'heart'
оучител҄ь [u.tʃi.te.lʲɪ] 'teacher'
дльгъ [dlɪ.gʊ] 'long'
чловѣкъ [tʃlo.væ.kʊ] 'person'
оударити [u.da.ri.ti] 'to hit'
хвостъ [xvo.stʊ] 'tail'
хлѣбъ [xlæ.bʊ] 'bread'
женихъ [ʒe.ni.xʊ] '(bride)groom'
стоуденъ [stu.de.nʊ] 'cold'
змиꙗ [zmi.ja] 'snake'
гвоздь [gvo.zdɪ] 'nail (to build things with)'
костьѭ [ko.stɪ.jõ] 'with a/the bone'
гоубител҄ьство [gu.bi.te.lʲɪ.stvo] 'destruction'
землꙗ [ze.mlja] 'soil'
you are right, thanks for telling (:
I prefer to not be referred to with masculine pronouns and nouns such as “he/him/his”.
Nortaneous
greek
greek
Posts: 631
Joined: 14 Aug 2010 13:28

Re: Languages with interesting phonotactics

Post by Nortaneous »

Zekoslav wrote: 05 Oct 2019 10:01 Allowing /s/ + stop clusters to begin a syllable is actually one of Indo-European peculiarities, since these clusters also go against the sonority hierarchy. In that context Western Romance languages which put a vowel before these clusters are actually making things more normal!
It's not that weird -- in many languages in the Tibetic linguistic area, for example, every cluster that isn't of the form CR violates the sonority hierarchy! Typically you have (sonority-violating) FC NC and (sonority-compliant) CR, e.g. /sp- mp- pr-/, but you don't have (sonority-compliant) PF PN FN etc.

My guess is that there are more languages that allow FP- but prohibit PF- than vice versa. This might also hold for NP- and PN-, but NP- can often be analyzed as units. (Of course, not all languages necessarily have grounds for distinguishing between consonants and clusters...)
Solarius
roman
roman
Posts: 1112
Joined: 30 Aug 2010 01:23

Re: Languages with interesting phonotactics

Post by Solarius »

IIRC there's a Qiangic language which recently underwent monosyllabicization, but in such a way that the permitted clusters were identical in onset and coda--i.e. [ClaCl] was valid but [ClalC] wasn't.
Check out Ussaria!
Nortaneous
greek
greek
Posts: 631
Joined: 14 Aug 2010 13:28

Re: Languages with interesting phonotactics

Post by Nortaneous »

Solarius wrote: 22 Oct 2019 21:49 IIRC there's a Qiangic language which recently underwent monosyllabicization, but in such a way that the permitted clusters were identical in onset and coda--i.e. [ClaCl] was valid but [ClalC] wasn't.
Ronghong Qiang, described in LaPolla's grammar
User avatar
Omzinesý
runic
runic
Posts: 3053
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Languages with interesting phonotactics

Post by Omzinesý »

Northern Saami has quite few CVCV words but very often a geminate.
Finnish kota Northern Saami koɑhti - koɑʰtti

Estonian word strength is also interesting:
linn 'town'
liin 'linen'
lina '?' (at least allowed)
# lin (too short a word to be allowed)
Both vowels and consonants can be of three different lengths.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3053
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Languages with interesting phonotactics

Post by Omzinesý »

Nortaneous wrote: 22 Oct 2019 15:42
Zekoslav wrote: 05 Oct 2019 10:01 Allowing /s/ + stop clusters to begin a syllable is actually one of Indo-European peculiarities, since these clusters also go against the sonority hierarchy. In that context Western Romance languages which put a vowel before these clusters are actually making things more normal!
It's not that weird -- in many languages in the Tibetic linguistic area, for example, every cluster that isn't of the form CR violates the sonority hierarchy! Typically you have (sonority-violating) FC NC and (sonority-compliant) CR, e.g. /sp- mp- pr-/, but you don't have (sonority-compliant) PF PN FN etc.

My guess is that there are more languages that allow FP- but prohibit PF- than vice versa. This might also hold for NP- and PN-, but NP- can often be analyzed as units. (Of course, not all languages necessarily have grounds for distinguishing between consonants and clusters...)
Greek and Latin had /ps/ and /ks/, not /bz/ or /gz/. No language with many plosive + fricative clusters comes to my mind.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
qwed117
mongolian
mongolian
Posts: 3921
Joined: 20 Nov 2014 02:27

Re: Languages with interesting phonotactics

Post by qwed117 »

Omzinesý wrote: 08 Apr 2021 09:12
Nortaneous wrote: 22 Oct 2019 15:42
Zekoslav wrote: 05 Oct 2019 10:01 Allowing /s/ + stop clusters to begin a syllable is actually one of Indo-European peculiarities, since these clusters also go against the sonority hierarchy. In that context Western Romance languages which put a vowel before these clusters are actually making things more normal!
It's not that weird -- in many languages in the Tibetic linguistic area, for example, every cluster that isn't of the form CR violates the sonority hierarchy! Typically you have (sonority-violating) FC NC and (sonority-compliant) CR, e.g. /sp- mp- pr-/, but you don't have (sonority-compliant) PF PN FN etc.

My guess is that there are more languages that allow FP- but prohibit PF- than vice versa. This might also hold for NP- and PN-, but NP- can often be analyzed as units. (Of course, not all languages necessarily have grounds for distinguishing between consonants and clusters...)
Greek and Latin had /ps/ and /ks/, not /bz/ or /gz/. No language with many plosive + fricative clusters comes to my mind.
Should be noted that Greek nor Latin had a plain /z/ (Modern Greek's /z/ descends from either /zd/ or /dz/- personally, I prefer zdeta, but the fact that tsan exists as well makes zdeta weird in the larger scheme of things)
Spoiler:
My minicity is Zyphrazia and Novland
What is made of man will crumble away.

The SqwedgePad
User avatar
Ahzoh
mongolian
mongolian
Posts: 4087
Joined: 20 Oct 2013 02:57
Location: Canada

Re: Languages with interesting phonotactics

Post by Ahzoh »

Pabappa wrote: 30 Jul 2019 14:34 One thing I notice about Semitic languages is that they don't seem to have a lot of homorganic nasal+stop clusters. /mb mp nt nd/ etc.
Assimilation involving nasals and non-nasals is a common occurrence in Semitic languages. They also strongly dislike certain homoorganic clusters in the first and second radicals, but don't much mind them in the second and third. There are no roots like B-M-T or G-K-L or S-S-B but you'll see lots of M-K-K and D-M-M
Image Śād Warḫālali (Vrkhazhian) [ WIKI | CWS ]
User avatar
LinguoFranco
greek
greek
Posts: 504
Joined: 20 Jul 2016 17:49
Location: U.S.

Re: Languages with interesting phonotactics

Post by LinguoFranco »

I think Japanese can have some interesting phonotactics. On the surface, it appears largely CV/CVN, but it also has lots of devoiced vowels, so you get words that kinda sound like they have clusters to those unfamiliar with the language.
User avatar
Ahzoh
mongolian
mongolian
Posts: 4087
Joined: 20 Oct 2013 02:57
Location: Canada

Re: Languages with interesting phonotactics

Post by Ahzoh »

LinguoFranco wrote: 03 May 2021 23:05 I think Japanese can have some interesting phonotactics. On the surface, it appears largely CV/CVN, but it also has lots of devoiced vowels, so you get words that kinda sound like they have clusters to those unfamiliar with the language.
The devoiced vowels are more likely elided vowels, so it's even more likely they have true clusters. As far as I'm aware, devoiced/voiceless vowels is physiologically impossible.
Image Śād Warḫālali (Vrkhazhian) [ WIKI | CWS ]
User avatar
Creyeditor
MVP
MVP
Posts: 4292
Joined: 14 Aug 2012 19:32

Re: Languages with interesting phonotactics

Post by Creyeditor »

Ahzoh wrote: 04 May 2021 08:18 As far as I'm aware, devoiced/voiceless vowels is physiologically impossible.
Why should that be?
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: 2319
Joined: 19 Sep 2011 19:37

Re: Languages with interesting phonotactics

Post by Salmoneus »

Whisper the words "fat feet". If you can tell the two words apart, then no, voiceless vowels are not impossible.
User avatar
Ahzoh
mongolian
mongolian
Posts: 4087
Joined: 20 Oct 2013 02:57
Location: Canada

Re: Languages with interesting phonotactics

Post by Ahzoh »

Whispering [some level of (supra)laryngeal turbulence/vibration] isn't the same as actual voicelessness [a complete lack of (supra)laryngeal turbulence/vibration], in fact you can even phonetically distinguish whispering from voicelessness. A whispered [g] from a voiceless [k].

There's a number of languages (like Comanche) claimed to have voiceless vowels but they never occur phonemically and they always occur between consonants, never in isolation. The only proof I've seen is that the lips round during segments containing "devoiced rounded back vowels", but that could just be allophonic spreading of features.

But I remember reading something, a paper or conversation on CBB/ZBB about voiceless vowels not occurring in Japanese, but that it's simple elision.
Image Śād Warḫālali (Vrkhazhian) [ WIKI | CWS ]
User avatar
Vlürch
sinic
sinic
Posts: 390
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: Languages with interesting phonotactics

Post by Vlürch »

Japanese devoicing/dropping probably varies a lot, I think? No idea what the rules are for when there can be full dropping and when there can't, but it seems like it happens more with verbs than nouns? Maybe?

I'm pretty sure I've heard eg. 分かりました (wakarimashita, "understood") with full dropping as [ɰᵝa̠ka̠ɾʲima̠ɕt̪a̠], but I don't think I've heard eg. (shita, "tongue") as [ɕt̪a̠] but only as [ɕi̥t̪a̠] with just devoicing rather than full dropping, although I do think I've heard eg. 明日 (ashita, "tomorrow") with full dropping as [a̠ɕt̪a̠] sometimes, so it's probably not as simple as it being an issue of whether it's a noun or not...?🤔

But well, sometimes they're not devoiced at all, like, I'm pretty sure you shouldn't devoice eg. 死体 (shitai, "dead body"), and Wiktionary doesn't indicate it either, so... I have no idea why that is, but I'm not even close to fluent in Japanese; it could be that the explanation is something that's obvious to native speakers and maybe there are very clear rules to all of this. Dunno. My ears are by no means perfect either, so... uh... take this with a grain of salt.
User avatar
Creyeditor
MVP
MVP
Posts: 4292
Joined: 14 Aug 2012 19:32

Re: Languages with interesting phonotactics

Post by Creyeditor »

Ahzoh wrote: 04 May 2021 19:58 Whispering [some level of (supra)laryngeal turbulence/vibration] isn't the same as actual voicelessness [a complete lack of (supra)laryngeal turbulence/vibration], in fact you can even phonetically distinguish whispering from voicelessness. A whispered [g] from a voiceless [k].

There's a number of languages (like Comanche) claimed to have voiceless vowels but they never occur phonemically and they always occur between consonants, never in isolation. The only proof I've seen is that the lips round during segments containing "devoiced rounded back vowels", but that could just be allophonic spreading of features.

But I remember reading something, a paper or conversation on CBB/ZBB about voiceless vowels not occurring in Japanese, but that it's simple elision.
It is true that whispery voice is different from voicelessness (at least articulatorily). The reason why you can easily distinguish voicing in whispered speech is that there are additional acoustic cues for voiced consonants (e.g. in English vowels are longer before voiced consonants). I am also not an expert on Japanese voiceless vowels. Voiceless vowels might or might not exist phonemically.

Phonetically, voiceless/devoiced vowels are plausible both articulatorily and acoustically, all else being equal. Articulatorily, the same gesture can be perfomed that is performed for a vowel with modal voice. The only difference is that the state of the vocal folds is changed to let more air pass through. Acoustically, voicing in continious sounds is gradient, not binary. In the acoustic signal, the energy concentrated in f0 distinguished between voiced and voiceless sounds. A devoiced vowel will thus have more noise and a less clear formant structure. This includes having less energy concentrated in the fundamental frequency. I have seen such vowels in Praat and I can assure you they exist in many languages.

Nevertheless, it is of course possible that in Japanese these vowels are actually elided. I have heard several times of sounds that are neither detectable in the acoustic signal, nor do they correlate with a certain articulatory gesture, yet speakers have a cognitive category for them and are convinced that they are pronounced. Of the top of my head I can only remember a lecture note on German coda laterals in certain contexts, but this might also be true for Japanese. I also heard the term Illusory Consonants used in similar contexts.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
Pabappa
greek
greek
Posts: 484
Joined: 18 Nov 2017 02:41
Contact:

Re: Languages with interesting phonotactics

Post by Pabappa »

Finnish has some words like helppo, humppa, olifantti, kurssi, etc with a sonorant followed by a geminate consonant. Does anyone know if there are examples of this type of cluster in native vocabulary? Some of these words are quite old, but are still originally loans. I had thought I knew of more, but my memory was faulty .... for example I thought the word for beach had /ntt/ but it's just /nt/.

But if my impression is right, the pattern requires a sonorant at the beginning, and then a geminate consonant that is one of /p t k s/, which could either be an artifact of the source languages or some constraint within Uralic that also applied to much rarer native words of the same type.

Also do we know if this cluster pattern is a native development within Uralic .... even if it is found only in loans ..... and if it has influenced or been influenced by the somewhat similar situation in Latvian & Lithuanian where sequences like /ar/, /in/, /um/ etc behave as if they were traditional diphthongs rather than a cluster of a vowel and a consonant?
I'll take the theses, and you can have the thoses.
User avatar
Vlürch
sinic
sinic
Posts: 390
Joined: 09 Mar 2016 21:19
Location: Finland
Contact:

Re: Languages with interesting phonotactics

Post by Vlürch »

Pabappa wrote: 05 May 2021 14:06Does anyone know if there are examples of this type of cluster in native vocabulary?
There's tarttua ("to grab") that's been reconstructed as *tarttudak in Proto-Finnic, and possibly goes back to Proto-Uralic since Hungarian has tart ("to keep"). Also varttua ("to mature"), but apparently that's derived from varsi ("stalk") so it probably doesn't count even if tarttua counts.

Depending on what you want to believe, myrkky ("poison") might count as a native word since apparently Hungarian has méreg of the same meaning that seems like an obvious cognate, but on the other hand according to Wiktionary the Finnish term is alternatively considered a Germanic loanword from a word meaning "celery" somehow. To be honest, I always assumed it was a straight-up Swedish loanword, as in I was sure the Swedish the word for poison was *myrk and had been living my life under the assumption that it was one of the few Swedish words I knew, but turns out it isn't.🤯

Trying to find the etymology of herkku ("delicacy") and herkkä ("fragile") started leading me down a rabbit hole that I just wanted to get out of because it seems endless and headache-inducing, but I guess they're almost certainly loanwords from somewhere and are etymologically connected not only to each other (which I'd never realised before, although "delicacy" and "delicate" in English obviously are but that hadn't hit me before either) but possibly also all kinds of other words. [O.O]

I can't find any information on the etymology of pinkka ("pile"), but it sounds like a Swedish loanword so that's what it probably is?

There are probably some more words like that where dispute regarding their etymologies exists, but those are the only ones I can think of off the top of my head that didn't turn out to be absolutely 100% certified loanwords.
clawgrip
MVP
MVP
Posts: 2241
Joined: 24 Jun 2012 07:33
Location: Tokyo

Re: Languages with interesting phonotactics

Post by clawgrip »

Generally, Japanese vowels /i/ and /u/ only are devoiced/dropped between two voiceless consonants, or between a voiceless consonant and the end of a word. Also, two vowels cannot both be devoiced in a CVCV sequence.

The reason that it's somewhat difficult to determine if a vowel is devoiced or dropped entirely, and also the reason that these two vowels can be dropped without ambiguity, is because they both induce secondary articulation on the preceding consonant, which is present even if the vowel is not. /u/ [ɯᵝ] causes lip compression, and /i/ causes palatalization (e.g. /hu/ and /hi/ are realized as [ɸɯᵝ] and [çi]).

This means that a word like kikyō is pronounced something like [kʲkjoː], while kukyō is [kᵝkjoː]. Something like this, anyway. So even without the vowel, the secondary articulation disambiguates. None of the other three vowels induce secondary articulation, and so none of them can be devoiced/dropped.

This is also why many Japanese speakers have trouble pronouncing English words such as "who" and "see", as they often have difficulty identifying and eliminating these secondary articulations in their own speech.
User avatar
LinguistCat
sinic
sinic
Posts: 246
Joined: 06 May 2017 07:48

Re: Languages with interesting phonotactics

Post by LinguistCat »

I've also noticed that devoiced vowels don't occur when the vowel is accented (directly before the pitch drop), which makes sense considering that would make it more marked than other vowels in the word.
User avatar
Omzinesý
runic
runic
Posts: 3053
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Languages with interesting phonotactics

Post by Omzinesý »

Pabappa wrote: 05 May 2021 14:06 Finnish has some words like helppo, humppa, olifantti, kurssi, etc with a sonorant followed by a geminate consonant. Does anyone know if there are examples of this type of cluster in native vocabulary? Some of these words are quite old, but are still originally loans. I had thought I knew of more, but my memory was faulty .... for example I thought the word for beach had /ntt/ but it's just /nt/.

But if my impression is right, the pattern requires a sonorant at the beginning, and then a geminate consonant that is one of /p t k s/, which could either be an artifact of the source languages or some constraint within Uralic that also applied to much rarer native words of the same type.

Also do we know if this cluster pattern is a native development within Uralic .... even if it is found only in loans ..... and if it has influenced or been influenced by the somewhat similar situation in Latvian & Lithuanian where sequences like /ar/, /in/, /um/ etc behave as if they were traditional diphthongs rather than a cluster of a vowel and a consonant?
Those clusters is what you first check to see if a word is a loan.
/nd/ in the loan-giver becomes/nt/ in Finnish and /nt/ in the loan-giver becomes /ntt/ in Finnish.

Older native words that had such clusters were simplified.
tun(ne)ttu => tuttu 'known' => 'friend'

Germanic *xelp- has been borrowed three times. You can see that the Proto-Germanic loan lacks the geminate.
kelpo 'proper' PG
helppo 'easy' Old Norse(?)
jelppaa 'helps' (slang) Swedish


But there is sanko ~ sankko 'bucket' that has alteration in Finnish. It derives from sanka 'hande (of bucket)'.

Borrowing to both directions between Finnish and Saami has also been intense, like Ánte has noted. So a Uralic wotd can be a loan in Finnish.
tanner '(battle) field' Uralic
tunturi 'small mountain' Saami
tundra from Russian that got it from some Uralic language.

But generally, they belong to nativized vocabulary and are synchonically not different from the others.
My meta-thread: viewtopic.php?f=6&t=5760
Post Reply