zompist bboard

These are always fun.

Nias has a bilabial trill and a stop with trilled release - [dr] acts as a single consonant. Also, it contrasts /v v\ w/, which you don't see very often, and its orthography has probably the only use of <ß> outside of German.
Khoekhoe has 10 pulmonic consonants and 20 clicks, and an aspiration distinction only on the velar stop.
Nivkh has a "defective" vowel system, and a process of consonant alternation where, for example, /t_h/ alternates with /r_0/.
Wari' has a voiceless dental bilabially trilled affricate (/tB\_0/ as a single consonant), /S x_w/ as the only fricatives, and five front vowels with only one back vowel. Yanesha' also has a weird inventory of fricatives, with only /B z` G/, in addition to having /l_j/ as its only lateral.
Kwaza has one front rounded vowel, /9/. Northern Yukaghir is similar, with /2/ as its only front rounded vowel.
Xavante has no velars except for /w/.
Gilbertese contrasts plain and velarized bilabials, but has no secondary articulation contrast on any other consonants.
Hixkaryana has a retroflex tap with lateral release and a vowel system of /{ e O M u/.
Marshallese has the most... interesting vowel system I've ever seen.
And, of course, Ubykh.

Anyone have more? (There are probably a few well-known ones that I forgot.)

Ubykh - and a lot of other West Caucasian languages - aren't even all that weird; they have large consonant inventories, but those a pretty well balanced and symmetrical.

Really weird inventories have weird gaps. You've mentioned some good examples.

TheGoatMan wrote:Ubykh - and a lot of other West Caucasian languages - aren't even all that weird; they have large consonant inventories, but those a pretty well balanced and symmetrical.

Yeah, its consonant inventory is pretty much entirely regular, but I figured it was worth mentioning just for the "holy shit" factor on first seeing it. Also, vertical vowel inventories aren't that common.

Really weird inventories have weird gaps. You've mentioned some good examples.

Weird gaps or weird consonants. It's just that a lot of languages with weird consonants also have weird gaps. (The only exception I can think of is Czech. Maybe Swedish and English if you stretch your definition of "weird".)

Somali has a few interesting gaps, contrasts voicing in continuants only at the pharyngeal POA and in addition has /q͡ʡ/.

I'm quite sure Arabic's /q/ is also something closer to [q͡ʡ], as it has a quality which sounds further back than uvular. When I have pronounced it as a pure /q/ my Arabic speaking friends have told me to make it 'heavier'.

I knew I was forgetting something.

Pirahã has [n] as an allophone of /g/, [B\] as an allophone of /b/, and possibly [k] as an allophone of /hi/, and can be analyzed as having either no nasals or no velars.
Central Rotokas has voice and place as the only distinguished features of consonants.
Estonian has 25 diphthongs and three levels of length on vowels and consonants.
Danish is just insane.

English is really weird it has dental fricatives~~~~

finlay wrote:English is really weird it has dental fricatives~~~~

<

That weird?

I think he was taking the piss out of what Nortaneous was saying above:

Nortaneous wrote:Maybe Swedish and English if you stretch your definition of "weird".)

There's a reason I said "if you stretch your definition of 'weird'". And what I was talking about was the rhotic, which seems to vary a lot between dialects but is usually realized in my dialect as a slightly rounded uvular approximant with retroflex, and (mostly word-initially) labiodental coarticulation. But I know that's not that weird.

Qimant has one front vowel, three central vowels, and two back vowels.
Xamtanga affricates word-medial /t q/.
Big Nambas has a series of linguolabial consonants, one prenasalized stop, /nd/, and three voiced fricatives and one voiceless one, all at different places of articulation.
Yeli Dnye has labial-alveolar, labial-postalveolar, and labial-velar stops - /tp t_-p kp/ as single phonemes.
Amis has an epiglottal trill, "releases" /s/ like a stop (adding a short, possibly voiceless, epenthetic schwa after the consonant - the plosives are the only other consonants to do this), and has a phoneme that varies dialectically between /D/, /K\/, and /d_d/.
Mapudungun has an interdental series and a central semivowel, /1_^/.
Badaga contrasts two degrees of rhoticity in all of its vowels.
Jinhai Wu has a vowel inventory of /a e ɯ ɨ i ɞ ɵ ø u y ɿ/.
Arapaho has no low vowels.
Southern Sami has only opening diphthongs.

Bumping a topic I quite like.

Nortaneous wrote:Arapaho has no low vowels.

• And St'at'imcets has no close vowels. It also contrasts pharyngeals, pharyngeals with glottal stop clusters, and glottalized pharyngeals, among other things.
• Aleut has natively no bilabial obstruents, as well as no glottal stop (thereby making an exception of the "/p t k ʔ/" rule). It does have a phoneme that varies as [v] ~ [θ] ~ [z] between dialects. There is also a voicing contrast in both nasals and fricativs without there (natively) also being one in stops.
• Jabo has four level tones and all possible contour tones that can be combined from these.
• Maxakalí arguably has only obstruent consonants.
• N|u has consonants such as a palato-uvular affricate /cχ/, and contrasts velaric and uvularic clicks, but lacks a /t/ (no, it doesn't have a dental or ejective or aspirated one either).
• Funky initial (yes initial, they claim; it would be interesting even as a final inventory tho) cluster assortment from Bonan: /mp nt ntɕ ntʂ ŋk tʰχ χtɕ rtɕ ltɕ ft fk ʂp ʂk/
• Toda has pretty much every goddamn coronal fricative there is (even a retroflex lateral) (conlangers take note: and it still doesn't have palatal fricativs), plus a semi-crapshoot vowel system /i y ɨ u e ɵ o æː ɑ/.
• Nobody appears to have told Buyang that phonation, POA and secondary articulation tend to be orthogonal dimensions.
• Dahalo is pretty much my definition of "nuts". ANADEW for "kitchen sinky". (Particularly out of place: random labialized alveolars here and there.)
• Finally, Hmong.

Tahltan has an absurdly high affricate inventory and an equally absurd fricative inventory.
Serer has voiceless implosives.
Yagaria, if you trust UPSID, has /L\/ as its only lateral.
Golin has no tense vowels.
Sicilian shifted /l:/ to /d`:/, and had a few other weird sound changes.
And how did I forget Shona and its whistled sibilants?

Tropylium wrote:• Nobody appears to have told Buyang that phonation, POA and secondary articulation tend to be orthogonal dimensions.

holy shit I never thought I'd see a sound inventory that has a bit that approximates the sierpinski triangle

And what's up with Kazakh vowels?

http://en.wikipedia.org/wiki/Kazakh_language#Phonology

Additionally, every source appears to give different values for the vowels.

That vowel system reminds me of Faroese.

I think Faroese's less strange than that... Kazakh on Wikipedia is something like


+ the diphthongs iə, yʉ, uʊ

OTOH other sites disagree so it might be wrong. That / ɘ/ seems suspect to me, here it's /I/ instead.

This thread should be called "Weird natlang inventories".

Piotr wrote:I think Faroese's less strange than that... Kazakh on Wikipedia is something like

Code: Select all

   ʉ
     ʊ
   ɘ 
   ə
æ     ɑ

+ the diphthongs iə, yʉ, uʊ

OTOH other sites disagree so it might be wrong. That / ɘ/ seems suspect to me, here it's /I/ instead.

The WP page neglects to inform that the sequences /iɘj ɘj/ are pronounced [i:], and /yʊw ʉʊw ʉw ʊw/ [u:].

Also I suppose /ʉ/ would be [ʊ̶] rather than [ʉ], and /ɘ/ something like /ɪ_-/.

The IPA just doesn't really suit central Asian languages, it seems.

Tarasoriku wrote:This thread should be called "Weird natlang inventories".

I'm pretty sure nobody would complain of being off topic if you brought some bizarre harmony system or historical development to the table; it's just that inventories are apparently easier to find (and notice as being weird on a glance).

Weird harmony systems?

Also: "Some Finnish speakers find it hard to pronounce both 'b' and 'p' in foreign words (e.g. pubi), so they voice (bubi) or devoice (pupi) the entire word." (from Wikipedia)

Tropylium wrote:I'm pretty sure nobody would complain of being off topic if you brought some bizarre harmony system or historical development to the table; it's just that inventories are apparently easier to find (and notice as being weird on a glance).

Yeah, someone should post that fucking Armenian numeral thing that people always mention for weird historical developments.

Nortaneous wrote:Also: "Some Finnish speakers find it hard to pronounce both 'b' and 'p' in foreign words (e.g. pubi), so they voice (bubi) or devoice (pupi) the entire word." (from Wikipedia)

Ah yep. It gets even worse with more complicated loanwords. "Anegdotes" are commonplace, "gebards" are known for their greit speed, "gebab" is occasionally served, a "pekonia" (funny since it's also the partitiv of "bacon") may be grown on one's windowstill, and even the hypercorrect "Goga Gola" has been attested at least once. One case has even made it into standard Finnish: "biisi" is a loan from "piece of music".

There's also the joke that in theology conferences, everyone in the audience must try to held back laughter whenever Finns have something to say on the topic of "baconism" ([peɪkənisəm]). But yeah, THIS may now be getting off topic.

Nortaneous wrote:Danish is just insane.

GFMFT.

Oh, also, if I may, I'm adding English under the category, Also Insane.

As for weird historical developments, I've just read somewhere on this board that the whistled sibilants in Shona are hypothesized to come from Proto-Bantu mere bilabial stops.

Tropylium wrote: Ah yep. It gets even worse with more complicated loanwords. "Anegdotes" are commonplace,

Pronouncing a voiceless velar in this word would actually seem a bit !Xoo-ish to me; luckily, we don't have to...

Nortaneous wrote:Southern Sami has only opening diphthongs.

All Sami languages have only opening diphthongs - if you don't count clusters of vowel + semivowel as diphthongs. Personally, though, I would.

Tropylium wrote:One case has even made it into standard Finnish: "biisi" is a loan from "piece of music".

I'm afraid that's still very much a slang word; standard would be kappale. But yeah, Finnish (and especially Helsinki area) slang is full of this stuff. I don't think it's all hypercorrection, though; it seems adding prosthetic s's and voicing stops is simply a way of making a word sound "slangy". My favorite example is skeba, "guitar", which appears to be derived from Finnish keppi, "stick".

Angami has actual aspirated nasals (and an aspirated lateral), and labiodental allophones of labial and labialized consonants before /@/. (yes that means [F_0_h])

More weird historical developments, from this page:

The most remarkable claim about the history of Basque that one will encounter in the present work is made by Michelena himself - an apparent reversal of the merger of j and x in Gipuzkoan Basque. It is claimed that, as in Navarre Basque, these phonemes merged as x, but later in Gipuzkoan etymological j shifted to a velar while original x remained palato-alveolar. This would constitute a violation of the logic of sound change as we understand it, and we might conclude that there may not have been a phonological merger at all. However, Michelena has an explanation and this relates to the function of palatals being associated with expressive meaning in Basque. The hypothesis is that in words so marked, x simply failed to shift, possibly a beautiful example of functional pressure on a sound system.

so basically: /j/ -> /S/, then /S/ -> /j/ except in diminutives(?), and since /S/ originally mostly appeared in diminutives, it looked like the change reversed itself