Phonemes which are found in <5 languages or so

[alice]

Two candidates come to mind: Czech ř and the paperclip sj in Swedish. Any others?

[Vijay]

Malayalam has both /ɕ/ AND /ʃ/ (definitely NOT */ʂ/). How common is that?

Oh, also, most of the trills in Toda and [ɭ̊˔].

[Sumelic]

Two candidates come to mind: Czech ř and the paperclip sj in Swedish. Any others?

What does "paperclip" mean in "the paperclip sj"? Is it a description of the shape of the IPA symbol <ɧ>? My impression was that this phoneme gets a dedicated IPA symbol more because there is no consensus on the correct phonetic description than because it is thought to represent a particular phone that occurs uniquely in Swedish. It seems kind of like if we just had a symbol that meant "the 'r' sound in English" rather than using broad-transcription /r/, /ɹ/ or /ɻ/ to represent a phoneme that can have many places of articulation and secondary articulation (as outlined in the Wikipedia article Pronunciation of English /r/). Actually, the English "r" sound is relatively rare; although I would guess it occurs in more than 4 other languages, the only one I can think of at the moment is Chinese (approximately? I believe onset /r/ in Chinese can be more fricative-y than English /r/, but coda /r/, in the accents where it occurs, is as far as I know not realized as a fricative in Chinese)

[t̪ʙ̥] seems rare.

[Grunnen]

Two candidates come to mind: Czech ř and the paperclip sj in Swedish. Any others?

What does "paperclip" mean in "the paperclip sj"? Is it a description of the shape of the IPA symbol <ɧ>? My impression was that this phoneme gets a dedicated IPA symbol more because there is no consensus on the correct phonetic description than because it is thought to represent a particular phone that occurs uniquely in Swedish. It seems kind of like if we just had a symbol that meant "the 'r' sound in English" rather than using broad-transcription /r/, /ɹ/ or /ɻ/ to represent a phoneme that can have many places of articulation and secondary articulation (as outlined in the Wikipedia article Pronunciation of English /r/). Actually, the English "r" sound is relatively rare; although I would guess it occurs in more than 4 other languages, the only one I can think of at the moment is Chinese (approximately? I believe onset /r/ in Chinese can be more fricative-y than English /r/, but coda /r/, in the accents where it occurs, is as far as I know not realized as a fricative in Chinese)

[t̪ʙ̥] seems rare.

The city dialect of Leiden in the Netherlands is locally famous for having an /r/ that sounds more American than like anything else. Also a great many speakers of Dutch use an approximant /r/ in coda position. And some time ago this recording was posted somewhere of a Himalayan (I think) language that sounds a lot like American English, especially regarding the /r/. I can't seem to find the recording right now.

Maybe some clicks in southern Africa and some ejectives in the PNW would make it into this category?

[gach]

And some time ago this recording was posted somewhere of a Himalayan (I think) language that sounds a lot like American English, especially regarding the /r/. I can't seem to find the recording right now.

That would be Mongsen Ao, fascinatingly closely related to Sangtam which has the dental-labial trilled affricates.

[Vijay]

And some time ago this recording was posted somewhere of a Himalayan (I think) language that sounds a lot like American English, especially regarding the /r/. I can't seem to find the recording right now.

That would be Mongsen Ao, fascinatingly closely related to Sangtam which has the dental-labial trilled affricates.

Don't most of the languages in that region sound a lot like American English, especially regarding the rhotic?

[Zaarin]

Maybe some clicks in southern Africa and some ejectives in the PNW would make it into this category?

/xʼʷ χʼʷ/ are believed to be found only in Tlingit.

[gach]

Don't most of the languages in that region sound a lot like American English, especially regarding the rhotic?

They do a bit. Aspirated stops and rhotic approximants aren't exactly rare in the area surrounding the Himalayas. Mongsen Ao's /ʉ/ is a nice addition to that.

[M Mira]

Staying on the topic of rhotics: how about /ɚ/ and /ɝ/?

[mèþru]

I'm pretty sure rhotic vowels have separately evolved many times.

[Frislander]

I think the velar laterals of some languages of the Chimbu-Wahgi group might count, though given that some of those haven't got that much documentation (indeed I'm not sure how many re that well-known other than Melpa), so that may push the total up to above five. In the same area there's also Archi, which I believe is the only language with velar-lateral ejective affricates, and certainly the only one with labialised versions of them I'm sure.

Also whatever the fuck's going on with Toda, that palatalised retroflex trill is unique to that language I'm sure, and that lateral retroflex fricative.

[Vijay]

Yep, that's why I mentioned Toda earlier! It's the loveable oddball of the Dravidian (language) family.

[Tropylium]

Obligatory: labioalveolars such as /n͡m/ in Yélî Dnye.

Some polyphthongs I imagine it would be a bit hard to find parallels for:
– /oæ/ in Southern Sami (contrasts with /oa/ and /oe/)
– /ɶæ/ in Eastern Mansi (contrasts with /æː/, unclear if contrasts with short [ɶ])
– /uːo̯i̯/ in Livonian (contrasts with /u̯oi̯/)

[Qwynegold]

My impression was that this phoneme gets a dedicated IPA symbol more because there is no consensus on the correct phonetic description

Yes, that's correct.

than because it is thought to represent a particular phone that occurs uniquely in Swedish.

No, it is pretty unique to Swedish. I've once heard a claim that there's a dialect of German that also has it, but I don't know. Anyhow, the matter is further complicated by that it's pronounced differently in different dialects of Swedish. Here's an explanation.

[Qwynegold]

Malayalam has both /ɕ/ AND /ʃ/ (definitely NOT */ʂ/). How common is that?

At least some dialects of Swedish have both.

There was some interdental consonant transcribed as h̪͆ (or was it x̪͆?) in some now extinct (?) language, but I can't remember which. Does anyone recollect this?

[Vijay]

Malayalam has both /ɕ/ AND /ʃ/ (definitely NOT */ʂ/). How common is that?

At least some dialects of Swedish have both.

Really? Because the Wikipedia article on Swedish phonology seems to say that the varieties in southern Sweden have /x/ and /ɕ/, the varieties in northern Sweden have /ʂ/ and /ɕ/, and the ones in Finland have something like /ʃ/ or /ɕ/ and /t͡ɕ/ or /t͡ʃ/.

[mèþru]

There was some interdental consonant transcribed as h̪͆ (or was it x̪͆?) in some now extinct (?) language, but I can't remember which. Does anyone recollect this?

There's the voiceless bidental consonant /h̪͆/, which is transcribed /x̪͆/ in the context of the only language it is known to exist in: the Black Sea dialect of Shapsug (which is a western/Adyghe variety of the Adyghe-Kabardian continuum). I don't know whether it is alive or extinct.

[finlay]

Always thought the swedish sound sounded like /W/, which is rare enough itself and dying out in english

[Frislander]

There was some interdental consonant transcribed as h̪͆ (or was it x̪͆?) in some now extinct (?) language, but I can't remember which. Does anyone recollect this?

There's the voiceless bidental consonant /h̪͆/, which is transcribed /x̪͆/ in the context of the only language it is known to exist in: the Black Sea dialect of Shapsug (which is a western/Adyghe variety of the Adyghe-Kabardian continuum). I don't know whether it is alive or extinct.

I'm pretty sure it's still alive, though likely endangered: the only Northwest Caucasian language which has become extinct during recorded history so far is Ubykh.

Also from the same dialect continuum, the Abzakh dialect has a palatalised glottal stop /ʔʲ/, and most varieties also feature a labialised glottal stop /ʔʷ/. In the other branch of the family, Abazgi, bothe Abkhaz and Abaza feature a palatalised uvular ejective /qʲ’/ and the palatalised uvular fricative /χʲ/, and the Bzyp Abkhaz dialect features the pharyngealised uvulars /χˤ χˤʷ/ (all of which and more were to be found in Ubykh).

[Yaali Annar]

Phonemes with exactly 4 occurences in Phoible:

/'n̪̥ ɛu ye e̞u kǁ͓ n̠ʒ ʈʼ eo χʲ yə kpʲ β̃ eə k̰ʷ s̪ˤ jɛ t̪s̪ʲ ɾ̤ ɲc t̠ʃʰʲ ʉ̃ tːˤ ø̞ː we ɤ̟ ʂʰ v̤ uai ɤ̤ t̠ʃ͉ lʱ uʌ ɹ̩ ps ɐ̹̆ ɪ̆ t̪θʰ n̪ʲ|nʲ ʊ̜ d̠ʒʰ ə̰ uɪ ɟ̟ ʈːʰ ɐu ɛ̝ i̞ː ɮʲ ɯi qːʷ ɟʷ k̟ æi wõː ɡ̰ rʱ ae̞ ɾʷ kǂ ðː ao̞ ɤ̞̃ ɟʝʷ xʼ õ̰ n̠d̠ aɯ ɛɪ zːˤ dzʱ iau kpʰ xk rˤ ʃ͉ ʁː jɔ ʃ̺ r̠ ṵ̃ d̤̪|d̤ ʌ̤ z̥ ɴq ø̃ː ɨə dɮ ɥ̃ woː s̪ː|sː z̪ˤ βʷ ɨa nɡ t̺s̺ː bˠ ɠɓ aːˤ ɾ̥ʲ z̪ʲ ðʲ oˤ ɲ̰ mˠ kǃ̠ r̝ dːˤ ä̠ d̻z̻ wi ʊ̤ ju ħʷ çː sːˤ ɻʲ ɟʱ ˩˧ tɬːʼ n̪t̪ ɔ̝ w̤ ɡǁ ë t̪s ʡ'/

Phonemes with exactly 3 occurences in Phoible:

/'ŋǁ͓ ɻʷ jẽ ɳʈ s̻ː mʰ ɔ̆ ɛo õ̞ː dz̺ ʼm u̜ uːˤ i̤ˑ d̤ʒ̤ʱ kǀʼ ɔ̤ˑ iˠ çʷ ɡǂ ya ç̟ lʼ ɳʈʂ tsʷʼ ʊa ɡ̤ː iæ ʊ̈ ɡ̟ ts̪ t̪s̪ʰʷ|tsʰʷ t̪ɬ̪ʰ|tɬʰ ɔ̞ æ̰ dˠ s̪ʷ|sʷ tˀ nɗ χˤ ɑʊ ɔe ɮ̪|ɮ kxʼ kǃx yø t̪s̪ʷʼ|tsʷʼ ʰpː ɒ̃ː dʼ æɪ̯ ⁿɢ tˠ ae̯ d̥ʑ̥ mˠʷ ʌ̃ː d̪ˠ aɪ̃ t̪s̪ː|tsː sːʼ y̤ ɛˑ ɖ̤ʱ ɟʝː ˧˦˧ kŋ r̠ʲ t̰̪ ʱɾ ʌu ntʷ l̥ː ɫ̺ qχʷ ɪ̞ ɲcʰ ɘ̃ ɑe œ̞ ɡ̤ʷ d̪ð cʲ c̤ ũĩ ʃːʷ ɡᶣ s̪ʲ jo ɡbʲ tɕʼ wʱ ɳʈʂʰ bz ʼn d̤̪ː ʃʰ oːˤ ˀd fˠ qχːʼ ṳˑ ɟ̤ ɟʲ d̤z n̥ʲ ʰtsː d̰̪|d̰ ɡǃx uɑ̃ cçʷ ʊɪ bˤ ɨ˞ n̪ˠ ɗʲ ᶑʼ æu̯ ɛ̞ kːʷʼ ⁿtɕʰ r̪ː R̰̪|R̰ e̞o̞ ɴqʰ ʄ̥ ɟ̰ ŋɣ ɛə ö ˦↓˦ ʊ̆ cːʼ nˠ iːˤ ɖr u˞ iaː ɵ̞̆ ɡ̤ʱ n̪d̪z̪ wə ʕʷ d̥z̥ ɛ̤ˑ mˤ tʲʼ kǀ͓ʰ ʁˤ l̤̪|l̤ ï ɡːʷ s̪ʰ|sʰ ɭ̥ ʌi t̪s̪ʲ|tsʲ ʱd tɬːʰ o̞ˤ nsʷ ḛ̃ b̤ʱ q̰ʷ iu̯ z͇ ʱn ɛi̯ o̝ ɺ̪|ɺ äi̯ ŋ̥ǀ͓ˀ kˤ i̯e qʲʼ xːʷ tːs ʕ̝ ʰtː e̠ χˤʷ ʰt̠ʃː ɳʱ w̥ wɛ øy kxːʼ ɖ̤ː a̤ˑ dʑʷ ɛ̆ ŋǃ̠ ĩ̯ üː l̪ː ɹ̥ ɡ̤ǃ ʈr eɪ̃ e̯ e̤ˑ ʰkː'/

Phonemes with exactly 2 occurences in Phoible:

/'ay ḛ̞ tn ɲ̟̥ ʈɳ tɬː ŋ̤ ĩ̞ mʼ ɡ̃ ɻ̊ uẽ ɕʷ ˨˥˩ t͈ː ʊ̙ː kˀʷ øu ɵy d̤ʒ̤ ɕᶣ ɾ̠ o̤ˑ s̰ ɡʼ m̤b̤ ɪ̤ n̠̥ ʒ̺ ʋː ɾ̪ˠ oːu̯ iã ɡɣ ɔːˤ ʃˠ ʃ͈ː ɵ̃ t̪ʷ|tʷ ʊu z̪ː ndɾ b̤ː ɪ̈ː kǁx sᶣ nʰ ʊə ṽ dˀ dɾ tɕː kǀx ũ̯ æ̤ː ãĩ oə n̤d̤ɮ̤ mːʷ i̯ɛ ʒˤ ŋ̥ǁ͓ˀ ɔ̝ː ɐɪ cʰʷ ɗʒ ɔˑ n̥ˠ l̥ˠ ɢ̥ kᶣ yə̯ æ̤ ˀr nʼ ɨ̠ ɛ̰ː uə̯ bːʷ ŏ i̜ ɒ̤ z̞̩ l̠ʲ iɪ̯ uːə kʘ ɓʲ ɔ̟̜ ɵ̤ u̙ ɕ̟ ɯ̙ ʱl n̪t̪|nt ˧˨˧ aʲ ɑɯ ⁿʈʰ n̪t̪ʰ k͈ː sːʲ n̤d̤ ɮ̪ c̰ iɛ̃ ẽ̞ĩ tɕʰʷ ʊˤ m̥ʲ ɡǂx lˀ mɓ l̠˞ ʐ̩ vʱ nˤ ɜi iːɛ p͈ β̞ː θ̪ ʊ̤ː ɲʱ ɣː pˤ ʎ̟̥ kǃ̠ʰ mpʲ ɡbʷ k̙ æe kǁ͓ʰ æːʊ rʰ ø̞ˤ yi wei̯ ʼw ts͇ ɡv x̠ t̪s̪ː tz̤̥ ɛ̝ː b̪ ɤi ŋ̥ǃ ɑi̯ nɟ ˀŋ ʎː t̪ʰʲ dz̤ ə̯ uːʌ ntsʼ ɰ̰ t͈ ɓʷ ʱn̪ ɘ̠ ʱj yi̯ nr ʊ̙ zˠ yu̯ tl ɹ̪|ɹ d̪̥ t̻s̻ʼ ʂ͇ ë̞ æ̃ɪ̯̃ ɤ̞ː ɺʲ oɪ ˥˧˥ pm dz̪ ˨˦˧ lːˤ ɸʼ ɡbː qχʷʼ nːʲ ɡǀx ɪ̰ nh ɲᶣ kǃʼ eɔ yø̞ ɔ̰ː ɬˠ ɛʉ pɦ ɬ̪ʼ|ɬʼ ɾˤ oʊ̃ ɖɽ t̪ʰˠ ̰˩ kːʰʷ nd̪ ɔ̃ĩ ʱb ŋʱ ɪ̙ ʌ̤ː ɾˀ ʊ̯ɔ r̪ʲ|rʲ mbv kǂx ŏ̞ kɦ ʼŋ ɔ̜ː s̠ yæ ɥˤ eːi̯ ɾ̥ˠ uæ s͈ː ɬʷ ɨ̹ t̠ʃˤ əy ⱱ̟ ɽ̤ʱ t̪ʷ ʉ̃ː aːu ɦʷ ðˤ ii̯ t̪ɬ̪ʰ ʉ̟ tɾ ʈɽ t̪ː|tː k̃ tɕʷ z̪͇|z͇ ⁿd̪z̪ nl ʱw ˀp n̠d̠ʒʷ ɔˤ f͉ ɡ̤̥ ɨɛ ʉ̞ uaː n̤̪|n̤ ˨˩ çʲ n̠t̠ʃʼ bˠʷ kpː o̞̜ ɔ̽ ɞʏ̯ n̠ʲ dx ʁˤʷ uei ɡǀ͓ iːə t̪n̪ ɬ̪ː|ɬː ⁿd̪ ɲ̩ ow k͈ d̤̻ʱ kǂʰ d̠ʒʼ b̤ʲ ʀʁ b̤̥ pʰˠ ɲj l̪̪ r̥ʲ i̯a kǁʼ ɽʰ pʰʷ ɑ̝ qˤʼ ɢʁ n̰̪ o̯ t̪ɬ̪ ø̃ ɒ̆ t̪s̪ːʼ|tsːʼ ãũ dl ˥˧ ɸʷ aʊ̃ bˀ z̹̩ ɽ̩ ɡʲʷ ãˤ aɵ̯ tɕᶣ v̩ i̠ ʒ̊ ɪ̯u̞ ɥ̥ eˠ ḭː t̠ʃ̺ʰ ʁʲ ə̰̃ ii ʼj ʰkʷ qχːʰ ʼɲ ʰc pʰˠʷ βː ɔ̃ʊ̯̃ ʑʷ n̪ʷ|nʷ yœ ŋ̥ʷ ʌːˤ aɨ e˞ ŋ̤ǃ qˤ lʰ iou k̟ʰ pʷʼ ʒˠ qːʼ lːʲ ŋǂ d̤z̤ bβ uã ɧ kf ɗʷ o̞a ɦ̤ ŋkʲ ̰˥ i̘ jau̯ iə̯ ɛ̠ ɪɛ ɡǁx ɸˠʷ d̤̥ ɠ̥ õ̞ĩ iʌ ɯ̽ː ĕ̞ ʙ œi ø̤ ŋ̥ǀ͓ʰ ʱɟ s̪̻ nzʷ dʑᶣ dz̻ ɗʱ fːʷ uu̯ eɪ̯ ʱŋ tɕʷʼ ʊːˤ ɯ͓ ɾʱ t̠ʃːʷ ʐʲ æˤ æɪ dʲʷ t̪ˀ|tˀ iɛː t̪̻s̪̻ z̤ ˧˨˥ ŋ̥ǃ̠ˀ ŋ̥ǃˀ ŋ̥ʲ ɪˤ'/

There are like.... 1000++ nonce phonemes in the database.

[gach]

That list highlights how we should rather be looking at distinguishing features than narrow phonetic notation when in search for rare phonemes. How is /kʘ/ for example different from a plain bilabial click or how relevant is the stress mark in /'ay/? I'm quite sure that there are many more languages that allow a [low, unrounded] to [high, front, rounded] diphthong on stressed syllables. I'm also quite sceptical about the tonemes in these lists (/˦ ˨˩ ˧˨˧/ etc.). Surely they appear in the 2- and 3-language lists only because most phoneme descriptions don't include tones in their phoneme lists.

Well, the problems with Phoible data.

[finlay]

θ̪

[Frislander]

θ̪

*snorts with stifled laughter*

But seriously yeah Phoible is fairly useless in several respects: another problem I find is how it sticks to the traditional IPA chart so rigidly that it leaves things like ejectives, aspirated consonants and affricaes out of the main table in a separate list which is just like a pain to read. Just take the Archi entry for example; that must be like 5/6ths or more of the inventory that isn't in the actual table.

[gach]

There are a lot of issues with Phoible. A classic one is reporting the Finnish palatal glide as /e̯/. I have no idea.

[Qwynegold]

Malayalam has both /ɕ/ AND /ʃ/ (definitely NOT */ʂ/). How common is that?

At least some dialects of Swedish have both.

Really? Because the Wikipedia article on Swedish phonology seems to say that the varieties in southern Sweden have /x/ and /ɕ/, the varieties in northern Sweden have /ʂ/ and /ɕ/, and the ones in Finland have something like /ʃ/ or /ɕ/ and /t͡ɕ/ or /t͡ʃ/.

AFAIK, all dialects spoken in Sweden have [ɕ] for /ɕ/, while the realization of /ɧ/ varies. Some posh people use [ʃ] for /ɧ/. I've also met someone non-posh who used [ʃ]. I'm not sure where he was from. I think it was like Dalarna or something.