I'm having a difficult time trying to find languages that have a phonemic contrast between /ʂ/ and /ʃ/.

I can hear the difference without difficulty because /ʂ/ sounds like a lower frequency range of noise than /ʃ/ does.

According to the paper "A Perceptual Study of Polish Fricatives, and its Implications for Historical Sound Change" by Marzena Żygisa and Jaye Padgett:

Polish [...] contrast[s] denti-alveolar [s, z, ts, dz], alveolo-palatal [ɕ, ʑ, tɕ, dʑ], and retroflex [ʂ, ʐ, tʂ, dʐ] places of articulation. In addition, a palatalized palatoalveolar sound [ʃʲ] exists as an allophone of /ȿ/ when this phoneme occurs before [i] or [j] as in To[ʃʲ]iba ‘Toshiba’.

In "Typology of the Syllable-Initial Consonants in the Chinese Dialects" by Wai-Sum Lee, I found /ʂ/ and /ʐ/ in more than one variety of Chinese fricatives, but not /ʃ/.

Are there languages that phonemically contrast /ʂ/ and /ʃ/?

  • 2
    I think Swedish has this contrast since /ɕ/ is a regular phoneme and /rs/ is realized by most as [ʂ]. I don't recall any minimal pairs at the moment, but the /rs/ phenomenon happens even across word boundaries and so it shouldn't be hard to construct minimal pairs. Feb 12 at 23:21
  • 1
    As one of the more notable examples, I believe Sanskrit features this contrast, however this is now only spoken as a literary language and it has no L1 speakers. Regardless, it still is a natural language. Feb 13 at 4:02
  • @Vegawatcher: agreed that standard Swedish has both [ʂ] and [ɕ], but I’d have thought the contrast is just phonetic rather than phonemic, since as you say, [ʂ] occurs just as the realisation of /rs/? Finlandsvensk dialects also have phonemic /ʃ/, for the “sj-sound”, but they don’t have [ʂ] for /rs/; but I’m not sure what they do with the “tj-sound”, i.e. the /ɕ/ of standard Swedish, and on brief searching I can’t find a clear answer or a recording. So possibly those dialects contrast /ʃ/ and /ɕ/?
    – PLL
    Feb 13 at 17:16
  • Sounds like ś and š in Montenegrin?
    – Davor
    Feb 15 at 0:41
  • 1
    @EdAvis - basically the issue is that Montenegrin pronunciation drifted so far away from "sj", they were pronouncing the word differently compared to how its written, which is a problem in a phonetically spelled language. So they invented a new letter to support the regional pronunciation. In any case, those two are pronounced completely differently, so it's not really a replacement, more of a fix for the previous inadequacy of the phonetic alphabet used in Montenegro.
    – Davor
    Feb 16 at 14:34

7 Answers 7


Skimming Phoible, stopping with languages beginning with n, I found as putative examples from: Abkhaz, Acoma, Arara do Acre, Basero, Basque, Bench, Burushaski, Cajamarca Quechua, Camsa, Candoshi-Shapra, Caodeng Rgyalrong, Capanahua, Cashibo-Cacataibo, Chacobo, Cham, Chamicuro, Chasta Costa, Chipaya, Chon, Cupeno, Curripaco, Dagur, Dongxiang, Eggon, Gaan Yajich, Gimira, Gserpa, Guambiano, Hindi, Hmong, Ishkashimi, Jacaltec, Kannada, Khalong Tibetan, Baniwa, Ladakhi, Luiseno, Malayalam, Manange, Mangghuer, Matas, Matsas, Matses, Mochica, Muniche, Munji.

Hindi and Kannada are probably correct, but Basque is a matter of interpretation (what are "s", "z" and "x"?), likewise Hmong ("sh" and "x").


As you mentioned Chinese, Standard Mandarin only contrasts /ʂ/, /ɕ/ and /s/. For example 殺/ʂᴀ⁵⁵/ 蝦/ɕᴀ⁵⁵/ and 撒/sᴀ⁵⁵/. In fact the complete contrasts are between the three groups /ʈ͡ʂ ʈ͡ʂʰ ʂ/, /t͡ɕ t͡ɕʰ ɕ/ and /t͡s t͡sʰ s/.

Some variants of Chinese contrasts /ʂ/ and /ʃ/, notably Jiaoliao Mandarin contrasts: 升ʃəŋ1 vs. 生ʂəŋ1. 1 means the tone type 陰平. enter image description here In this area, Wendeng and Rongcheng have /ʂ/ and /ʃ/ that are pretty standard. The two corresponding contrasting groups are /ʈ͡ʂ ʈ͡ʂʰ ʂ/ and /t͡ʃ t͡ʃʰ ʃ/.

Some subdialects of Central Plains Mandarin (like Guanzhong and Longzhong) also seem to contrast /ʂ/ and /ʃ/ but I can't find first-hand documents.

  • 3
    This four-way distinction between /ʂ/, /ʃ/, /ç~ɕ/ and /s/ in Jiao-Liao 胶辽 Mandarin is particularly fascinating. The /ʂ/ vs /ʃ/ distinction is historically conditioned by the Middle Chinese distinction between the retroflex stops [知], retroflex affricates/sibilants [莊/生] and the palatal affricates/sibilants [章/書], but the details of which lexemes map to which affricate/sibilant vary somewhat between locality.
    – Michaelyus
    Feb 15 at 16:47
  • What’s the meaning of the /a/’s in the Mandarin words being small caps? That’s not an IPA vowel, as far as I know. Jun 18 at 23:59
  • @JanusBahsJacquet It's an open central /a/. Chinese scholars use it to denote the Chinese /a/ for simplicity.
    – lilysirius
    Jun 19 at 0:07

Ubykh is an extinct Northwest Caucasian language (and thus in the same family as Adyghe) that contrasted the following ʃ-like phonemes: /ʃ/, /ʃʷ/, /ɕ/, /ɕʷ/, /ʂ/.


Querying UPSID for three or more voiceless sibilant consonants and skimming through the results, I find Tarascan and Pashto as candidate languages with a contrast between /ʂ/ and /ʃ/.


Adyghe language contrasts /ʂ ʃ/ and their voiced counterparts.


Using UPSID (which I just learned about), I found that Guambiano has this contrast too.


Gwitch'in has two whole separate series of postalveolars and retroflexes

  • Hi Dániel, and welcome to the site! Could you add some more detail here? For example, what else do these two series contrast against?
    – Draconis
    Jun 17 at 20:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.