Does any living language contrast /kʷ/ and /kw/?

If yes, is there a way I can hear a minimal pair spoken?

3 Answers 3


In theory, yes. Tashlhiyt Berber is said to have a contrast, but that does not mean that there are any minimal pairs. That article points to literature, saying that it is generally agreed that they are different. However, the article slightly undermines the claim by noting that [kw] and [kʷ] differ in terms of syllabification: which (potentially) means that the distinction is not [kʷ] vs. [kw], it resides in distinctive syllabification. That means that it depends on what you mean by "minimal pair". Tigrinya has a similar potential, that k+w sequences can be constructed and there is a /kʷ/ phoneme, but the actual phonetic values have other low-level properties associated with the sequence vs. single segment. In both cases, there's a good argument that kw is different, phonologically, from , but I suspect that minimal pairs are a deal-breaker.


Thai can be what you are looking for.

It has onset clusters /kw/, /kʰw/. Quite often, they are realized as labialized velar consonants /kʷ/, /kʰʷ/.

However¹, final stops like /-k/ are accompanied by a simultaneous glottal stop, thus making syllable boundaries well defined by intersyllabic juncture. This prevents any C-to-C coarticulation and resyllabification of final consonants.

So if there are two adjacent syllables, one ends with the final /-k/ and another starts with /w-/, they remain audibly separate.

Compare (press 🔈 speaker icon to listen):

  • ทุกวัน [tʰúk wan] "every day"
  • กวาง [kwaːŋ] "deer"

¹) P. K. Peggy Mok. Language-specific realizations of syllable structure andvowel-to-vowel coarticulation


Just because a language contrasts two sounds, doesn't mean there should be minimal pairs (cf. English /h/ and /ŋ/).

The IPA uses a plain w to symbolise the [w] sound (war) and a superscript ʷ for labialisation (i.e. secondary articulation). There's a constriction at the velum for w, but ʷ doesn't have any constriction at the velum, it's simply the roundedness of the lips.

Now the sequence [kw] consists of two sounds having two points of articulation (primary articulation) the constriction is made at the velum for both the sounds (velar throughout) whereas [kʷ] only has one point of primary articulation at the velum (the ʷ means that the lips are rounded at the same time of articulation of the [k]). The best way to understand this is to think of the English words queen and cool (ignoring [-iːn] [-uːl]): in the former case, there are two sounds having two primary articulations (i.e. [w] after [k]). The latter, however, has only one primary articulation (that of [k]) but if you look closely at your lips, you'll see that they're rounded in anticipation of the following rounded vowel ([u]).

Labialisation isn't phonemic in English, but phonetically, compare the following words:

  • [sʷuːn] and [swuːn] (soon and swoon)
  • [sʷuːp] and [swuːp] (soup and swoop)

For many speakers of English, the [s] in soon and soup is labialised due to anticipation of the following rounded vowel. It means English contrasts [sʷ] and [sw] only phonetically.

Minimal pairs for [kʷ] and [kw]

There are a few minimal pairs distinguishing [kʷ] and [kw] (not /kʷ/ and /kw/) in (British) English:

  • [ˈkʷɒɹᵊl] and [ˈkwɒɹᵊl] (coral and quarrel)
  • [kʷɔːl] and [kwɔːl] (call and quarl)
  • [kʷɔːn] and [kwɔːn] (corn and quarn)
  • [kʷɔːk] and [kwɔːk] (cork and quark)
  • [kʷɔːt] and [kwɔːt] (court/caught and quart)
  • [kʷɔːts] and [kwɔːts] (courts and quartz)

Some of these qu- words are quite rare and I've never heard then before but they do exist.

Some Salishan and Wakashan languages are said to contrast /kw/ and /kʷ/, but I haven't been able to find minimal pairs.

The Wakashan language Makah has a phonemic contrast between /k/ and /kʷ/. Waka'wala contrasts /kʲ/ and /kʷ/. Nahuatl also has a phonemic contrast between /kʷ/ and /k/ and I assume it also contrasts /kʷ/ and /kw/.

  • I do not accept that any of your English examples for kʷ or for sʷ is valid. There is no labialisation of the onset in any of these words,
    – fdb
    Commented Jun 15, 2021 at 16:34
  • @fdb: there's definitely labialisation when I pronounce them. I guess it's idiolectal Commented Jun 15, 2021 at 16:41
  • According to the Oxford University website: Labialization of velar and uvular obstruents [in Tashlhiyt] is usually contrastive, e.g. D/ikti/ 'hot' vs. D/ikwti/ 'recall'. Commented Jun 27, 2021 at 6:52
  • 1
    +1 I'm an enthusiast not an expert, so forgive me if I'm misunderstanding. But "labialise" means "form [a sound articulated elsewhere] with rounded lips", right? If that's so, then I (a speaker of Australian English) can confirm that I labialise consonants before rounded vowels like /ʊ/ and /u/; "cook", "do" have rounded lips at the onset. I would sooner believe that it's dialectal than idiolectal: if I try to unround them, it comes out like [kə̯ʊk], [də̯uː], which sounds not-Australian! Commented Jul 31, 2021 at 9:02
  • @TimPederick: Yep, that's exactly what labialisation means. I speak what's commonly called Modern RP and I definitely labialise (most) consonants before rounded vowels. :) (I'm not sure why fdb disagrees) Commented Jul 31, 2021 at 9:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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