Are there any languages which distinguish consonants exclusively by manner of articulation instead of place of articulation and such, akin to a vertical vowel system?

  • 1
    Do you mean something other than the usual meaning of "matter of articulation"? For example, in the usual meaning, [t] and [s] are consonants with the same place of articulation (alveolar), but different manner of articulation (plosive vs fricative)
    – Cecilia
    Commented Jul 8, 2022 at 10:15
  • No, I mean (I think) the normal meaning of manner of articulation. E.g. Fricative/Approximation/Nasal/Plosive and so.
    – Qaziquza
    Commented Jul 8, 2022 at 11:26
  • 1
    Then, for example, English is such a language, as it distinguishes the aforementioned [t] and [s], two consonants with same place but different manner of articulation.
    – Cecilia
    Commented Jul 8, 2022 at 19:23
  • 1
    Ah, I meant exclusively by manner of articulation. I will edit the question to avoid ambiguity.
    – Qaziquza
    Commented Jul 8, 2022 at 19:28

3 Answers 3


I don't know of any, and I wouldn't expect to find any.

From an information-theory point of view, consonants tend to convey much more information than vowels. I can't think of any natural language that has more vowel phonemes than consonants, even if you include tone as a property of the vowel (rather than of the syllable); there are just a lot more possible ways to obstruct an airflow than to let it go through, so there's a much larger space for the phonemes to occupy without crowding each other and becoming difficult to distinguish.

Vertical vowel systems take some of the information load from the consonants and duplicate it on the vowels. If vowels before labialized consonants become rounded, now you have a bit more redundancy there, and you don't have to worry as much about distinguishing labialization on consonants. This is most useful when each consonant is conveying a lot of information, like in Ubykh, with its 84 phonemic consonants and 3 phonemic vowels.

A "vertical consonant system", presumably, would take some of the information load from the vowels and duplicate it on the consonants. But in natural human languages, this isn't very useful—the consonants are already carrying a lot more information than the vowels are. Over time, I would expect this to simplify into only marking the information on the consonants—and now it's just a normal consonant system, not a vertical one any more.


I am not aware of any natural language with that feature, and I quick-checked that Damin hasn't it either.

But you can certainly construct such a language, there are enough different manners of articulation to make up a sufficient consonant phoneme inventory.


No, but it is possible. You can go from /m n p t k s w j/ to /n t ts s j/ like that: /w/ > /m/ > /n/, /p/ > /f/ > /h/ > nothing, /k/ > /kx/ > /cç/ > /tʃ/ > /ts/

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.