I am creating a language called Knashta, and one of the phonemes is /p͡r/.

I believe this sound would be a trilled affricate, and I'm guessing that it's name would be a voiceless bilabial alveolar trill.

Am I right or wrong?

  • 2
    Affricates must have a fricative as the second component, your /p͡r/ doesn't have one, so it's not an affricate.
    – Yellow Sky
    May 11, 2014 at 16:54
  • A trilled affricate is not a regular affricate. It begins with a stop and ends with a trill. An example would be [mbʙ], the voiced prenasalized trilled bilabial affricate. If you don't like the term affricate, you can also call it a 'post-trilled consonant'
    – Igor
    May 11, 2014 at 16:58
  • Still, [ʙ] is a trilled fricative. And I like the term affricate, only it doesn't fit your sound.
    – Yellow Sky
    May 11, 2014 at 17:32
  • The term trilled affricate is used in literature, whether or not we like it as fitting this type of sound.
    – LjL
    Apr 24, 2020 at 0:47

2 Answers 2


As it happens, there is a model of speech production -- the Convertor-Distributor Model (Phonetica 57:128-38) -- advocated by Osamu Fujimura, which denies that there is ordering of elements in a syllable position such as "onset". To handle clusters like [pr], the model has bigger inventory of "segments", so [pr] would be a voiceless labial alveolar trill (assuming a trilled r). I don't agree with the model, but the point is that your proposal corresponds to something that exists in linguistic theory.


This is a cluster.

What constitutes affricates and the like, well it's not part of this arrangement. Anything like this trying to be incorporated in a nomenclature like affricate is serving two masters. Which is bad. Creating a language still falls within the parameters of the research Chomsky presented in grammar, phonology, and phonetics. What's more - a language that is so called created is not and never a natural language and does not stem from the natural development all languages follow to the latter.

  • 1
    This question isn't about the merit of constructed languages. Also, please don't use caps. It's not necessary.
    – Igor
    May 11, 2014 at 20:19
  • 2
    Welcome to Linguistics SE, Messiah! I fail to see how your contribution answers the question. Do you think you can make that clearer?
    – robert
    May 11, 2014 at 20:50

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.