In Spanish, there is no orthographic distinction between the diphthongs [wi ~ uj] and the vowel hiatus [ui]. Is there a way to predict how "ui" is pronounced in a word?

Possibly via phonetic environment (i.e. the preceding consonants)?

It seems to me that, for example, the sequence "cui" is generally a diphthong (e.g. cuídate), and that verbs in "-uir" are generally in hiatus (e.g. construir). Can such generalisations be made? The RAE seems to imply the realisation of "ui" are heavily dialect, sociolect, and idiolect dependent:

Por otra parte, algunas de estas combinaciones vocálicas (las formadas por... dos vocales cerradas diferentes) pueden, en una misma palabra, fluctuar en su pronunciación entre el hiato y el diptongo, dependiendo de diversos factores, como el mayor o menor esmero en la pronunciación, el origen geográfico o social del hablante, etc.; así ocurre, por ejemplo, en gratuito, que puede pronunciarse con diptongo (gra - tui - to) o con hiato (gra - tu - i - to)...

  • Que interesante!. De la información que proporcionas parece deducirse que no hay regla general y que algunas veces todos los hablantes estamos de acuerdo pero en otros casos algunos usan una forma y otros la otra.
    – DGaleano
    Feb 19, 2020 at 13:06
  • I think you may have a typo [wi ~ uj] should be [wi ~ ui] I think. Unless its my ageing eyes of course.
    – mdewey
    Feb 19, 2020 at 13:44
  • @mdewey not a typo: /j/ us the ipa symbol for the semivowel equivalent of /i/ en.wikipedia.org/wiki/Voiced_palatal_approximant
    – jacobo
    Feb 19, 2020 at 13:48
  • Thanks for that. I did not edit it myself as it did seem unlikely to be a typo. I am not that familiar with IPA sadly.
    – mdewey
    Feb 19, 2020 at 14:39

1 Answer 1


Regarding the hiatus [u.i], all verb infinitives ending in -uir have it, and it's inherited by all their inflections and derivations where it may apply: substituir, substituido; recluir, recluido; fluir, fluido, fluidez; huir, huida, etc. I think this is so because we are used to verbs getting stress in the last vowel of the infinitive; we only recognize three verbal classes (-ar, -er, -ir). It might also have to do with the fact that in many of the finite forms of the verb, the vowel /u/ remains clearly part of the root while the /i/ is changed to /ʝ/. This reinforces the idea that the /u/ must be left alone.

Now, it shouldn't be strange to find speakers that turn the hiatus into a diphthong in some cases, like when the stress moves elsewhere. I pronounce huidizo with a hiatus but fluidificar (if that were a word) I think I would pronounce with a diphthong. (There are not many examples to choose and try.)

Outside the cases above, most instances of ui are pronounced as the rising diphthong [wi].

The falling diphthong [uj] is usually orthographically marked as uy, and the only examples of it that come to mind are muy and cuy, both monosyllables. The proper place name Ibicuy is also [ibi'kuj]. I don't think I've ever heard unstressed [uj], although the difference might be hard to perceive. Notably, the variant cuis (for cuy) is pronounced [kwis].

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.