I assume, considering the Onset principle, that there are not many languages that have a structure with VV or VVV but are there languages that have a CV.VV structure? If there is, I would assume that these vowels tend to gravitate towards cardinal vowels in order to keep too many phonemes being reduced/neutralized?

Likewise, I assume languages that have long consonant clusters (like Polish, Georgian or Armenian) have a specific set of consonants available based on what is likely to keep its perceptual cues. For example, these languages allow long clusters of fricatives because a fricative is more likely to keep its perceptual cue (according to Wright 2004) than a string of stops. Assuming this, then these languages would have a reduced inventory of other consonants like stops, nasals and liquids, no?

My professor suggested turning this into a research paper so I'm trying to get a feel of what info is available. Are there other languages other than the ones that I've mentioned that have long vowel or consonant clusters?

  • If you do pursue the matter, you ought to be careful to distinguish phonemic forms from pronunciations. Some languages written phonemically with clusters of consonants may have those clusters broken up by epenthetic vowels, and clusters of vowel phonemes may be broken up by epenthetic glides. Or, you may be dealing with an incorrect phonemicization. Another complication is a possible confusion between "vowel" and "syllabic sound".
    – Greg Lee
    Jan 31, 2015 at 19:40

1 Answer 1


Questions about there being "many languages" with some property are basically unanswerable since there's no contextually reasonable number that constitutes "many". Anyhow, there are quite a number of languages with VV and VVV -- basically, you just need a language that doesn't particularly care about vowel sequences. You also find CV.VV in languages such as Ancient Greek and many Bantu languages. I don't know what it would mean for these languages to gravitate towards cardinal vowels. In these languages, there are no stress-related reductions so there are not many vowel neutralizations. Vowels are really easy to parse, so there's no inherent problem with long vowels sequences.

The languages that you mentioned as having long consonant clusters actually have a decent supply of consonants (you can add Berber, Lushootseed and Bella Coola to the list), so I don't think you can say that they have a reduced set of consonants. Perhaps you mean that only a subset of the consonant inventory can appear in consonant clusters. This is fairly typical, and is observed in English where we don't allow stops after onset stops, obstruent clusters in onsets other that sC (and ʃC for some dialects), and so on. In fact, it might be the case that no language which allows onset or coda clusters up to length N allows all possible permutations of that length -- there is always some sequencing restriction. It is a theorem of the Clements & Keyser theory of syllabification that longer clusters are more restricted than shorter clusters (because cluster restrictions refer maximally to two consonants, and any three-C sequence has to pass all of the 2-C filters, and passing all the filters gets much harder as the sequence grows).

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.