In English, diphthongs are single phonemes, and monolingual English speakers hear them as a single vowel. However, this does not mean that English speakers will hear all diphthongs in other languages as single sounds. Is there any research on which diphthongs English speakers perceive as single sounds and which ones they don't?


2 Answers 2


The only definition of "single sound" that exists in phonology is "single segment", which is different from the phonetic view (whereby "church" has a bisegmental articulation at the beginning t plus ʃ). There is a tie diacritic that can be used in IPA to indicate "functions as one segment". Notorious examples of ambiguous segments are aspirates (th), NC sequences a.k.a prenasalized segments (mb, mp, mf), consonant+glide sequences a.k.a. palatalized and rounded segments, and diphthongs. Fujimura's model of syllables includes pre-spirantized stops (sp, th...), and similar notions turn certain seeming consonant clusters of Georgian into single complex segments.

There is no general test of subject behavior that establishes "single segment behavior" from "cluster behavior". Asking "how many vowels are there in X" or "how many consonants are there in X" is a really poor way of determining the internal representation of sounds in a language. Typically, speakers resort to counting orthographic units, so that θ in English has a count of 2 and expect that speakers of Greek assign a count to 1 to the cluster ks. The first step would therefore be establishing a reliable basis for comparison. This has yet to be done.


If it's a matter of perception, it would be fairest to say, I think, that a diphthong means precisely a cluster of vowels that speakers perceive and treat as a unit. For example, in elementary school you learn about the "short i" /ɪ/ and the "long i" /aj/ and you treat them both as "first-class citizens" in your inventory of sounds.

It doesn't take too much time to demonstrate to a native speaker that some of these vowels actually have multiple components. I've frequently had high school students just slow down their articulation until they feel that they're saying /a/ and then /j/. The same is true of /aʊ/ and /ɔj/.

However, it's harder to get people to differentiate the sounds in /ej, ij, oʊ, uw/. They don't always perceive the boundary between the phones and if you demonstrate then they tend to feel that you're just saying /e, i, o, u/ even if you stop before the off-glide. I think this is because we don't have separate sounds for /e, i, o, u/, whereas many dialects have /a/ and /ɔ/ or at least align these with other discrete phonemes like /ɑ/.

So my intuitive, anecdotal response to your question would be that the diphthongs that are stubbornly perceived as a unit are those whose components don't align with their own phonemes.

  • Yes, this is all true for English.
    – Lambie
    Jan 5 at 19:49

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.