I am learning IPA and I learned that vowels are nasalized before nasals. When we are transcribing nasalized vowels, we write the tilde over them to denote that they are nasalized. My question is "are diphthongs nasalized before nasals"?

Example: we have "tine" transcribed as [taɪn]. Is the diphthong [aɪ] nasalized? If it is, how do I transcribe it?

We transcribe nasalized vowel for example [ɪ̃]. But how do I transcribe a nasalized diphthong?

[ãɪ̃]? Or [aɪ̃]? Which vowel of the diphthong is nasalized? Are both of them nasalized?

  • 6
    That depends on the language, the dialect, the speaker, the particular utterance… I think the only thing that can be said as a general rule is that in a sequence /aɪn/, it is very unlikely the output will be [ãɪn] with only the first part of the diphthong nasalised. The other three options, [aɪn ~ aɪ̃n ~ ãɪ̃n] are all, a priori, more or less equally likely to occur. Sep 19, 2020 at 9:46
  • 1
    @curiousdannii, Nonsense! Whether or not it can be answered in general, it is about linguistics. Where else could the OP ask this question if not here? Sep 20, 2020 at 4:50
  • 1
    @curiousdannii, I'm not sure as to how this question is not meaningfully specific. This is a very good question. Where's the rule that says that this kind of questions are not allowed? Can you show me such rule? Sep 20, 2020 at 4:54
  • 1
    @JoyfulSadness You're quite right. This is an excellent question for this site. Oct 24, 2020 at 23:28
  • 1
    @jlawler English is a language that nasalises vowels both before and after nasals. (We have lazy velums!) Oct 24, 2020 at 23:30

1 Answer 1


Unlike French, there is no difference in underlying forms between oral and nasal vowels. It is widely believed that there is some nasalization of vowels before nasal consonants in English, so the obvious question to ask is, how can you possibly decide where if anywhere the nasalization is. The best approach is impractical but maybe worth a try. Get some hardware that measures nasal and oral airflow, and see when the nasal airflow starts to rise. Cohn did that in her UCLA dissertation, using a couple of California speakers, and compared those results to productions from French and Sundanese speakers, which have phonological nasalization. English is different, at least for her speakers: it is clear that there is a continuous increase in nasal airflow over the second half of a pre-nasal vowel, so the vowel isn't all-oral and it isn't all-nasal.

A cheap substitute for fancy machinery is micro-listening, using a bunch of recordings of word pairs like "made/mane", "pain/paid", "kin/kid", with the final consonant being either a voiced stop of a nasal. Using Praat or similar sound editor, trim the final consonant, and some contiguous portion of the preceding vowel. It is possible that the left portion of "pain" and "paid" become indistinguishable after editing out only a tiny bit of the vowel; or it may be that the two words are still recoverable even from just 40 msc of [pɛ] vs [pɛ̃]. If that is the case, you have evidence that the vowel is fully nasalized. With "mane / made" you also have the possibility of progressive nasalization, so an analogous question can be asked about "bane/bade" vs. "mane/made" – is there nasalization in "made" compared to "bade"?

In US English (some varieties) there is a marginal minimal pair, "can" and "can't", the latter having a nasal vowel. In UK English the vowels are totally different. The prospects for crossover to phonemic status are greater in US English, so barring an accent with surprisingly more nasalization, I expect that what they are telling you about "nasalization" is something about low-level phonetic details – which is best studied instrumentally and quantitatively. I think it is a misanalysis to say that vowels in English are "nasalized", rather they have some degree of coarticulation with an adjacent nasal consonant.However, this is just for your information: I'm not suggesting that you argue with your teachers.

  • 2
    (Stressed) can and can’t both have a nasalised vowel for me. The difference is that can is [kʰæ̃ːn], while can’t is [kʰæ̃ːʔt̚] utterance-finally or before a consonant (with assimilation of the final [t̚]), and [kʰæ̃ːɾ] before a vowel. Sep 20, 2020 at 8:10
  • Thank you for the thorough explanation
    – user30668
    Sep 22, 2020 at 10:57
  • 1
    I find this answer perplexing. In relation to a language such as English in which nasalisation is not phonemic, nasalisation refers exactly to the result of co-articulatory processes whereby part, or sometimes, all of a vowel is produced with air escaping through the nasal cavity because the velum is lowered because of a preceding or following nasal. As mentioned in a comment above, there's no rule for how much of a vowel may be nasalised in such an environment. It differs from speaker to speaker, and with other contextual factors. But it may well be all of it, especially if ... Oct 24, 2020 at 23:09
  • 1
    (cont) ... you measure it perceptually. It's perfectly possible for the velum to lower during that period of the vowel where there's a positive VOT, in other words during the aspiration that will be present in a word like pain. The experiments you talk about are common, everyday demonstrations used by phoneticians in under and postgrad classes that I've attended. I agree that it's less likely that the vowel will be fully nasalised in tine, but then if it's increasingly nasalised toward the end of the diphthong, I think OP is owed an answer as to how to represent this in the transcription! Oct 24, 2020 at 23:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.