I’m wondering how exactly do you make a “w” and “y” sound in English. These two are considered the glides of English, but what exactly makes it a glide? What are the characteristics of a glide sound? For instance, is there a difference in sound between “yi” and “ii.” I know those are not actual words, but someone once told me that a y sound is basically an i sound in the beginning, quickly followed by the main vowel of the word. So for the word “you,” it’s “iu”. Same for w, except it’s an “u” sound instead. So is there a difference between “wu” and “uu”? I never really thought of the y and w as two vowels joined together. Kinda just treated it as one particular kind of sound, though hard to describe.

3 Answers 3


Glides (or "semivowels") are sounds that are not phonetically dissimilar from vowels but behave like consonants—that is, they cannot constitute the nucleus (peak) of a syllable.

From a purely articulatory point of view, [j] and [w] are just short occurrences of [i] and [u] (except that in [j, w], the constrictions may be slightly more prominent) and can be alternatively transcribed in IPA as [i̯] and [u̯]. But most linguists consider English [j] and [w] as distinct sounds (phonemes). The reasons are primarily phonological, which include:

  • When at the beginning of utterances, east, Uber, etc. are often preceded by a glottal stop. This doesn't happen with yeast, womb, etc.
  • We say "a year", "a week", etc., not "an week", "an year".
  • They cannot be stressed like vowels.
  • They behave like other approximants, /r/ and /l/, in that they become voiceless when preceded by a word-initial voiceless plosive, as in cue and queen (compare crew and clean).
  • They can precede almost any vowel, as in Yiddish, wit, yet, wet, yap, wax, yacht, watch, wood, young, one... Analyzing [j, w] as vowels would entail adding a considerable number of diphthongs and triphthongs into the phonemic inventory of English.

It is possible to analyze the non-syllabic components [ɪ, ʊ] of the diphthongs /aɪ/, /aʊ/, /ɔɪ/, etc. as /j, w/, but since /j, w/ otherwise do not occur in syllable codas, these diphthongs are usually considered distinct phonemes rather than sequences of two phonemes. (Phonetically, however, there is little to no difference between [j, w] and non-syllabic [ɪ, ʊ] of diphthongs. The non-syllabic components of English diphthongs are often conventionally transcribed with [ɪ, ʊ], but the actual quality of the sounds varies depending on accent and may be closer to [i, u].)

  • Would you remind me again what’s the difference between phonetically and phonologically? You would think I can tell the difference by now, but I’ve always had a hard time differentiating the two when they are within some sort of context. How I differentiate them is phonetics is based on sounds of the human speech while phonology is based on the sounds of a language instead. This was the definition I sticked with whenever I couldn’t differentiate. What did you mean when you said phonogically and phonetically? Mar 10, 2019 at 6:20
  • 1
    See linguistics.stackexchange.com/q/180. Put crudely, phonetics is concerned with actual physical sounds humans make whereas phonology deals with the distribution of sounds within a language and their theoretical relations with one another and with other aspects of the language like morphology. They are closely entwined and the definitions depend on context, but when contrasted, phonology refers to theoretical, abstract aspects of speech sounds, and phonetics refers to physical, tangible aspects.
    – Nardog
    Mar 10, 2019 at 8:07
  • Is there a reason why it’s called a “glide”? What is “gliding” anyway? It’s funny how I never thought of the y or w as a non-syllabic i or u. I always thought of the y and w as it’s own unique sound. I’m certainly not thinking of the letter i or u when I say words like “yeast” or “womb” when compared to words like “it” or “opps.” Is that the phonogical aspect of it in English that you were referring to? Mar 14, 2019 at 1:49
  • It's called a glide because it glides from/to vowels. If you see them on a waveform or a spectrogram you can't tell when it starts or ends (like this). See also this recent question.
    – Nardog
    Mar 14, 2019 at 16:54
  • By phonological I more or less mean "taking into account not just the physical sounds but also how they are used and perceived in the given language". English speakers perceive the /j, w/ as consonants, as evidenced by the alternation of the singular indefinite articles, and that's something linguists certainly take into consideration when determining whether the glides are phonemes, with the caveat that that perception is merely a symptom, not a cause, of the grammar stored in speakers' brains and doesn't necessarily explain why they perceive them that way.
    – Nardog
    Mar 14, 2019 at 16:59

The glides y,w are phonetically very similar to the corresponding vowels i,u in English, but they are shorter and more constricted. Since "ii" and "uu" don't exist in English, you really can't compare how they differ from "yi" and "wu". The suppose tense vowels "i" and "u" in English are phonetically lower and more diphthongal compared to [i:], [u:] as encountered in numerous other languages, and a narrower phonetic transcription would be [ɪj, ʊw]. It is not unusual for yi,wu in other languages to be phonetically indistinguishable from [i,u], especially if [i,u] in the language are closer to the cardinal vowels [i,u].


By experiment I find that "i'yi" and "u'wu" are quite different from "i'ii" and "u'uu".

The emphasis on the second syllable in "i'ii" is produced by a change in pitch accompanied by a glottal stop. The same applies to "u'uu".

When I say "u'wu" I can detect a near closure of the lips on "w".

When I say "i'yi" there is a definite lifting of the tongue towards the alveolar ridge.

In both cases there is a constriction of the airflow, either by the lips or the tongue.

As an experiment I tried doing both of these actions at once. It produced a sound that I've never heard in English! It sounded quite a lot like a jaw harp (jews harp) - you can listen to these online.

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.