I don't understand why ⟨o⟩ is called the "close-mid back rounded vowel" while ⟨ɤ⟩ is called the "close-mid back unrounded vowel" - they sound completely different and they feel completely different in my throat. The difference doesn't seem to be rounding.

Try "deriving" ⟨ɤ⟩ by starting with ⟨o⟩, then unround your lips. You may not even start with your lips rounded... I find that I don't ever round my lips much when I use this vowel, which seems like another argument against roundedness being the distinguishing feature between these vowels.

But anyway, for the sake of argument, start with your lips rounded and then, keeping the rest of your speech apparatus stationary, unround your lips. Completely open your mouth, even. Nothing significant changes, and the result sounds nothing like any recording of ⟨ɤ⟩ I've ever heard. It just sounds like a laxer ⟨o⟩.

I also want to say that I see the same problem with ⟨ɔ⟩ and ⟨ʌ⟩. Recordings of these vowels don't sound like rounded and unrounded versions of the same vowel. You can produce either of them either rounded or unrounded, and the real distinction between them seems to be that ⟨ʌ⟩ is slightly more front than ⟨ɔ⟩.

This might be a special case of a more general question. Roundedness seems to be the only thing distinguishing the names of multiple vowel pairs that sound completely different in ways that are unrelated to roundedness. So why don't those more dramatic differences factor into the names of these vowels, instead of roundedness?

I think this is different from the related question why roundedness is considered so essential that it's one of the few characteristics that is used in the formal names of vowels, when it seems so dispensable. But I suppose that's worth asking too. If you speak with your lips permanently puckered or permanently lax, it hardly sounds different at all.

I agree roundedness can convey meaning in terms of how lax or tense your speech is. But the actual rounded version of ⟨ʌ⟩ (by that I mean, not ⟨ɔ⟩, but the result you actually get when you say ⟨ʌ⟩ with your lips protruded) is so marginally different from ⟨ʌ⟩ that it's hard to imagine them not being allophones in most languages.

And indeed, this "actual" rounded ⟨ʌ⟩ I'm thinking of doesn't seem to appear on the vowel chart. I can't find any reference to it anywhere. Maybe because the name it would logically be given is already taken by ⟨ɔ⟩. That one is clearly more than different enough to be distinguished in many languages. But for some reason its name suggests an extremely close relationship with ⟨ʌ⟩, yet I hear radically different sounds.

So, is something wrong with the way I'm hearing or sounding these vowels? Am I missing something?

I suspected maybe the difference between ⟨ʌ⟩ and ⟨ɔ⟩ was indeed more than just roundedness, but the naming scheme doesn't offer enough degrees of gradation to really express that difference. So the more subtle difference between them, roundedness, was chosen instead. Maybe in languages that distinguish between them, ⟨ʌ⟩ is canonically unrounded and ⟨ɔ⟩ is canonically rounded. So I can concede roundedness is at least a theoretical difference.

But if I was gonna name these vowels (that is, the sounds I hear in recordings of them), I'd call ⟨ɔ⟩ a back vowel, and call ⟨ʌ⟩ a central-back vowel. You could include roundedness too, but at least include the more significant difference, so people don't drive themselves crazy trying to figure out how to say ⟨ɔ⟩ by forming a rounder ⟨ʌ⟩.

So maybe the ultimate question is, why are there so many degrees of openness, but only one degree of gradation between front and back? Is this a problem for IPA? How did these get named in such a nonsensical way and is there any effort to fix it?

Thank you!

  • 3
    I think you must be misinterpreting something, because pronouncing a standard IPA [o] and then unrounding your lips does indeed yield [ɤ]. By your user name, I am going to guess that your first language is Mandarin, and it’s worth noting that neither Mandarin nor English really has a standard IPA [o] sound in their sound inventories. The vowel/nucleus found in syllables like 博 is not [o] but a diphthong that starts slightly higher (and usually rounded) and ends somewhere a bit lower, more fronted and less rounded than [o], something like [ɤ̞̹̈] (in fact, sometimes entirely unrounded). Apr 3 at 9:38
  • No, my first language is English. I can't speak Mandarin. I just think Xiang Yu is a badass. Anyway, I'm not using sounds from English, I'm just trying to reproduce what I hear in recordings of the vowels.
    – Xiang Yu
    Apr 4 at 18:08

1 Answer 1


They are called that because that is how they are defined.

If the vowels feel different when you pronounce them, and sound different when you hear them, it is either because the vowels in question are being pronounced incorrectly or transcribed inaccurately.

Inaccurate transcription is pretty common and can occur for several reasons.

The first is that transcription is often imprecise, and the person making the transcription may choose a symbol that is "close enough" but easy to typeset than one that is more accurate but hard to typeset. Additionally broad transcriptions will often leave off features that can be regularly predicted (e.g. aspiration in English), meaning that producing the transcription exactly will not sound correct.

Additionally, a transcription may be based on a former variety of the language and the pronunciation may have shifted. A clear example of this is in British English. The symbols developed as a transcription for the former prestige variety (Received Pronunciation) are still the most commonly used, despite being wildly inaccurate as a description of the current prestige variety. The Phonetician Geoff Lindsey has several videos on this topic (starting with this one here) and advocates for a new transcription based on the phonetics of the current prestige variety that he terms Standard Southern British English.

Some key things to note are that, for instance, whilst the vowels in COT, CAUGHT, and CUT are still typically transcribed as [ɒ], [ɔː], & [ʌ] (following the pronunciation in RP), a more phonetically accurate transcription would be [ɔ], [oː], & [ʌ].

Someone familiar with British English and this traditional (but outdated) transcription could easily run into the same confusion you have when feeling their way around the IPA vowel chart, as if they try to round their CUT vowel they don't arrive at a short version of their CAUGHT vowel, but rather at their COT vowel. This doesn't mean that the ʌ is actually fully open (and not open-mid), but rather than the transcription of the COT and CAUGHT vowels is not accurate to standard modern British English pronunciations.

  • Erm, In think you mean /ɒ/, /ɔː/ & /ʌ/ not [ɒ] etc - which brings us to the major point that language-specific transcription systems do not use the IPA symbols that they borrow merely to faithfully represent the sound that they would stand for if used in the IPA system proper -- a point that Geoff keeps omitting in his vlogs. The IPA-proper symbol for STRUT would be [ɐ], not [ʌ] btw. Apr 3 at 11:25
  • @Araucaria-him I debated the bracketing issue. Per the IPA themselves square brackets are for narrow transcription, and slashes for broad, but common use outside of phonetics is to use square brackets for phonetic transcription and slashes for phonemic transcription. I decided to follow the latter convention and use square brackets as we are discussing the phonetic transcription. As for the STRUT vowel this is not the case, [ɐ] was accurate (and used) for some intermediate stages of RP, but both earlier & later RP as well as SSBE have a back [ʌ]
    – Tristan
    Apr 3 at 11:53
  • Lindsey does also address the fact the symbols are not always being used solely to faithfully represent their sound. He points out that the transcriptions given are actually pretty accurate to the speech they intend to describe, just a slightly broad transcription (e.g. no raising/lowering etc diacritics), and when occasional typesetting concerns come in he mentions this too
    – Tristan
    Apr 3 at 12:01
  • As discussed in this answer, whilst it is true that transcriptions do not always faithfully represent the pronunciation, the choices aren't arbitrary, and with a few exceptions (e.g. typesetting concerns, something that's much less of a problem now in the unicode era than back in the days of typewriters) people generally do aim for a reasonably accurate (if not always precise) transcription, but that such a transcription can become inaccurate
    – Tristan
    Apr 3 at 12:02
  • I don't think I explained properly. Within [ ] there is no language-specific convention regarding symbols. If you had turned up for your IPA exam and for the word rug or root written [rʌg] or [ru:t], you'd basically fail it. Because English /r/ is realised as a postaleveolar approximant and not an alveolar trill, you need [ɹ] and not [r]. In '[ru:t]', the GOOSE vowel is short due to pre fortis clipping, but the length marks on [u], which are optional in square brackets, say very specifically that it is long (even if it was long as in rude, the length marks could be missed off). Apr 4 at 8:04

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.