STRUT for Dummies

In a previous post I discussed the FOOT vowel of SSB, the rounded close-mid central vowel [ɵ]. This post looks at the central area below FOOT, a large area of the vowel space where we can find considerable variation in quality with little functional load in terms of word contrast.


This variability in lower-central vowel quality is not a recent phenomenon. For the old prestige accent Received Pronunciation (RP), Daniel Jones described the STRUT vowel as ʌ, “the HALF-OPEN BACK UNROUNDED vowel” (An Outline of English Phonetics; I’m using the downloadable first American edition of 1922). This is Jones producing his secondary cardinal 6, [ʌ]:

His own STRUT vowel sounds more central, [ʌ̈]:

And many RP speakers had a STRUT vowel still further from [ʌ], as exemplified in a clip which recently received attention on, of the first ever broadcast in 1940 by the then Princess Elizabeth. We can hear her [ɐ] for STRUT in just, love, comes:

This [ɐ] vowel – which was quite widespread, as old recordings and films tell us – clearly overlaps with the international a vowel of Spanish and innumerable other languages. Jones acknowledged a as a (non-RP) pronunciation of STRUT, and even recommended it as a second-best realisation for foreigners:

If all efforts to obtain the precise sound ʌ fail, the best substitute is a, which bears a considerable resemblance to ʌ, and is actually used as a substitute for it in some English dialects (including London).

I agree with Jones’s pragmatic desire to help foreign learners, but [a] is no longer recommendable for STRUT, since it has become the contemporary TRAP vowel, thanks to the Anticlockwise Vowel Shift which differentiates contemporary SSB from RP. (The old RP æ became so stigmatized – or ridiculized – that even the Queen gradually shifted hers, as famously reported by Jonathan Harrington of the University of Munich.)

The less posh speakers of southern England have remained relatively conservative, retaining for STRUT the [ɐ] quality we heard from Princess Elizabeth. Here’s Kirk Norcross, from The Only Way Is Essex, saying “I was a bit gutted actually… I would love to do…”:

For SSB, however, STRUT has moved away from [a] and towards the triangle defined by ə-ʌ-ɑ. Such qualities can be heard from a single speaker. Emma Watson’s STRUT is most commonly a [ʌ̈] not unlike Daniel Jones’s, but it can slide towards [ə] or [ɑ]. Here are two pronunciations of lovely, both with nuclear intonational accents. The first is more [ʌ], the repetition more [ə]:

While this emphatic pronunciation of stunts might be confused with the word stance:

And here Kate Winslet with the same [ʌ]/[ə] variability:


2. commA

Jones defined the final vowel of commA as ə, “a MIXED [= central] UNROUNDED VOWEL, HALF-OPEN and SLIGHTLY LOWERED”, adding that “the average English ə is very similar in sound to ʌ“. He goes on:

It varies slightly in quality according to its position in the word, being distinctly opener when final…
Many English people actually use ʌ for ə when final, pronouncing
bitter, clever as ˈbitʌ, ˈklevʌ. There is no objection to this pronunciation.

The recording of Princess Elizabeth exhibits this commA allophony, with a pre-pausal variant identical to the STRUT vowel, which for her was [ɐ]. This extends to the centring diphthongs; here are mother, Canada, Australia, remember, and share:

Again, we find that similar qualities are still alive and well in London/Estuary. Here’s Kirk Norcross from Essex saying older, singer, this year:

But as with Jones’s advice that [a] is acceptable in STRUT words, I’d resist his “no objection” attitude to the RP/Estuary allophony in commA syllables, which are more homogeneously [ə] in SSB today.


[ə] also continues to be the most common quality for SSB NURSE. Jones transcribed it for old RP as əː, and John Wells’s Accents of English had the same view thirty years ago:

Phonetically, it is a relatively long unrounded mid central vocoid, [əː]

Which raises the question why the symbol “ɜ” is so widely used for it. I haven’t been able to find an explanation of “ɜ” in Accents of English (apologies if it’s there somewhere), aside from the 1979 IPA chart reproduced at the front of each volume, where it’s defined unhelpfully as “Variant of ə”.

“ɜ” is not, I think, a widely agreed-upon phonetic entity. Africanists use the symbol for an “advanced tongue root” (ATR) vowel, whereas the feature ±ATR is treated with diacritics on the current IPA charts. But the IPA now additionally gives “ɜ” its own plot of land on the vowel quadrilateral, immediately below schwa – which seems at odds with Jones’s view that the RP schwa has a jaw position “somewhat lower than in the case of the long əː.” And the current Wikipedia entry on the Vowel contains this curious statement, which shows that phonetic symbols can sometimes obfuscate more than elucidate:

The rarest vowels cataloged are ɜ (has just been cataloged in Paicî and Received Pronunciation English) and ʊ̈ (Early Modern English and Russian).

My impression is that using “ɜ” for the British NURSE vowel is chiefly a transcriptional sleight of hand to preserve the dogma that “schwa is never stressed”. In the reality of English, however, schwa can definitely be stressed. This is the case for the long monophthong of NURSE, and also for the first element of the GOAT diphthong.

I suspect that the “unstressable schwa” idea is popular because it’s simple, easily remembered and sounds authoritative. There may also be the good intention of reminding students about English vowel reduction. But it tends to instill or reinforce an assumption that schwa is a second-class vowel used only in casual or lazy speech, and therefore a low priority in learning. This profound misconception – nothing characterizes a foreign accent in English more than an absence of schwa – is so prevalent that Michael Ashby, Director of UCL’s Summer Course in English Phonetics, makes a point of contradicting it each year in his opening lecture.

A good schwa should be the foreign learner’s highest priority in pronouncing English. And learners aiming at a British accent will never produce native-sounding NURSE or GOAT vowels unless they can produce stressed schwa.

Of course, as I said at the beginning, variability is characteristic of the lower central vowel space of English, and this is found in both NURSE and GOAT. NURSE can be heard more open and back than [əː]; I tend to associate this with younger, relatively posh females. Listen to her said by Emma Watson:



In GOAT, a fronted first element is widely perceived around the world as a key element in the stereotypical posh British accent. It was occasionally used by Princess Elizabeth, and has re-appeared more recently in the “GOAT fronting” which Emma Watson also sometimes uses:

Also common in the first element of GOAT is the same kind of relatively open pronunciation we heard as variants of commA and NURSE. We get this in London/Estuary; here Kirk Norcross says show and I dunno:

Daniel Radcliffe has a GOAT vowel which begins similarly, and can become more open and back under expressive emphasis:



Of course, the variability we hear in commA, NURSE, GOAT and STRUT is unsurprising, since vowel quality across these four lexical sets carries little or no functional load. NURSE, a long vowel, is distinct from the others in quantity. GOAT is distinct by virtue of its diphthongal trajectory. STRUT and commA are distinct from the other two by being short and monophthongal, and have practically no contrastive value with respect to each other: they’re defined in a mutually exclusive way, and alleged “minimal” pairs such as an equal and unequal are not structurally identical.

For many speakers in America, Wales, and England outside the southeast, commA and STRUT are not felt to be different vowels at all – though such speakers may differentiate an equal and unequal by means of stress. This is true of “educated Northerners” like me, and I suspect it lies behind the wide STRUT-variability of CNN’s weather anchor Jenny Harrison, born in Stratford-on-Avon:

Then there’s the ə/ʌ STRUT-variability of the Anglo-American pop-rock music through which many people around the world get most of their exposure to native English:

So, with all the variability across commA, NURSE, GOAT and STRUT, how do we follow Daniel Jones in attending to the needs of the general learner? What guidance can be offered to “Dummies” – in the sense of Wiley’s book series – the vast population of English learners who want phonetic help but not the fine detail which is the domain of specialists?

My basic philosophy is to teach the simplest system consistent with sounding native. So my advice to Dummies for commA, NURSE, GOAT and STRUT is to use ə in all of them: comm[ə], n[əː]s, g[əw]t and str[ə]t.

Like Jones, I teach English schwa as a somewhat lowered mid central vowel. To sound native, foreigners need to learn this vowel and be able to stress it and (for SSB) to lengthen it. And it must be kept distinct from both TRAP and PALM/START. This is not easy for most learners: happy is the English teacher whose foreign students have all mastered three qualities to differentiate Sam, psalm and sum/some. To insist on a fourth quality which has minimal functional value, and which is not necessary to sound standardly native, is in my opinion unnecessary for most learners.

Of course for more advanced students, or for actors aiming at a specific dialect, we can and should cover the “footnote” details: for example, that some speakers have a more open/back vowel in NURSE, or in GOAT, or in stressed STRUT syllables especially under emphasis. But, for Dummies, all four lexical sets can be treated as containing ə.

39 replies
  1. dw
    dw says:

    “[a] … has become the contemporary TRAP vowel, thanks to the Anticlockwise Vowel Shift which differentiates contemporary Standard British from R.P. “

    Do you have any audio to demonstrate this? [a] in TRAP sounds characteristic only of the North of England to me.

    In the accompanying clip of Jilly Cooper, I hear something close to [æ] in “actor” (although the less distinct, unstressed, vowel of her hesitating “and”s is more [a]-like). Do I need to syringe my ears?

    • Ed
      Ed says:

      Even Brian Sewell uses [a] in TRAP, and he’s widely considered to be a posh speaker.

      Even before the change in BBC English, [a] in TRAP was characteristic of 90% of regional English. I don’t think that it’s a good idea to teach [æ] to foreigners, as it gets misinterpreted by native Brits as /e/ rather than /a/.

          • dw
            dw says:


            Although this is surely not a typical contemporary speaker — he sounds like almost a parody (whether consciously or not) of U-RP!

            Even here, though, I hear a sound closer to [æ] than [a] at some places — for example in “quite grand enough” at 0:21

      • Adrian
        Adrian says:

        Although I think there’s still a lot of æ-ness to the short-a vowel of a lot of British-English speakers, I agree that textbooks, dictionaries and teachers are doing their students a disservice by emphasising it.

    • Geoff Lindsey
      Geoff Lindsey says:

      @dw No syringe needed, you’ve just been understandably misled by the widespread use of phonetic symbols that no longer match the Standard British vowels they’re supposedly describing. This is Daniel Jones demonstrating [a], cardinal vowel 4:
      Clearly that’s an appropriate symbol for the Standard British vowel of TRAP as used today by speakers as diverse as Jilly Cooper and Daniel Radcliffe. But dictionaries persist in using [æ], which was only accurate back when TRAP had a higher vowel, as in this 1957 clip from the BBC’s Rag, Tag and Bobtail:
      Or family and gather in the Queen’s Christmas message from the same year:
      By contrast the world’s most common open vowel, the a of Northern British English, Spanish and countless other languages, is neither as front as [a] nor back like the [ɑ] of PALM, but central, and so in IPA terms is strictly [ä] or [ɐ].

      • dw
        dw says:

        It could be that I have lived in the US too long, but [a] (even as articulated by Jones) just doesn’t sound like the TRAP vowel of a contemporary typical non-Northern accent of England. I take Ed’s point about Brian Sewell above, but his accent is by no means typical in any way.

        I mention in an earlier comment that in the Jilly Cooper segment you give in the main article, I hear her “actor” as closer to [æ] than [a]: do you disagree?

        • Geoff Lindsey
          Geoff Lindsey says:

          First, Jilly Cooper. I’ve digitally slowed down the initial vowels of her two utterances of actor in the post:

          I’d say the second one has [a]. But the first one has a slight glide (confirmed by spectrogram), indeed beginning “close to [æ]” (as you said in your first comment) and then lowering to [a]. For purposes of convenient comparison, here are Jones’s cardinal 4 [a]:

          and [æ] demonstrated by John Wells and Peter Ladefoged:

          But it’s clear from other words in Cooper’s speech that she no longer has the classic R.P. æ which she presumably had in the 70s. For example, she repeatedly uses a strongly stressed and with an [a] somewhat backer than Jones’s (the latter being a maximally peripheral reference vowel); one of these begins the clip in the main post.

          Of course Jilly Cooper is an older speaker (75) with a decidedly conservative accent. More widely, TRAP has definitely lowered away from [æ]. Here is the first vowel of actor in the British and American pronunciations of the online Oxford Advanced Learner’s Dictionary, which again I’ve slowed digitally:

          [æ] is an appropriate symbol for the American vowel. But the British vowel is clearly different; to my ear, it’s even further away from [æ] than Jones’s [a].

          For some, TRAP has moved still further back. You may like to check out this comedy clip which parodies the very back TRAP vowel of some young, posh speakers.

          • Lipman
            Lipman says:

            The Queen’s 1957 ash vowel in family is a nice example for the length I find typical of some speakers. I suspect it’s longer than in mainstream RP because it’s the result of a smoothed centering diphthong, the older or alternative U-RP realisation of the ash vowel.

        • Ed
          Ed says:

          My point in mentioning Brian Sewell was that there are people in the south of England who use [a] in TRAP, so it is non-regional. The fact that someone as old and conservative as Sewell uses it [at least some of the time] suggests that it’s been around for a while.

          Can you find anyone from the north of England who uses [æ] in TRAP? If you can, you might end up on a TV programme of greatest discoveries.

          • Lipman
            Lipman says:

            No, I don’t mean the bad-lad split but a general tendency for a(n optionally) longer æ, in fam’ly just as well as in <national, even if trisyllabic.

  2. Jack Windsor Lewis
    Jack Windsor Lewis says:

    I can’t think of a better accolade than being appointed stand-in by JCW.
    “the HALF-OPEN BACK UNROUNDED vowel” (An Outline of English Phonetics; I’m using the downloadable first American edition of 1922).
    This was the second edition strictly speaking but it was really a second impression of the first.
    It was the last time DJ used Sweet’s term ‘mixed’.
    The 1932 third edition, which used ‘central’, coming in line with IPA, took ten years to come and was a very fundamental re-write.
    → Ed: You might like to look at my Blog ‘The Bleck Hendbeg Problem #004 at

    • Ed
      Ed says:

      Your blog illustrates what I saying brilliantly. Most Germans have no problem saying [a], so the fact that they end up using [ɛ] for TRAP (and being misinterpreted by Brits as a result) suggests a problem with the way that they are taught English pronunciation.

  3. Tom
    Tom says:

    The Upton IPA system is used in the Shorter OED dictionary (both the paper version and the WordWeb Pro version) and in the on-line version of [N]ODE:

    (The paper version of [N]ODE does not include pronunciation of (common) words)).

    Most speakers of Spanish, Italian, Swedish, German, Slavic languages pronounce CUT as [kät] (as in Australian English, Cockney and in Pittsburgh), CAT as [ket] (as in NZ English and the Northern cities region of the US (Chicago, Cleveland, Detroit, Buffalo)).

    Speakers of Norwegian, Portuguese, French, Indian languages (except for Bengali) don’t have problems with the [ɐ] vowel, so it’s easy for them to pronounce CUT as [kɐt], or with a more raised vowel, typical of some varieties of American English (Ohio, Texas) and Wales, so LOVE can sound like LERVE in England.
    (LERVE is used as an informal spelling of American pronunciation of LOVE in British chat rooms and forums).

    Norwegian ø and Portuguese (denasalized) â are not unlike [ɐ].
    [ɐ] is the inherent vowel in all Indian languages, except for Bengali which uses [o] as the inherent vowel.

  4. Mitko Sabev
    Mitko Sabev says:

    Thank you for this very interesting discussion, and everything else that you post here.

    [STRUT and commA] are defined in a mutually exclusive way, and alleged “minimal” pairs such as an equal and unequal are not structurally identical.
    For many speakers in America, Wales, and England outside the southeast, commA and STRUT are not felt to be different vowels at all – though such speakers may differentiate an equal and unequal by means of stress.

    In this context, how would you treat the variants of words like the following?
    product /ˈprɒdʌkt, ˈprɒdəkt/
    industry /ˈɪndəstri, ˈɪndʌstri/
    (John Wells has this last one, /ˈɪndʌstri/, as a non-RP variant).

    You could claim that the second syllables in /ˈprɒdʌkt/ and /ˈɪndʌstri/ bear secondary (or any kind of non-primary) stress. This would probably do for the purposes of phonemic transcription (although it is rather unusual to assign secondary stress to a syllable immediately following the one with primary stress). On this interpretation, the difference between the two variants of both words would be the same (mutatis mutandis) as the difference between the vowels in the second syllables of selfish–shellfish, or panic–picnic.

    Don’t you think, however, that in /ˈprɒdʌkt/ v. /ˈprɒdəkt/ there’s also a difference in the quality of the second vowel? Or does this difference only exist in the southeast of England?

    I know this may sound like splitting hairs, but I’m genuinely interested in your interpretation.

    Thank you.

    • Simon Bramwell
      Simon Bramwell says:

      I’m no expert, but for those who can’t hear the difference between between ə/ʌ, I think there is no clearer illustration of the difference than comparing an SSB pronunciation of ‘nurse’ with a modern Scottish pronunciation of ‘nurse’. Clearly quite distinct vowels.

  5. Jahid Akon
    Jahid Akon says:

    Great analysis! Makes good and clear sense to me. But what about the last syllable of ‘usual’? [ˈjɵwʒɵwəɫ] or [ˈjɵwʒɵɵɫ] or [ˈjɵwʒɵɫ]?

    • Geoff Lindsey
      Geoff Lindsey says:

      This is a very good question. I think the smoothing process as I describe it applies to stressed diphthongs; therefore [ˈjɵwʒɵɵɫ] should not result. I would consider [ˈjɵwʒəɫ] to be a weak form of usual rather than an output of smoothing/compression.

  6. Ed
    Ed says:

    Your recent post on Eurovision has led me to read this again.

    I’m not clear on why ʌ came to be used for RP STRUT. Already in the 19th century, Joseph Wright was using ɐ for this vowel, which is what came to be the main variant on the BBC. Why did Jones use ʌ? I understand that he concentrated initially on transcribing his own speech. Was he in a minority of ʌ users in an area of mostly ɐ usage?

    • Geoff Lindsey
      Geoff Lindsey says:

      Sorry I missed this question, Ed. I don’t know the answer, but I certainly hear ʌ rather than ɐ in the bits of Jones’s speech I’ve heard.

  7. Al-Qadir Moghadam
    Al-Qadir Moghadam says:

    I have a dummy question: if I pronounce the vowel in ‘bad’ as /bäd/, will it sound wrong? /æ/ is somewhat difficult for me, and I can’t hear the difference between /ä/ and /a/.

    • Geoff Lindsey
      Geoff Lindsey says:

      (Strictly, Al-Qadir, you shouldn’t be using slanting brackets to discuss precise phonetic qualities.) In strict IPA terms, [ä] is further back than the typical SSB TRAP vowel, and may be heard as STRUT. Remember that in reality the vowels of languages are not precise points, but rather areas of the auditory-acoustic vowel space, as discussed here. TRAP and STRUT are areas which for many speakers overlap somewhat. If you want to train your ear to distinguish [a] and [ɐ] (which is more or less the same as [ä]), you could use the synthesized audio demonstrations here.

      • Al-Qadir Moghadam
        Al-Qadir Moghadam says:

        Yes, I should have used the square brackets, I forgot that. I listened to the audio files in Wikipedia and I think I can see the difference now. The [ä] sounds slightly muffled compared to [a], and is really closer to [ɐ], [a] sounds more clear and open. I think I’ve been using [a] all along without knowing, because my vowel in ‘cat’ certainly sounds different from the one in ‘cut’, and I do use [ɐ] for ‘cut’. But I think the [ä] in my native language is significantly more open than the [ɐ] in ‘cut’, I just haven’t noticed it was a bit muffled compared to [a]. However, if the speaker had /ə/ for STRUT, wouldn’t the difference be less significant?

        • Geoff Lindsey
          Geoff Lindsey says:

          Sorry, I’m not sure I understand your last question. A major point that I was trying to make in this post is that many non-natives use their ‘a’ for STRUT. This was okay back in the days of RP, but now it may well sound too much like TRAP. TRAP should be kept ‘clear’ and ‘bright’, STRUT should be more ‘muffled’ – in fact it can even be pronounced as schwa, which is how millions of natives pronounce it.

  8. Al-Qadir Moghadam
    Al-Qadir Moghadam says:

    I was asking if [ä] could be acceptable for the TRAP vowel, because it seems that for those speakers who have [ä] in their native inventory, like me, most of the time they hear the TRAP vowel as closer to [ä], while the STRUT vowel usually sounds closer to sth like /ɜ/, or a schwa. It seems this is true for Italian, for example, as I saw an Italian confirming this in a language forum. As long, of course, [ä] is pronounced open and clear, and different from [ɐ], and since the schwa is quite common for the STRUT vowel now, I think the difference between the two would be quite noticeable. I’m asking this because the [a] vowel, as a front vowel is somewhat rare, as it is [æ], so this would make it easier for learners.

  9. Boris
    Boris says:

    There are speakers of modern Standard British, such as Emma Watson, as shown by one of the recordings in your post, whose default realisation of GOAT (but not MOUTH) seems to have a fronted second element, a vowel somewhere in the range of [ʏ] or [ʉ]. How does this phonetic development interact with your postulated linking /w/? Does the fronting of the second element disappear when GOAT is followed by another vowel?

    • Geoff Lindsey
      Geoff Lindsey says:

      I wouldn’t say that I postulate “linking /w/”, as that suggests the insertion of the consonant phoneme of way and away. In sequences like go‿out and see‿it, the linking w or j sound is, I would suggest, intrinsic to the first vowel: gəw‿awt, sɪj‿ɪt. You rightly point out that GOAT may have a fronted second element (as may GOOSE). For such speakers, broad symbols like əw and ʉw are not phonetically precise. To answer your question, I’d say tentatively that speakers who have a very front offglide pre-pausally may have less fronting before a vowel, and the degree may depend on the following vowel. An acoustic study of young speakers saying eg so, so easy, so early, so awful; too, too easy, too early, too awful would be interesting (I can suggest it to colleagues as a student project). But there’s variability, and I’ve known at least one speaker who had a very j-like glide in eg No‿I’m not.

  10. Jack
    Jack says:

    What do you think of Well’s suggestion of /ɐ/ for contemporary STRUT? It’s like a more open schwa. A lot of speaker seems to have that one in open syllables, as in better [ˈbɛtɐ]. Of course, ɐ ≠ ä ≠ a.

  11. Egg
    Egg says:

    The final schwa still sounds a bit lower than elsewhere to me, and pretty much identical to STRUT. So you have the same vowel in both syllables in words like butter, cover, mother, thorough etc.
    This is why, despite the fact that both English bitter and German bitte are transcribed as [bɪtə], German bitter sounds basically the same as its English friend: [bɪtɐ]


Trackbacks & Pingbacks

  1. […] to Geoff Lindsey, in Standard British English the STRUT vowel is actually [ə]. In his blog post on the STRUT vowel, he notes that old RP speakers (esp. Queen Elizabeth) use [ɐ] for the STRUT […]

  2. […] If you really want specific advice about phonetics, I’d say the following. Since I definitely know of some accents where /ʌ/ is more “back” than /ɑ/, but I don’t actually know of any accent where /ʌ/ is “lower” than /ɑ/, it’s probably best to focus on vowel height as the distinguishing factor and aim to produce /ʌ/ as a central-ish non-low vowel: somewhere around the quality [ə] that you will have to be familiar with anyway to use for reduced vowels in unstressed syllables. Many (although not all) American English speakers identify the quality of /ʌ/ with that of /ə/, so this should be a fairly “safe” choice. This will probably be at least somewhat “higher” than Hebrew /a/. This is pretty much the advice Geoff Lindsey gives in his blog post “STRUT for Dummies“. […]

  3. […] RP is highly variable so /ɐ/ probably occurs sometimes, and especially at the ends of words. See STRUT for Dummies for more info on its realisation in […]

  4. […] “u”-Laut, um den es hier geht – und für den englischsprachige Phonetiker den Begriff STRUT-Vokal geprägt haben – ist ein solcher Fremdkörper. Genau betrachtet ist er weder Bestandteil […]

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *