As you are probably aware, our contemporary English content is now available through (, and our old English dictionary site no longer exists.

As a result of this, this forum is now closed.

The English dictionary community team would like the opportunity to say a huge thanks to all of you who participated by posting questions and helping other community members.
We hope this forum was useful, and that you enjoyed being a part of it.

If you would like to get in touch with any OED-related queries, please write to
[email protected]

And if you would like to contribute suggestions to the OED, please do so by visiting:

Thank you very much indeed, and good bye!
The community team


  1. There seems to be a mismatch between the spoken pronunciation of this word and the phonemic transcription, as given in the entry under susurration/susurrus, at: . The recording of the spoken pronunciation seems to be (to my ear) /ˌsjuːsʌ'reɪʃən/ . The written transcription, however, is: /ˌsuːsʌˈreɪʃ(ə)n/. Maybe /sju:/ is right; or /su:/ is right, or maybe both. It is confusing having this mismatch.

  2. The entry points out that this word is used in literary style. I wonder how the dictionary makers go about deciding the pronunciation of a word such as this. Used mostly in literary style, and therefore mostly written form, examples of its pronunciation must be rare, as the word is. Come to think of it, there are probably words in the language, rare and literary, which have never been spoken in recent times, at least not since the invention of mechanical means of recording. In a descriptive approach to dictionary making, how is their pronunciation to be ascertained? And how described in the dictionary entry?


  • DavidCrosbieDavidCrosbie ✭✭✭
    edited October 2018

    The OED has /s(j)uːsʌˈreɪʃən/ and the sound file matches.

    I looked in John Wells's Longman Pronunciation Dictionary. He doesn't include susurration, but he does repeatedly report similar variation.

    For example suit is listed with suːt as the more common British pronunciation with sjuːt as the second.

    The same is true of suzerain and words in super-, including superb.

    Generally speaking a syllable spelled CONSONANT u CONSONANT reflects a historic * *juː** sound. After many consonants this is still the case in British RP.

    Some words lost the j very early, typically

    • after r — hence ruːl despite the spelling rule
    • after palatal ʃ/ʧ and ʒ/ʤ — hence ʃuːt despite the spelling chute
    • after CONSONANT+l — hence bluː despite the spelling blue

    The so-called jod was lost more generally in other dialects, notably in American accents. In RP, yod-dropping has extended to after s — as in Susan — but more conservative speakers otherwise retain the jod.

  • I've found some more words with initial su- spelling. Some are from Latin such as sucrose, supine. But there's also the recent borrowing Suez and at least two from Old French: sue and suet. (Suit is another.)

    Wells lists all of them with or juː. For sue he lists sjuː as the more common.

    The Latin derived words have always been spelled with su-, but the OED shows interesting historical spellings for the Norman borrowings. These include (along with many other spellings)

    • suit — seute, sewte, siyte, sywete, sywyte, siewte
    • sue — siwe, sewe, seue, seu, siu, siwi, sywi, siwy, siue, seuwe, seewe, sieu, syew
    • suet — sewet, siyte, sywete, sywte, siewte

    According to John Wells in Accents of English our 'long U' sound stems from
    1. Middle English words with 'long O'
    2. Middle English words with
    * 'falling' diphthong ɪu — still heard in the Welsh pronunciation of threw
    * 'falling' diphthong ɛu

    This happened together with the 'Great Vowel Shift', so the end result had

    • sound quality u
    • length from ME long O words
    • j-glide from ME diphthongs

    There's plenty of evidence in the form of written forms before standardised spellings. It's not just conjecture before

    the invention of mechanical means of recording

    According to Wells, words which now or previously had this juː sound are typically spelled with

    • ue
    • u CONSONANT e
    • u
    • eu
    • ew

    In many cases it works the other way: the spelling is a clue that the word is pronounced with juː in British RP — or used to be.

  • Thank you for the reference to the OED which I had not looked up. However, the online dictionary at is, I imagine, to be regarded as self-standing and independent, its users not needing routinely to refer on to the OED for an answer to their inquiries.

    Where there is a transcription, the implication of the dictionary entry is that the recording is an accurate rendering of that transcription. In the OED this is the case, as you point out, although there are two alternatives given within the one transcription. This is by means of the brackets surrounding the symbol /j/ : thus: (j) . The recording thus accurately exemplifies one of the alternatives.
    At the case is altered. The transcription is /suːsʌˈreɪʃ(ə)n/ and the recording exemplifies something different, viz, /ˌsjuːsʌˈreɪʃ(ə)n/ This is what is confusing to the user of the online dictionary.

    Two remedies suggest themselves. Either, the entry gives two pronunciations through the (two-alternative) transcription /s(j)uːsʌˈreɪʃən/ with a recording of one of the alternatives (as in the OED); or, if only one pronunciation (through a single-alternative transcription) is to be given, then the recording should match accurately that given transcription.

    That’s enough for a blog comment for now, I think. It may be that your most interesting points on yod-dropping, historically and currently, and spelling and pronunciation, may be deserving of a separate post!

  • Where there is a transcription, the implication of the dictionary entry is that the recording is an accurate rendering of that transcription.

    In an ideal world, yes. But that would mean employing specialist readers and specialist recoding-producers with more than basic training in Phonetics.

    Many teachers (and, I presume, many actors) have learned to read IPA transcription and thereby recognise the word transcribed. Some teachers (fewer actors, I suspect) have also learned to write transcription.

    Hopefully, the majority of IPA-literate teachers/performers/producers can recognise when a transcription is grossly at odds with their own preference when they believe they're speaking RP. But su- is a grey area. Individual pronunciations vary, even among RP-speakers. I can't believe many listeners notice whether a speaker says sujː or suː in these words.

    The problem is recognition — successful recognition, that is, not failure of recognition. The reader recognises the word despite the IPA. The hearer recognises the word despite the individual pronunciation. In this case, the performer recognised susurration — if only as a word comparable to sue, suit and suet — and pronounced it accordingly. The producer recognised the word and accepted a final version for clarity, sound level etc.

    [An aside.
    I once worked for an outfit that used phonetic transcription in teaching materials — to the point that students could do pronunciation homework without the support of a teacher.
    The teacher's role was to prepare for and administer tests, though we were allowed a tiny bit if actual teaching.
    A crucial skill was to assess a students memorised pronunciation against a phonetic text. To avoid the recognition problem, the teacher-training course involve reading phonetic transcription syllable-by-syllable forwards and backwards against a metronome.]

    To show how much of a grey area this is, compare
    Both offer s(j)uː and an actor's voice with (to my ears) some sort of /s/ and some sort of/j/.
    Offers suː and an actor's voice that (to my ears) is a palatalised [sᴶ] if not a fricative [ɕ]

    1. The reader recognises the word despite the IPA. The hearer recognises the word despite the individual pronunciation.

    This is a possible scenario as to how the error came to be made, but it does not controvert the notion that the error should be fixed. The point is that in a dictionary which gives accepted pronunciations of words, the recording and the transcription should match; or, put differently, that a recording should exemplify the given transcription. This is not the case at as I have shown.

    1. There is no problem in my experience, with having actors (or teachers) produce required pronunciations for recordings, as I have seen at recording sessions when actors were required to do exactly this. The author is on hand to ensure accuracy, and actors are able to alter renderings, without (again, in my experience) any real difficulty.

    2. Many teachers and teacher-trainers of English, native and non-native speakers, use the OED (or one of its spin-offs) for advice on pronunciation. Many know phonemic transcription. I think most will easily distinguish between /su:/ and /sju:/ and spot the difference, especially if pronunciation is one of the key points they have gone to the dictionary to find out about. They will be listening carefully for it. And they will use the transcription to help them. At
      the transcription does not assist, as it is contradicted by the recording. A recording-transcription mis-match is confusing and unhelpful. Dictionary users do not know whether to follow the recording or the transcription, or both. And, worse, they start to doubt the evidence of their own ears. (It is also injurious to the reputation of this old lady of English letters, the OED).

    3. The references you point to show that there is clearly a grey area. But it has been dealt with, accurately I am sure, by Professor Wells in the Pronouncing Dictionary, as you mentioned; and accurately dealt with by the full OED under susurration. So it’s all doable. But it is not done in the online version at: It can and should be done there, too.

  • I think, Barry, we can both agree that a mistake was made. I believe the mistake was to omit (j) in the onscreen transcription.

    I suspect OUP's problem is in the addition and integration of the transcription/sound file facility into a traditional book-based framework. Even the use of IPA transcription is relatively new to mainstream general dictionary production.

    And even the OED is not immune to error. I recently had cause to look up the General American pronunciation of pitta|pita (the Greek bread). The AmE transcription offered is ˈpɪdə (indicating a 'flapped' t), but the sound file has a 'higher' i vowel. This was for a comment I posted on the blog of the highly expert Lynne Murphy, who specialises in BrE/AmE variation. She questioned both the vowel quality and the t value. In her opinion the usual American pronunciation has the FLEECE vowel and a unflapped t.

    Personally, if I want to know the usual RP pronunciation of a word, I go to a specialist dictionary. I own only two (plus a BBC dictionary for proper names) , but my English Pronouncing Dictionary is somewhat dated, so I turn to the Longman Pronunciation Dictionary.

    I fully expect OUP and other publishers will improve their online service, which is clearly the dominant form if the future. They'll need to iron out errors as they are spotted and reported. So I suppose it's up to you and me, Barry.

  • @Simone, are you the appropriate person to ask to pass on feedback concerning susurration and pitta|pita to the relevant teams?

  • I believe the mistake was to omit (j) in the onscreen transcription.
    Yes, this seems the easiest way of looking at it, and leads to the easiest way of fixing it.

    So I suppose it's up to you and me, Barry.
    I suppose it is; we’ll do our best anyway, David.

  • Hi @BarryCusack, @DavidCrosbie
    My apologies for joining this discussion so late!
    @BarryCusack, many thanks for flagging this up - I've got in touch with the editors, hopefully I'll be able to get back to you soon!
    @DavidCrosbie, I'm not the person able to answer these questions, but I'm certainly the point of contact. So if you ever notice anything else that may be wrong, do flag it up, please! :)

  • Hi @BarryCusack, @DavidCrosbie
    I hope you're both well.
    I've now heard back from my colleagues in the editorial team, here is the reply I've got:

    Susurration: Our audio pronunciations should indeed be an accurate rendering of the transcription. As is noted with reference to the OED pronunciation, the /j/ is an optional element in the first syllable, i.e. “syoo-“ or “soo-“ are both considered possible pronunciations. There is an error in the ODO entry, though, because a j-less transcription is paired with audio which does have /j/. The soundfile is also quite an old, low quality one.
    The mismatch will be fixed, thank you for bringing it to our attention.

    The “literary” label belongs to the entry in general and doesn’t pertain in particular to the pronunciation. When researching the pronunciation of very rare words, we would consult older reference works, consider the etymology and other aspects of the word, look for suitable analogues, perhaps discuss with expert users. Poetry metre can also offer clues to stress placement.

    Pit(t)a: The OED entry gives British and US pronunciations only with short /ɪ/, but the US pronunciation ˈpitə, to rhyme with “Rita”, is much more common. I will amend the entry for the next OED update.
    Again, thank you for bringing it to our attention.

  • For susurration, a most happy outcome. Thank you very much, Simone.

Sign In or Register to comment.