The correct IPA synmbol for the vowel in BrE 'bed' is not /ɛ/. This vowel, which dictionaries and course books transcribe phonemically as /e/ or/ɛ/ has no symbol on the standard IPA chart. In general BrE RP it is somewhere between the cardinal vowels [ɛ] and [e], slightly closer to the former.
It might not be exactly /ɛ/ (I don't think there's any English vowel that's exactly the IPA symbol it's represented by, except maybe for the schwa), but it certainly is much closer to that than to /e/.
This website has pretty accurate transcriptions (IMO) and uses [ɛ] for both RP and many other English dialects.
Thaks a lot mara,
Best wishes for you,in such final unaccented positions, e.g. in city, lady, memory, /ɪ / is increasingly replaced by a short variety of /i:/ by many speakers. (The contrast between /ɪ /and /i:/ is in any case neutralized in word-final position, so no ambiguity arises from such pronunciations.)![]()
Do you have any sample? Because Australian English has a real [e] or [ɛ ̝] and to me it sounds very different than what I've heard in RP.
Wikipedia has this chart for RP
That's something between [ɛ] and [e̞], yet closer to [ɛ] ([e̞] is midway between the first and second horizontal line).
Does head in "refined RP" sound like this? That would be an Australian [hɛ ̝d].
No, the COD approximates to the closest vowel. That website uses precise transcriptions. And I don't see any [e] or anything close to it in RP.So does the COD. However, most writers in the UK use /e/. It really doesn't matter, so long as the writer makes it clear what the symbol represents. What is important is that phonemic symbols be placed inside //, not [];
The COD does not 'approximate to the closest vowel'. It uses a symbol, which the phonetics editor has chosen, to represent the phoneme, This phoneme has many slightly different forms - the vowel in bet, for example is shorter than the vowel in bed, and my vowel in both words is almost certainly not exactly the same as yours.
When a trained phonetician produces a narrow transcruption of the English word bed pronounced by me, any trained phonetician in the world will be able to produce that word almost exactly as I said it. However, if I give the phonemic transcription /bed/ (or, if I am using different symbols, /bɛd/) then you have only a rough idea of the sound. You know, for example, that most British speakers will use the same vowel that we use in said,head,any,leisure,leopard; you also know that it is a different vowel from the one we used in term, the, pretty, there, me, etc.
When we use the IPA symbols to transcribe phonetically, each symbol denotes one precisely defined sound. When we transcribe phonemically, we can use any symbol we wish, so long as we explain what symbols we are using for what sounds. In practice, most writers use an IPA symbol for a sound close to the phoneme they are transcribing, but many modify with some degree of romanisation - using a letter from their own alphabet which is readily understood. That is why many British writers use /e/ for the bed vowel; most use /r/, when [ɹ/] is actually the closest IPA sound to the phoneme heaard inred and arrow.
For most speakers of BrE there is no exact [e] or [ɛ] vowel. Even if there were, we would be under no obligation to use the IPA symbol in phonemeic transcription.
But then by that logic I could use any kind of symbol as long as I make it clear what it means. I could use /e/ for the vowel in caught, because hey, there's no exact IPA symbol for that vowel.
But, since we have agreed to use the IPA alphabet it's better to stick to transcriptions that are both simple and accurate. I can't agree to use /e/ for bed and then /eɪ/ for say, because of the implication it carries. If typing /ɛ/ is so hard then let's agree on using some other alphabet that contains only the characters on your keyboard, like SAMPA.
There is a great deal of discussion among phoneticians about which symbols are most suitable for phonemes, and the discussion can get quite heated, as it did when the COD moved to /ɛ/ for the sound we are talking about. Because different features are important to different writers, each writer remains free to use the system s/he prefers, though one imagines that anyone coming up with a system in which /e/ represented the vowel in caught, would not have many followers.
In narrow phonetic transcription, the IPA symbols and diacritics are tightly defined. In phonemic transcription we use the system that suits us best.