bet / bet /

bear / ber /

bee / bi: / 

beer / bir / 

(The non-rhotic "r" cannot be shown accurately.) 

I wonder my understanding of IPA for British English may be totally inapplicable to American English.

However, my problem in this particular case is not exactly that of a transcription issue.  My problem is that the voice recording in the MW dictionary website for bet and hair are OBVIOUSLY DIFFERENT in the sound of the vowel, but both of them use the same / e / in the transcription.  Different sound, same transcription.

I can accept this "different sound, same transcription" situation for consonants because they are the "same" sound in the sense that they are allophones of the same phoneme.  For example, clear L vs dark L, flapped T vs non-flapped T.  I have no problem with those because they don't cause any confusion to me.  However, for vowels, I would say this is a big problem in the transcription system.  To me the quality of the vowel for bet is different from that of hair (at least this is what I heard from the voice recordings in MW and native speakers that I met).  This kind of transcription is just plain wrong!  How can a dictionary use the same symbol for two different vowel sounds? ? ?

My problem is that, the pairs of words we discussed before are denoted with different IPA symbols in British English, which is mostly what I learnt, while they are denoted with the same symbol in American English.  My only conclusion is that, they are indeed different in British English and warrant the use of different symbols, but they are much closer in American English and are considered only as variants.

Now I know I knew nothing about American English.

My problem is that, as I have mentioned before, they are indeed denoted differently in British English.  The first one is a simple vowel while the second one is a diphthong / e /.  The question is whether the diphthong / e / ONLY occurs before an R.   If it occurs before some other consonants, such as D, L, T, N ( / -et /, / -el /, / -ed /, / -en / ), using / er / to indicate a different sound from / e / is not a good idea because you'll have problems showing the differences between / -et /, / -el /, / -ed /, / -en / and / -et /, / -el /, / -ed /, / -en / respectively.

Certain combination of sounds just do not exist in English and that's why they don't need to use diacritic to indicate that the P in SP is non-aspirated whereas an initial P is aspirated.  The bottom line is that the pronunciation keys have to be non-ambiguous.  This is the same reason why they don't need to use diacritic to indicate tapped intervocalic T or final dark L.  (Clear L at syllabic final position does not exist.)

I did a search in my pronunciation dictionary and found that the combinations / -et / and / -el / do not exist.  / -ed / and / -en / only exist when the spelling has an R between / e / and the following consonant.  Since British English is non-rhotic, the corresponding American pronunciation would have an R before that consonant, resulting in / rd / and / rn /.  So, / e / ONLY occurs before an R!  As a result, using / er / in American English to represent the counterpart of the British English / e / is non-ambiguous after all!

Kooyeen, many thanks indeed for discussing this issue with me.  Without your insights, I would not have been able to find out the underlying issues that has been troubling me.

EDIT:

It appears that / e / was the result of the influence of the following R sound on the / e / sound.  It is therefore actually the equivalent of / er / in American English.  However, because British English is non-rhotic, the R influence results in the / e / becoming a diphthong, i.e. a schwa is added after the / e / to become  / e /.

EDIT 2:

While adding the above EDITed text, I just found that what I guessed was probably right.

http://www.bl.uk/learning/langlit/sounds/text-only/received-pronunciation/london/

The vowels of youth

Listen first to the vowel sound she uses in the words air, there, their, where, somewhere and parents. In traditional forms of RP, they would be pronounced with a diphthong – that is two vowel sounds. Older RP speakers would start with an <e> sound – as in bed – before drifting into a weak vowel rather like the initial sound in about. This type of pronunciation, also applied to words such as dare, hair and bear, was until relatively recently common in many English accents. The diphthong emerged once speakers began to omit the <r> sound at the end. Speakers throughout the UK once pronounced this <r> sound, but it is increasingly restricted to speakers in the West Country and far South West of England, a small area of Lancashire and most of Scotland and Ireland. It is also present in most US English accents. The <r> sound was initially replaced by the weak vowel at the end of the diphthong, but nowadays most younger RP speakers omit this final part of the diphthong and simply use a long <e> sound – thus shared is pronounced with exactly the same vowel as in shed, only the vowel is noticeably longer. This demonstrates perfectly how successive sound changes can radically alter the pronunciation of a set of words. Most RP speakers, like Michelle, now only distinguish between pairs such as fairs and fez or flared and fled simply by vowel length. Older speakers tend to use a diphthong for the first word in each pair.

It means that the Received Pronunciation that I have been learning (and indicated on all the dictionaries) is only that of the older generation!  What a surprising finding from this simple question of fair-haired!