All my dictionaries indicate two phonemes:
They simply sound different, regardless.
>"As I've said, there's no difference between these two vowels, no matter what your dictionary says." "...apparently these lexicographers feel otherwise.">
I'm not prepared to question lexicographers over your brash/haughty claims, but the day when/if you author a dictionary, I might consider your authority over the long established references.
Maybe some ESL students do have problems with "sit/seat", "live/leave" etc., but to me it seems pretty hard to confuse those sounds.
The dictionaries referenced are not distinguishing between phonemes, as far as I can tell, although it's hard to be sure with their archaic pronunciation keys.
I don't recall mentioning my experience with writing dictionaries, so speculation in that domain is risky.
Anyway, ESL students have problems with "sit" and "seat" if their native languages do not make the distinction between these two vowels. This is the case for French students, because French doesn't have a vowel with the sound of /I/; they usually confuse it with /i/, which they do have. Many other languages have both sounds, of course.
ESL students always have trouble with phonemic distinctions that don't exist in their own languages. Those are the areas that one must work on when teaching them. For French students (a majority of my students), the distinction between /i/ and /I/ is a major problem, as are /T/ and /D/ (neither of which exist in French).
There are a bunch of other sounds the ESL learners have problems with, no matter the time or effort. One year or 15+ years (in the case of my flatmate) in a country has little effect on this outcome. I'd like to know if there is a way to turn the tables on time/effort and speech.
In my accent, (and probably in non-rhotic English in general) there are 3 phonemes in the following:
I would have thought that in rhotic English, this would be the case, even with an 'r' getting in the way. If not, then which 2 words above share the common vowel sound?
In standard pronunciations, the first word has a vowel (/I/) different from the other two (/i/). I'd transcribe them phonemically as /fId/, /fid/, and /fird/ in American English. In RP, I'd transcribe /fId/, /fid/, and /fi@d/, respectively (but I'm not an expert in RP, so I usually teach only GAE).
Other differences are not phonemic unless they serve to distinguish significant numbers of minimal pairs in real-world utterances. The number of true phonemes thus varies slightly, but in general regional distinctions don't count, as they distinguish too few words and in too trivial or unrealistic a way to be significant.
In RP, they are miminal triples (3 different phonemes) as indicated:
* bid, bead, beard
According to my American references:
"fid" and "feared" actually share a common representation, ie. /fid/, /fird/,
"feed" is represented with a different vowel, ie: /fèd/
On a side note :
To my senses, the non-rhotic English "beard" is the closest to the French "bide". (being a native to both languages)
Anyway, getting back to spelling, I was only concerned with the 5 to 10 thousand "grassroots" words that the average person can potentially use/need on a daily basis to get around. This is about 1% of the "million words" mentioned, but it's obviously the most critical, and whether its poor orthograpy (in terms of logic) is addressed is another matter.
I didn't give the other 99% much consideration, but how many people possess a workable vocabulary of 1 million words or even a tenth of this? Realistically, relatively few do and those few wouldn't be overly troubled or fussed.
I pronounce the "ee" and "ea" sounds exactly the same (long "i"), while of course I distinguish short "i" (for instance, "meet" and "meat" sound the same, whereas "sit" sounds differently). There are both phonemes in my native language, but they are spelt the same - just like "i".