Rufio wrote:It's not a difference in the phonology so much as a difference in the stress of the sentence of the whole, because of a change of emphasis.
I agree that sentence stress enters into the equation-- I won't run through the paradigm, but any word in the sentence 'some dogs are black' could easily carry sentence stress. But if you will listen to yourself (or to some other folks), I think you will hear /sm/ in 'I know some people from Japan' and /səm/ in 'everything is caused by some other effect'.
However, I really don't want to belabour the issue. There's no right answer: Stu is right because 'some' is not on the Article List; you're right because stress is an ongoing variable which is a complex dependent on each context. What is interesting to me is:
(1) the consistency of the pronunciation of 'some' (/sm/) as the unstressed plural indefinite article, which suggests that at least the differentiation of the word is in process in the language.
(2) the way in which it completes the article set (an egg, the egg, some eggs, the eggs). As both a teaching device and as a bit of language analysis, it immensely simplifies and clarifies this particular grammar point to separate the 'some' that means 'more than one' from the 'some' that means 'not all'. Much of my ongoing concern is finding the 'rules' to which the language adheres and teaching them, while eliminating or minimizing those (like the 'some/any' declarative/interrogative 'dichotomy') that students have learned but which produce as many exceptions as examples in authentic language.
The two 'some's I am calling homographs are one entry in Webster's ('1some'-- and if you have a Webster's, note the pronunciation comment). I would also be interested in your view on 'that', which are entries '1that' and '2that' in Webster's-- 'that' /thæt/, the relatively stressed demonstrative adjective/pronoun, and 'that' /thət/, the conjunction/relative pronoun. I find that the individual pronunciations and the difference between the two is much more consistent than the dictionary pronunciations would suggest. I think that they are, in spite of their common origin, now two different words that share little but orthography, and this is the direction in which the two 'some's are evolving.
And thanks for making this an interesting discussion-- it helps me clarify and correct myself also!