User:Soap/stress

Some people claim that in AmEng, /ʌ/ and /ə/ are the same vowel, but different people claiming this make different arguments for it. This is often a sign that they are all wrong, even if they are the majority.

An example of this in the currently ongoing BP discussion involves enwrap ~ unwrap. One person says that they both have schwa, and another person says that unwrap has schwa and enwrap has some other vowel, possibly /ɛ/. So these people don't even agree with each other.

So far as I know, the claim that /ʌ/ and /ə/ are the same phoneme was always a sort of linguistic trivia, based on their having no minimal pairs. I'd never seen the claim that they are actually pronounced the same at the surface until a year ago. The claim has several flaws:

1. The IPA /ə/ vowel is not particularly close to /ʌ/. Indeed the schwa vowel is a neutral vowel. Several vowels equally close to IPA /ə/ are /ʌ/, /ɪ/, /ɛ/, and /ʊ/. If someone were to say out loud [pət], how would a listener know whether the word was putt, put, or something else? If you're thinking it's obviously putt because that is the only one with a stressed schwa, which IPA vowel are you thinking of? Are you saying that putt pronounced out loud is actually IPA [pət], or that it's [pʌt] where [ʌ] is a stressed allophone of the /ə/ phoneme?
   1. If you're saying that the vowel is [ʌ], you're accepting that [ʌ] is an allophone of schwa. But why choose [ʌ] as the allophone when each of [ɪ ɛ ʊ] would do just as well?
   2. If you're saying that the vowel is [ə], I wonder why this phenomenon seems to have just been noticed in the past few years, or possibly even in the last one year. The Geoff Lindsey video linked early in the thread, which has been claimed to support the merger of the two vowels, in fact says in its first 13 seconds that the vowel varies a bit based on stress. This is the old familiar argument I've been hearing for more than twenty years: that the [ʌ] and [ə] sounds are allophones of the same phoneme, and therefore, by means of a technicality, they can be merged into one phoneme. But why do this? Many other languages have reduced vowels, and so far as I know, nobody has tried to merge them with any particular full vowels. English is very similar in setup to German, and I've never heard anyone argue that entdecken should be transcribed /ɛntˈdɛkɛn/ or /əntˈdəkən/.
      1. And why STRUT? It is just one of four lax vowels that could be chosen here. If we are basing our choice on acoustics, the only stressed vowel that approaches [ə] in GenAm that I'm familiar with is the unrounded allophone of /ʊ/ in the small word set pull full bull. But this is a tiny word set, and I think even this could be better transcribed as a syllabic /l/. I see no reason to unify the unstressed schwa in English with any stressed vowel.
2. If the STRUT vowel is analyzed as /ə/ when stressed, it presents no problem, but how would you handle the same vowel in unstressed position? We have two words, enwrap and unwrap, that would merge in pronunciation if we analyzed both with schwa. This is just one particularly convenient example. My guess is that some people would say that enwrap begins with an unstressed /ɛ/, but as of 29 October 2023 we continue to list /ə/ as a valid pronunciation for the vowel of the prefix en-, so analyzing CUT as schwa requires either 1) removing this, 2) marking it as non-US, or 3) explaining how we could have two prefixes both pronounced /ən-/ that mean different things, yet people aren't confused.
   1. One way around this is to posit secondary stress. I'm not sure, but my guess is that these people would put a secondary stress on unwrap. But why is it that secondary stress evolved in AmEng un- while other prefixes remained unstressed? Are we to suppose that in the process of shifting /ʌ/ to /ə/, all unstressed instances of /ʌ/ gained secondary stress? If so, would it not be simpler to just revert to the original analysis that /ʌ/ is a separate phoneme? (This is the coincidence I refer to below.)

All in all, I see a lot of mental gymnastics in this argument ... with people (who don't even agree on the other main points) all finding ways to repair the argument that American English has no /ʌ/, whether it relies on positing secondary stress everywhere an unstressed [ʌ] occurs (and perhaps ONLY where [ʌ] occurs), or shifting some /ə/ to /ɪ/ to eliminate collisions, or saying that the vowels are actually distinct after all but that [ʌ] is just the stressed allophone of /ə/ and therefore American English, unlike all other Germanic languages, has a stressed schwa.

other info

Reagan's rayguns

claims raygun, with secondary stress, against Reagan without, proves that secondary stress is phonemic. This is supposed to also prove that /ʌ/ is /ə/, though it seems like a circular argument to me .... if we start out by assuming /ʌ/ is /ə/, then anywhere they contrast we would need to posit secondary stress for the [ʌ] allophones. If we dont start out assuming that /ʌ/ is /ə/, of course, this means nothing. The raygun argument would make more sense (though still a weak argument; see above about the "coincidence" of secondary stress) if it were an ordinary word inherited from Old English, but it is clearly a two-word phrase ray gun, and two-word phrases nearly always have secondary stress. Essentially this means nothing.

This argument could even be used against the original intent, making it weaker than nothing, if I'm correct that the very reason this word pair was singled out was because the supposed [ʌ] allophone in raygun is what supposedly proves that it has secondary stress. I can't assume this, as the people supporting the merge might claim that they actually pronounce it the same as Reagan but for the greater stress on the second syllable. This "and even better, i turn the argument on you" response is not needed however. ray gun being two words is sufficient to prove it is not a minimal pair.

This, as above, is separate from the unresolved circularity of the claim that having secondary stress proves that [ʌ] and [ə] are allophones, since anyone can just posit secondary stress wherever it is needed to make the analogy hold, even if it requires assuming that secondary stress arose for no obvious reason in prefixes like un- which would be unstressed in British English and in presumably all other dialects of English.

permit

while not relevant to the narrow issue above, apparently some people believe secondary stress exists even in non-compound words, like permit (noun). if this is real, it would mean that permit is a homophone of purr mitt for these people.

anyway

I may be slowly losing my hearing, as I find myself unable to understand audio clips posted on social media. This is a natural effect of aging, and since most of the people I associate with are my age or older, I may not be aware of it. It's possible that young people really are pronouncing these words with [ə] and not [ʌ].