The vowels in ''father'' and ''bother''?
>>There is no trace of /r/ in the second example--it simply doesn't show up on the phomemic level in that particular phonological environment.<<
I assume you mean phonetic environment. And along those lines, one thing I find really annoying is when people claim that words ending in /r/ in most English English dialects have diphthongs or phonemic long vowels at their ends, as in reality, such cases should be considered as not having such, but rather *just* having an /r/ at the end of said words. I would actually take that further, and extend such to word-initial positions, or *at least* at the ends of individual morphemes, and say that all those diphthongs and long vowels that people commonly make reference to simply do not exist phonemically, and rather are purely aspects of realization alone. The only case when one might be able to truly claim that there are such phonemes is word-internally when not coming before morpheme boundaries, but even then, treating such things as not existing phonemically, and instead treating there to be /r/s in such positions which simply are never actually realized as [r\] probably would result in a simpler analysis overall, if one less intuitive to some laypersons, in the end.
<<I assume you mean phonetic environment.>>
Oops, you assumed right. Thanks for catching that. I was thinking one thing but typed another :)
I meant to say "extend such to word-medial positions" above, for the record.
Travis, I'm not sure I understand what you're saying. Do you want to make changes to transcription schemes for English English?
Lazar, what I mean is that I prefer a more abstract interpretation of the phonemic system in English English dialects, one which eliminates altogether diphthongs and or long vowels caused by non-rhotic-ness, and which instead interprets the underlying phonemic representation as essentially being still rhotic, with such long vowels and diphthongs only being present in surface forms. The reason for such is that it provides a more consistent manner of handling intrusive [r\] in dialects which do not insert it in *all* word boundaries where one word ends with a vowel and another starts with one. Furthermore, it allows a more consistent handling of [r\] at *morpheme* boundaries and eliminates the need for a complex series of diphthongs and or long vowels linked purely to positions where rhotics formerly were present.
In the kinds of naive interpretations of the phonology of English English dialects that I have been seeing, one would essentially need to special-case every single of those "formerly rhotic" diphthongs/long vowels to have phonological rules where [r\] is to be *inserted* in various cases, such as at word and morpheme boundaries where one word or morpheme ending in such is followed by another word or morpheme starting in a vowel, which in the end is effectively more complex than the above, in addition to of course requiring a far more complex vowel system interpretation than in the interpretation which I favor. Second, this interpretation is less consistent with the actual historical origins of such intrusive [r\]s and the continuing presence of [r\] *after* consonants than the interpretation that I favor, as it implies that intrusive [r\]s are being inserted where they previously were not, and that somehow only such [r\]s after consonants originally existed, whereas the interpretation I favor instead treats it as if such rhotics were there all along, but phonological rules have just been introduced so that they are assimilated to preceding vowels or simply derhoticized when not intervocalic, which is far more consistent with the history and overall pattern of such, obviously.
<<Lazar, what I mean is that I prefer a more abstract interpretation of the phonemic system in English English dialects, one which eliminates altogether diphthongs and or long vowels caused by non-rhotic-ness, and which instead interprets the underlying phonemic representation as essentially being still rhotic, with such long vowels and diphthongs only being present in surface forms. The reason for such is that it provides a more consistent manner of handling intrusive [r\] in dialects which do not insert it in *all* word boundaries where one word ends with a vowel and another starts with one. Furthermore, it allows a more consistent handling of [r\] at *morpheme* boundaries and eliminates the need for a complex series of diphthongs and or long vowels linked purely to positions where rhotics formerly were present.>>
Well I agree that there's something to be said for just showing all historical /r\/s in British English, but I don't think I would change the transcription of the vowels themselves. The long vowels /A/ and /O/ can exist in situations where there was never a historical /r\/ (like in "spa" or "caught"), and the "rhotic diphthongs" /I@/, /E@/, and /U@/ are necessary for showing all the pre-rhotic tense-lax distinctions. I pretty much use the RP scheme to transcribe my own pre-R vowels, and I'm rhotic:
there - /DE@`/
merry - /mEr\i/
Mary - /mE@`i/
car - /kAr/
chore - /tSOr\/
here - /hI@`/
serious - /sI@`i@s/
Sirius - /sIr\I@s/
Lazar, I was referring purely to long vowels and diphthongs resulting from non-rhoticness, and not ones resulting from the "breaking" of tense vowels or vowels that were long to start with, non-rhotic-ness aside. That is, I was opposing treating, say, "there" as /DE@/, "car" as /kA:/, "chore" as /tSO@/, "here" as /hI@/, and so on, which are like how I have seen some people phonemically transcribe words like such.
So would you transcribe those EngEng words phonemically as /DE@r\/, /kA:r\/, /tSOr\/, and /hI@r\/?
Sorry, I left out a length mark. It should be:
So would you transcribe those EngEng words phonemically as /DE@r\/, /kA:r\/, /tSO:r\/, and /hI@r\/?
Yes, to my limited knowledge of English English at least.
Yeah, that makes sense. EngEng dictionaries sometimes use a superscript /r\/ to indicate where linking R's should occur, but I agree that it makes more sense to just transcribe all historical /r\/s and treat non-rhotacism as allophony.
For example,
beard - /bI@r\d/ ---> [bI@d]
It's being totally impossible. Maybe the whole net's having a crappy day?
Happens Uriel. Sometimes we being so far away and having a crappy telecom system are unable to access overseas sites sometimes.
If you have a non-rhotic dialect with linking R but no intrusive R (e.g. "spa" and "spar" are the same before a consonant or a pause, but before a vowel "spa" is [spA:] and "spar" is [spA:r/]) then I think it makes sense to include the /r/ in the phonemic transcription in words with linking R (e.g. "spar" as /spA:r/).
However, I believe most non-rhotic dialects have both linking R and intrusive R (e.g. both "spa" and "spar" are [spA:r/] before a vowel, and [spA:] before a consonant or pause). In this case it seems artificial to me to treat them differently, so I'd either put /r/ in both (e.g. even "spa" as /spA:r/) or neither (e.g. "spar" as /spA:/). The latter is undoubtably the normal approach.
If "spa" is /spA:r/, then is "calm" /kA:rm/? You might even go for /spar/ and /karm/, with [A:] just an allophone of /a/ before /r/, though you'd have to work out why "sari" and "carry" (or "curry", depending on what you think /a/ is) don't rhyme.