Andreas Johansson skrev:
> Quoting Benct Philip Jonsson <[log in to unmask]>:
> 
> 
>>R A Brown skrev:
> 
> 
>>>representation where the second element is denoted by a [i] or [u] with
>>>the small inverted breve beneath it. Those diphthongs are also often
>>>denoted as [ai] and [au] or as [aI] and [aU]. In normal speech the
>>>tongue rarely, if ever, reaches that second position. For example,
>>>English /aj/ is often realized (by those who actually use a diphthong)
>>>as [aI] or [ae], with the second element being non-syllabic.
>>
>>It is definitely [ae] to my ear, but my L1 has no [I]
> 
> 
> You have no [I]? What vowel do you have in, say, _min_?

Why so surprised, Sir?  There are plenty of Scandinavian
dialects that have no height difference between short and
long realizations of vowels.  This holds true for me with
most vowels, e.g. /e/ and /E/ both occur both long and
short, short /o/ being [o], not [O] and thus qualitatively
identical to long [o:], /2/ and /9/ are different phonemes
for me, both occurring both short and long (although /9/ is
actually closer to [3\]).  The low vowels are a bit more
complicated since when long I contrast an /a/[a:] of rather
restricted incidence from /Q/[Q:], while as short there is
variation between [a]--[A]--[Q] depending on the consonantal
context.  I still don't think length is phonemic in Swedish
-- my lect just differs from Central Swedish in having two
phonemes /Q/ and /a/, but the distinction being neutralized
in short contexts.

Things are a bit more complicated with the high vowels. When
short there are clearly three phonemes which are realized as
[i], [y] and [u], but the long realization is different in
closed syllables, where they are diphthongized [ie], [y2]
and [u8] (sic!), and in open syllables where they have a
rather fricative offglide: [iz\], [yz\], [uB] or even
[z\=:], [z\_O=:] for the front ones.  So in a way you may
say I do have a height difference between short and long
realizations here, but all the allophones are higher than in
Central Swedish.  The situation with the sounds
corresponding to orthographic _u_ is again more complicated,
and resemblant of the _a_ situation: I have two phonemes, a
more restricted /8/ and a frequent /8\/ that remain distinct
when long but merge when short.  Again there is difference between
the long realizations: [8:] and [8\B] (that is [2_wB], but often
verging on [y_wB]) in open syllables, but [83\] and [8\9] in
closed syllables.  Also the short /8/ has an [u\] allophone
in certain contexts, mainly before nasals.

I must point out that there is nothing freakish about my
pronunciation: it is a quite normal West Coast
pronunciation, i.e. I hear this kind of pronunciation around
me every day, though of course most people are not aware of
the different allophones in their own speech.  I have said
before and I say again that Standard Swedish is a phantom as
far as pronunciation is concerned.  It may be that there are
more speakers of Central Swedish than of other accents, but
that's simply because Central Sweden is more densely
populated -- theirs is a regional accent just as that of the
West Coast, Norrland, Östergötland or Skåne, and people
living in these different regions speak rather differently.
Besides how people *think* they speak and how they actually
speak may differ: there is the famous case of the Lycksele
natives who denied that they had short [e], despite that
phoneticians' ears, recordings and measurements clearly
showed they had it, because they simply assumed that their
pronunciation was the same as the one they heard in the
media, with short /e/ merged into short /E/!



>>and no true diphthongs -- e.g. |aj| being [Az\] as
>>often as not.
> 
> 
> I'm tempted to analyze Swedish Vj as diphthongs - partly because V:j is
> essentially absent - but I'm not gonna be obnoxious about it. I'm unrepentantly
> obnoxious, however, about [au] in eg. _paus_ being a true diphthong!

In a way it doesn't work for me, since my /j/ is normally [z\],
but OTOH [z\=:] *is* a perfectly possible realization of /i/
for me!  For _paus_ I have [pABs], but again I'm not sure that
I don't have [B=:] as a possible allegro realization of /8\/,
since non-instrumental analysis of one's own allegro speech
is inherently difficult!

BTW one side effect of [B] being allophonically present
in my Swedish speech is that I use [B_o] or even [B]
as realization of English /w/, and [p\_o] -- which was
m father's main /x\/ allophone! -- for English /W/!

>> I am a semi-native speaker of German,
>>but I've seen German |ei| transcribed as [ae] as well...
> 
> 
> [ae] is probably the commonest transcription in my experience.

Yes, with corresponding [ao] and [o2] for the other two diphthongs.

One thing is certain: my Swedish "diphthongs" and their closest
German counterparts don't rime phonetically!

BTW, is it only Tristan and I who are given to meticulous
sub-phonemic analysis of our own lects? :)

-- 
/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    "Maybe" is a strange word.  When mum or dad says it
    it means "yes", but when my big brothers say it it
    means "no"!

                            (Philip Jonsson jr, age 7)