Re: Phonological questions, bunch 2

On Tue, 2005-10-18 at 16:16 +0300, John Vertical wrote:

> Yet another side question ... where does it stem to analyze English
> diphthongs phonemically as vowel + glide? Isn't <oy> the only one which is
> historically not descended from a long vowel? And since not all vowel +
> glide combinations are even possible - /ew/ strikes me as the primary
> asymmetry - I'd think analyzing them as nuclei was more logical than as
> rimes.
> I used the transcription /aI eI oI aU oU/ until I was pointed out that plain
> /I U/ are phonemically lax and thus /ai ei oi au ou/ matches the phonemic
> structure better.
I always thought that writing (e.g.) /aj/ or by Americans /ay/ was just
an orthographical habit indicating the nature of the diphthong, and it
was still considered to be a single nucleus. I'd be interested to hear
different i.e. that there's a phonemic analysis of /aI/ being /a/+/j/,
and consisting, as you say, of a nucleus and a rime. I'd make for some
interesting clusters by our rhotic friends e.g. /jrd/ in 'hired'!

I think your original transcription (i.e. /aI/) is the more common one
(for conservative RP at least) because, as they're considered a single
unit, the phonemic analysis of the underlying vowels isn't that
important. I don't think RP has the vowel /a/ separate from /aI/
and /aU/, for instance. Certainly the system you use is used in many
transcriptions of Australian English (tho for no other reason than it's
the standard system, so people use it, so it's the standard system; when
discussing the phonology, the sounds seem usually to be divided up into
[[long monophthongs] and diphthongs] vs [short monophthongs], with
frex /E@/ being considered a monophthong [e:]).

(In Australian English, they're normally denoted /Ae, &I, oI, &O, @u\/
(that last glyph being CXS for u-bar) when desiring to be more
phonetically accurate, indicating a greater variance in the second
elements of the diphthongs.)

> Now, _Finnish_ rising diphthongs are more clearly analyzeable as
> vowel+glide... the system is completely symmetrical, and syllabe structure
> also supports this analysis (CVCC and CVVC are allowed, but CVVCC occurs
> only in loan words.)
That of course doesn't prohibit diphthongs from being classified as long
vowels (i.e. CVVC). I imagine that it most languages with a length
distinction in monophthongs as well as having diphthongs but without a
length distinction, the diphthongs and long vowels behave similarly. I
was of the impression that Finnish diphthongs derive from long vowels
too. I'd be interested to note how it managed to get such a huge
collection of vowels: long, short, and a multitude of diphthongs!

--
Tristan.

Replies