Re: The Language Code (take 4)

On Tue, 2003-06-17 at 05:19, Mark J. Reed wrote:
> 1. bat                  /&/
> 2. bought, father       /A/
> 3. bet                  /E/
> 4. bit                  /I/
> 5. beet                 /i/
> 6. boat                 /o/
> 7. put                  /U/
> 8. boot                 /u/
> 9. but, begin           /@/
> Thoughts on the list?
I'm not suggesting this seriously, just to reduce it further, but you
could associate [i] and [u] with /II/ and /UU/ and if it's tenser, it's
just alophony, and the same with [ou] -> /VU/ (the backness and heightof
the [U] pulls the /V/ up and rounds it). (I'm using /V/ rather than
/@/.)

Thus:
I       U
E       V
&       A
And you get a nice rectangular system in six vowels :)

And now lets put dialects' (sorry) vowels in terms of it:
[I]  = /I/
[I:] = /IIr/
[Ii] = /II/
[e:] = /EIr/
[e]  = /E/
[&]  = /&/ (dad)
[&:] = /&&/ (bad)
[&i] = /EI/
[&u] = /Au/
[6]  = /V/
[6:] = /&r/
[6i] = /Ai/
[O]  = /A/
[O:] = /AA/ (gone)
[o:] = /Ar/, /AV/ (sore, saw)
[oi] = /Ui/
[u]  = /U/
[u\u] = /UU/
[8:] = /Vr/
[@]  = /V/

Well, that mostly works once you allow sequences of the vowels to be
realised differently. It also means it's necessary to put secondary
stress on /"hI%kVp/ 'hiccup', though this wouldn't be necessary if
created another vowel

While we're at it, let's merge [I] and [j] and [U] and [w], so that
'water' is /UAVtVr/. I could also argue for allophony between [@] and
[r] if we don't want to keep the same underlying form in AmE and AusE
(which means we can get rid of the combination /AV/ and just say that
'saw' is /sAr/). Of course, it only offers the advantage of merging [@]
and [8:] and [r], so we don't reduce the phoneme count, and you can
generally tell whether a [@] has an [r\] in rhotic englishes anyway, so
it doesn't offer much advantage.

--
Tristan.