--- In cybalist@yahoogroups.com, enlil@g... wrote
--- in http://groups.yahoo.com/group/cybalist/message/32212
> Richard:
> > Yep, three chrones allocated to two chronemes. Yucky but true.
> Not yucky and true, you mean. You're trying to prove that
> double-length isn't bizarre by using false examples. The above are
> NOT examples showing "double-long" in English. I already told you
> that. If you think _that's_ an example of double long, well then,
> my friend, it appears I have double-long vowels in mLIE! Afterall,
> I have *&, *&. and then *e/*a. You misunderstand what double-long
> means though. It means that there is a functional contrast between
> short, long and double-long. In English, there is no such contrast.
> Such a three-way contrast is rare as Piotr himself has already
> said.

It depends on whether you see a functional contrast in vowel length
between <bit> and <bid>. In Canadian English there are differences
of quality, most notoriously between <writer> and <rider>, where the
phonetic difference may entirely depend on the vowel. That
difference is definitely phonetic for some Canadians - one adduced
as further evidence the fact that his idiolect had the historically
incorrect vowel in <fire> - he doesn't rhyme it with other words in
<-ire>. In this example we see phonemicisation in progress.

If you accept the function contrast in RP, there is then
phonologically a four-way contrast - short short vowels (e.g. in
<bit>), long short vowels (e.g. in <bid>), short long vowels (e.g.
in <beat>), and long long vowels (e.g. in <bead>). Generally in RP,
the short and long vowels that are matched up with one another are
contrasted in quality. The monophthongisation of the vowel of
<bear> and <scarce> ([e&] > [E:]) yields a pure length contrast
between <Ben> and <bairn>. This contrast has a very slight
functional load in English in uninflected words - in inflected words
there is a contrast between <fed> and <fared> that depends entirely
on length. The multiway phonetic lengths fail to be phonemic
because the conditioning factor, voicing (or presence) of the final
consonant, is still realised.

For a cleaner example of superlong vowels in an English dialect, we
have the Scottish contrast of <tease> and <teas>. The inflectional
suffixes <s> and <d> are accompanied by extra length when added to
open syllables. This is perhaps more relevant to the issue of the
realisation of PIE *ph2te:(r) etc.