Re: Origins of [i\]

In Russian, it appears to have arisen as [i] after non-palatalized
consonants (especially the retroflexes). Then again, I seem to recall that
palatalization arose as a secondary feature of vowels at some point in the
history of the Slavic languages, so I probably am mistaken. At any rate,
[i\] and [i] seem to form a sort of complementary distribution.

In Mandarin, [i\] arose as a variant of [i] after the retroflexes [s\],
[ts\], [ts\_h] and [r\] and the alveolars [s], [ts] and [ts_h]. It's
actually pronounced closer to a 'zero vowel', in which the preceding
fricative/affricate is prolonged and turned into a syllabic nucleus, but
I've seen it transcribed as [i\].

So, [i] after retroflexes seems to be a common source of [i\]
cross-linguistically. Also, short [i] can plausibly de-tense to [I] and in
turn centralize to [i\]. Or, [@] can raise to [i\] under certain conditioned
environments. Or (at the risk of starting a YAEPT) it could be the 'default'
epenthetic vowel, as it seems to be in my dialect of Standard American
English: for me, /roses/ and /Rosa's/ are pronounced differently. /roses/ is
more like ["r\OU.zi\z], while /Rosa's/ is more like ["r\OU.z@z]. Same goes
with other epenthetic sequences: /betted/ is ["bE.4i\d], and so on. This
seems to be more a phenomenon of affix sequences, where the affix is
unstressed and requires some sort of epenthetic vowel to break up awkward
sequences. [@] sounds stilted to my ears in those environments.

On Sun, Feb 22, 2009 at 3:56 PM, Benct Philip Jonsson <bpj@...>wrote:

> Can anyone cite any examples of how [i\]
> has arisen in natlangs other than by
> unroounding of [u]?  In particular I
> wonder if there is any attestation of
> an [e] > [7] > [u\] path of change?
>
> The New Zealand English [I] > [I\]
> would seem to come close.
>
> /BP
>