Re: Velarization

Quoting Benct Philip Jonsson <conlang@...>:

> Andreas Johansson skrev:
>  > Quoting Benct Philip Jonsson <conlang@...>:
>  >
>  >> BTW some varieties of Swedish have what boils down to a
>  >> three-way distinction between /s`_G/~/s`/~/s\/.
>  >>
>  >
>  > This doesn't sound like anything I'm familiar with ...
>  > would you expand a bit on it?
>
> s`_G is merely an in context more revealing notation for x\.
*headdesk* I should realized it was a realization of /x\/. I was speculating if
it might be some bizarro reflex of *rD or the like.

> You got minimal pairs
>
> - /x\/ - /s\/ stjärna - kärna
> - /x\/ - /s`/ page - pars
>
> AFAIK there is no /s`/ - /s\/ minimal pair, due to the
> distributional oddities of those phonemes, but most native
> speakers would agree they're distinct.
Indeed.


To complicate matters, certain 'lects, notably my own, have what might be termed
the "split sj" - /x\/ is realized with a dark (back) allophone at the beginning
of a syllable and with a light (front) at the end of one, sometimes also at the
end of a morpheme as in _krascha_.

The catch is that the 'light sj' tends to assume the same pronunciation as the
tj-sound. Using my own (not entirely typical) pronunciation as an example, your
example words come out as:

x&:n`a ~ S&:n`a
pA:S   ~ pA:s`

In fact, I see no strong reason that the light sj should be considered an
allophone of dark sj rather than of tj in such 'lects.

                                          Andreas

Reply