Re: Quantity shift (was: Re: Native grammatical terms)

Quoting Andreas Johansson <andjo@...>:

> Quoting Benct Philip Jonsson <bpj@...>:
>
> > At 03:31 22.11.2003, Andreas Johansson wrote:
> >
> > >Not that it falsifies the point that there are complications to either
> > >interpretation, but in my 'lect, those both get a long vowel; [ho:d`] and
> > >[ho:t`], respectively.
> >
> > Maybe there are really DEEP differences between our 'lects! :)
>
> Perhaps ...
>
> While in my 'lect both long and short vowels can go before the voiceless
> retroflexes ([t`] and [s`]), only long ones are found before the voiced ones
> ([n´], [d`] and [l`]*). However, it seems the short vowel plus voiceless
> retroflex combos invariably has a morpheme boundary in them,
That should, of course, read "LONG vowel plus voiceless retroflex", as my
example below shows.

                                                        Andreas

> which makes the
> phonemicity of the length distinction (whether we want to place it the vowel
> or consonant) questionable in this position, apparent minimal pairs like
> _fars_ [fas`:] "farce" and _fars_ [fA:s`] "father's" notwithstanding.
>
> * I'm quite certain my ideolect didn't have [l`] only a few years ago, but
> I've somewhy acquired it since.
>
>                                                              Andreas
>

Reply