Quoting Henrik Theiling <[log in to unmask]>:

> Hi!
> Mark J. Reed writes:
> > And then there's "chassis", whose pronunciation changes in the plural
> > even though the spelling doesn't...  gaaah.
> I think it's the same in German.

Hm. I looked it up to see how its spelt in Swedish (turns out to be _chassi_)
and found something surprising - it gives the plural as _chassier_, which sounds
just plain wrong to my ears; I say /'x\asin/. It would probably be excessively
sophistic to conclude _er_ spells /n/ here ...

> And the adjective 'orange' is [?o:'RaNS] while the noun 'Orange' is
> [?o:'RaNZ@].  The adjective has normal inflection, of course:
>       die Orange ist orange   [di: ?o:'RaNZ@ ?Is(t) ?o:RaNS]
>       the orange is orange
> but   die orange Orange       [di: ?o:'RaNZ@ ?o:'RaNZ@]
>       the orange orange

In Swedish, I guess you have to spell the weak masc. sg /u'ranSE/ the same as
the base form _orange_ /u'ranS/. The neuter _oranget_ /u'ranSt/ and
fem./common/pl _orangea_ /u'ranSa/ are also funky.

The fruit is _apelsin_.

(I'm, incidentally, violently agnostic about the wider affiliations of the sound
I here denoted /S/. It may or may not be to be identified with /x\/ or /s\/. The
symbol chosen merely reflects my own phonetic realization in this lexeme.)

Andreas Johansson