Re: long consonants

#1/Max wrote:
> Henrik Theiling wrote:
>
>  To travel
> >around the world,
> >naming a few not-so-unknown langs: it exists in
> >Finnish, Estonian,
> >Japanese, Arabic, Greenlandic, Inuktitut, Swiss
> >German, and many
> >others.  (And I'm sure I forgot a few other famous
> >ones.)
> >
> >I'd say it's quite common.
> >
>
> ok thanks for these explainations I now feel more rassured about using
> them
> in conlangs and feeling them natural
>
> but still a question: are long consonants... geminated consonants phonemic
> in these languages?
They can be considered phonemic, in that they distinguish meaning in minimal
pairs:  Ital. anno [an:o] 'year' :: ano [ano] 'anus'. But it may depend on
how one defines "phonemic" and what theoretical model one uses...And of
course they can originate historically in a variety of ways-- assimilation
of cons. clusters, for ex.

IIRC in Arabic, gemination of C is a feature of some derivational processes.
>
> Are geminated consonants opposed to the short ones enough to make that
> pronounciating a germinated consonant as short would make the sense
> different or would it remains the same?
Yes, definitely. I don't have my lists handy, but there are lots of
min.pairs in the South Sulawesi languages-- some I remember, Bugis  mita
['mita] 'to see' :: mitta ['mit:a] 'long time'; Makassarese, near-minimal:
tallu '3', allo 'sun' :: alu 'pestle', talo 'defeated'
>
> And are they really phonemes in the sense it may be used in a root or do
> they simply *occur* when you paste a word that ends with the same
> consonant
> it is pasted to, like it is in the "penknife" example Sanghyeon Seo gave?
That can happen too, but it's purely phonetic; in such cases there's always
a morpheme boundary. Again Bugis: aseng 'name' + -ku 'my' > asekku 'my
name'; + -mu 'your' asemmu.  Similary in Arab. where the -l of the article
_al_ assimilates to certain initial consonants--
as/salaam '(the) peace', ar/rahman 'the omnipotent'.

Geminates are fascinating!!