Re: Rotokas (was: California Cheeseburger)

On 16 Jun 2004 "Mark P. Line" <mark@PO...> wrote:

> > I'd guess that [s] is an allophone of /t/ before /i/.
>
> It could also be an allophone of any other phoneme, if it's an allophone.
> The question would still remain as to why it's written as 's' and not as
> 't'. The orthography might be multilectal, for example: there might be an
> /s/ in one of the dialects that merges with some other phoneme in other
> dialects. That's why I said I'd ask.
  I do not know the reliability of the author(s) of the article
"Rotokas alphabet" in Wikipedia
<http://en.wikipedia.org/wiki/Rotokas_alphabet>, but it says that
"[it consists] of only the eleven letters AEIKOPRSTUV [...] The S
is sometimes written G, and the V is sometimes written B."

  This raises the possibility that [s] can be an allophone for /g/
too. Not considering the tap /4/, the majority of the sources tells
that there are just two voiced phonemes(?) in Rotokas: voiced
bilabial fricative /B/ (which is spellt as <v> or <b>) and voiced
velar stop /g/. This is not symmetrical: it is very likely that /g/
is realized (partly) also as a fricative [G] and after a (supposed)
devoicing we may consider the variants [x] > [s].

  As for the phonotatics and own the etnonym of Rotokas: Ethnologue
mention three dialects: Pipipaia, Aita, Atsilima. I have found that
the etnonyms Rotokas, Pipipaia and Aita are geographical names in
some edocuments, e.g.
<http://sshl.ucsd.edu/melanesia/nsolomons.htm>. Probably these are
much rather external names from the denomination of the inhabited
area than own etnonyms. Maybe they call themselves just "man" in
their dialect...

  If the above dialect name "Atsilima" is not a typo than (a) there
are some other allophones (e.g. /ki/ = [tsi]), or (b) consonant
clusters are allowed after all. Probably the above three dialects
slightly differ from each other and the merging of the dialects
cause the contradictory opinions.

  An example for the contradictory data: Herman Miller cited a
Rotokas sentence on 15th June: "vi wiliwili vorepieriva". It is
suprising because there are both word-initial "vi" and "wi".
According to this "v" and "w" (the latter is possibly [B]) should
be separate phonemes not just the allophones of the single phoneme
/v/...

Reply