Re: Rotokas (was: California Cheeseburger)

Quoting Roger Mills <rfmilly@...>:

> But OT3H-- Given what I've seen so far of Rotokas, and Mark Line's
> statements, it appears most likely that Rotokas simply allows sequences of 2
> like vowels which are phonetically long. So hypothetical [ka:] contrasts
> with [ka], but these are /kaa/, /ka/.  There might even be ?*/kaaaa/, no?
How would one discriminate between the /kaa/ and /ka:/ interpretations? Well, I
can imagine ways to do it, but it does not seem to me that any of those must
needs be applicable in every language.

> Some Philippine languages also have long vowels-- some arise due to
> morphophonemics (certain prefixes/derivations lengthen vowels) but the
> majority simply correlate with penultimate stress, so there are contrasts
> like ?*['pa:nat] vs. ?*[pa'nat] (in Tagalog e.g. would be written |panat| --
> |panát|).  That leads to six-of-one arguments as to whether length is the
> relevant factor, or stress.
This reminds me of the traditional debates 'tween me and BP re: whether
consonant or vowel length is phonemic in Swedish ...

                                                          Andreas

PS It's vowel length! :)