Re: phonemes

[And Rosta <cbmvax!uunet!UCL.AC.UK!cuvmb.cc.columbia.edu!ucleaar>]

Lojbab writes:
> Since the buffer is defined by convention to be a sound that is not found in]
> LOjban's phonemic inventory, I see no reason why it cannot appear anywhere.
> I don;t know why it WOULD appear somewhere else, except perhaps as another
> kind of hesitation sound between words.
>
> Remember that we define the buffer in terms of its appearing in consonant c
> clusters, saying it is a non-Lojban sound.  We never say anywhere that non-
> Lojban sounds have any meaning in any other part of a word - it is not
 relevant
> to Lojban phonology, of course - nor is it relevant to a speaker - we teach
> people what we want them to do, not the full range of things they could
> possibly do.

It is clear that in English one can't go round inserting [zabadaz] randomly
into the phonic string and still expect to be properly understood. In fact,
as far as I am aware, any insertion of extraNeous material into the
string will hinder comprehension. It won't be unconsciously ignored.

Does, then, the assertion that "a buffer vowel may be inserted between
consonants" have the same status as the assertion "one or more segments
may be inserted anywhere in the phon. string", in each case with the
proviso that sounds corresponding to lojban phonemes will be understood
as phonemes? If this is so, then what you're saying is that non-lojban
'noise' can be arbitrarily interspersed with lojban text, and this will
be ignored, and the inserter of the noise does it at the risk of
impeding communication.

If this is the case, then the buffer vowel oughtn't to be mentioned at
all - or at least it should be stressed that it falls outside the
grammar. Furthermore, I should like to know precisely which phones
map onto which phonemes, so that it may be clarified which sounds aren't
lojbanic.

---
And.