--- In cybalist@..., Piotr Gasiorowski <piotr.gasiorowski@...>
> I know that site. The argument that the relation between sound and
meaning is _in principle_ non-arbitrary is specious and easily
falsifiable. Just look at a page of text in a language you know
nothing about (e.g. Mari, Arrernte, Burushaski or Navaho) and try to
guess the meaning, even approximately (assuming that the text does
not contain familiar loanwords).
This counter-example does not work. The thrust of the thesis appears
to be that the sound _fine tunes_ the meaning - connotation v.
denotation. The thesis has more to do with interpreting nonsense
words in rhymes like 'Jabberwocky'.
There seem to be two propositions:
A: The sounds in a word supply it with connotations. However, the
same sound has different connotations depending on the broad area of
meaning. Arguments are made for this proposition.
B: These connotations are universal. There is only limited argument
for this view. In fact, I think the site actually contains evidence
against this proposition!
> The experiment proposed by the site owner is so hopelessly
subjective that it's sure to be "successful", but what of that? To go
over the top like that means to throw the baby out with the bath
water: interesting ideas degenerate into a monomaniacal idée fixe.
That's a shame. How did you show that the experiment
would nearly always
succeed? If the experiment cannot fail, it looks like another
warning about how good people are at finding patterns in random data.
> If the author were right, historical linguistics would be
impossible, since it's precisely where the sound-meaning association
is non-arbitary that we can expect words to develop in aberrant ways
(so that the association can be maintained).
The question here is what gives when there is a conflict. Obviously,
there are at least three options:
(i) Adjust the sound - if a suitable sound still exists in the
(ii) Adjust the meaning. Meanings certainly seem to wander.
(iii) Drop the word.
There is a fourth option:
(iv) Adjust the connotations. This goes against Proposition B.
Finally, a fifth option:
(v) Live with the conflict. The site implies that this is a
feasible long-term option if connotations are not important in the
choice of word ('concrete' words).
The reality and non-triviality of the proposition depends, I suppose,
on how much options (i) to (iii) are exercised. You point out that
option (i) is rarely chosen.
Did the thesis have anything useful to say on how conflicts are
resolved? The website says that 'concrete' words like 'father' can
last even though they lack the right connotations.
> ----- Original Message ----- http://www.conknet.com/~mmagnus/
> From: richardwordingham
> To: cybalist@...
> Sent: Tuesday, August 27, 2002 1:55 PM
> Subject: [tied] Re: lat. barbatus
> For the 'ba' theme (what is the correct technical word?) the site
on phonosemantics may be useful
background. That site considers single phonemes, rather than groups,
but the principle is the same.