[Avery Andrews 930317.1404]
Spending a bit more time on the information controversy, it seems
pretty clear to me that Rick Marken & Bill Powers have run afoul
of the bizarre fact that `information theory' has nothing to do
with content. This seems strange, but is true, and is the motivation
for a lot of what has been going on in places like Palo Alto for the
last ten years ago (so Dretske and lots of other people want a
`semantic information theory' to address issues of what content is).
The fact that information theory is not about content is probably one
of the reasons why linguists such as myself know virtually nothing
about it, but one thing I do think I know is that if the info coming
over one wire is P1, and that coming over another is P2, the info
coming over both is P1+P2. Now knowing only p(t) or o(t), we
can't recover d(t), but miraculously, knowing both, we can (if
we pretend that there is no noise), so in some sense or other there must
be info pertaining to d(t) in both p(t) and o(t), since 0 + 0 = 0. But,
equally mysteriously, in the realm of content (i.e. informative
information), 0 + 0 != 0. What we have to do is figure
out how the information can in some sense be there, and yet be
uninformative. Alternatively, we need to understand the sense in
which it is there, and the sense in which it is not.
This requires that people actively *think* and try to make sense
out of what other people are saying, rather than relentlessly defending
their positions. You can't expect the other guys to give you their
essential insights on a silver platter. To paraphrase the Godfather,
real knowledge can't be given, it must be taken.
I'd also like to point out that the situation
with d(t), o(t) and p(t) is basically the same as encryption with
a one-time pad: the pad (o(t)) certainly has no content at all,
the message (p(t)) has no accessible content for people who don't
have the pad, but if you have the message and the pad then you get
the content. Maybe the philosophy of cryptography, if there is such
a thing, has some application here.
···
----
[Bruce Nevin (Tue 93039 12:17:00, Thu 930311 12:08:14)]
>And now once more beyond the pale: in learning to be people with
>those around us we learn to discount and ignore some perceptions,
>or to cover or reinterpret them with the aid of imagined
>This is what has happened with socalled "psychic"
>perceptions, which have for many centuries been the occasion of
>painful and shameful death, and which apparently not all people
>access equally (variability analogous to color blindness).
This is also how people learn GB syntax, covering up the perceptions
to the effect that the analyses don't actually work. But I'm
not so sure about the psychic stuff. One reason I don't believe in it,
in spite of the fact that my own mother claims to have telepathic
experiences all the time, is from reading a book about a Crow Indian
warrior called Two-Leggings (can't remember the exact title,
unfortunately). He and his friends seemed to use clairvoyance in
an extremely matter-of-fact way to locate game, enemy war-parties,
etc., yet these skills were neither salable to the US Army, nor
usable against them, to prevent massacres & surprise attacks such
as Wounded Knee. Somehow, it vanishes into smoke when confronted
with our stuff. So I think that nonexistence is the best explanation.
But who knows. Maybe it's just that our songs are stronger.
(Bill Powers (930316.0900))
>Most of the effort seems to go into
>looking for regularities in language behavior. From the PCT
>viewpoint, this is something like trying to understand how people
>participate in sports by looking for regularities in the way they
>move their arms and legs.
This is because you don't have to go looking for these regularities -
there just out there, in wierd and wonderful profusion, like the fossils
that S.J. Gould likes to go on about. Real linguists are just the
people who are compelled to think about them, which is why you won't
ever find the linguist you're looking for. This is also why
linguistics is a lousy model for cognitive science in general, and
why it runs into the sand when it gets to discourse - linguistics
is about stuff that happens on a small time scale, where interaction
doesn't matter much, whereas discourse and most other cognitive things
are highly interactive (Martin Taylor has some stuff about this in his
papers). To get a good PCT linguistics, you have to start with both
more or less as they are, and resolve the contradictions, rather than
just blank out one of the subjects. You might wind up totally
redoing it of course (like when Phlogiston went into the scrap heap,
but you start with it as it is). I think that the things that
Bruce says about norms are on the right track, tho there's clearly
a long way to go. Martin's recent post also looks interesting, tho
I haven't figured out anything useful to say about it yet.
Avery.Andrews@anu.edu.au