RE: Popper, PCR, and Bayesianism (was group based judgment)

From: Colin Hales (colin@versalog.com.au)
Date: Thu Aug 08 2002 - 21:00:27 MDT


Chris Hibbert
> You keep coming back to this. It apparently seems to you as some kind of
> foundation for reasoning. I think of perception as a conduit, like
language.
> There are some things I can figure out without language, but most of my
> reasoning is done using language. I make use of perception and language,
> realizing that they are imperfect tools. There are results one can reach
with
> impeccable use of language, and experiments one can witness directly that
give
> wrong answers. You have to use the rest of your bag of tricks to evaluate
> them and decide what's correct. Perception isn't foundational, it's just
the
> main way I interact with the world.
>

I couldn't resist poking in here, briefly.

I have spent 1000s of hours writing code in real world control systems that
'sense' = "perceive". Less consider the simplest dumbest sensing device
there is: The humble limit switch. In a program it's usually registered as
ON or OFF by a '1' or a '0'. What does it mean? Holy smoke! Take your
pick....

1 = that the hardware/firmware combination has contrived to present me with
a 1?
1 = there is a voltage applied at the digital input to my computer?
...now?...how long ago?
1 = somehow a voltage connection electrical circuit has been created 'out
there'?
1 = something has caused the switch striker to fully actuate? Partly?
1 = that the device to which the switch is attached has been actuated?
1 = that the conditions permitting the device actuation have been achieved?
1 = some idiot painter is standing on the pullwire? (This actually
happened!)
1 = that the monster machine is now stopped?
1 = that George is now in the toilet?
1 = that production is now halted
1 = that it is the end of shift?
1 = that the machine is stopped for maintenance?
1 = mix and match all of the above, add more, stir................

In the life of a control system there are _always_ occasions where the
meaning assumed in the program and the meaning actually presented by the
sensing are different. Usually you end up called out at 3am to deal with a
fault of some sort!

Sensing = Foundation? Foundation for what! It's only a foundation for a
_guess_ at the meaning presented by the input and if you're lucky you're
right a lot of the time. That's all. The boundary of sensing forms a
boundary between the universe and potential meaning - a meaning which could
be anything at all, with varying degrees of certainty.

More formally: Perception is used to construe propositions (theories) that,
to a variable degree of certainty, certain conditions in the universe
exist - that a certain train of causality is in progress.

Having lassoed lots of propositions and put them in little liguistic buckets
with names on them, and then arguing endlessly over which proposition is in
which bucket how many buckets or even that the names mean what they mean -
the game is philosophy. What I said above probably fits into one or other
'ism. I don't know. What I do know is that when we finally put it into one
of the 'ism buckets it will have changed nothing and added no new insight
into what is going on in the universe or the mind. This is just my prejudice
peeking over the wall! ;-)

so...

>>> "Perception isn't foundational, it's just the main way I interact with
the world"

and

>>> Pan-Critical Rationalism (or whatever we want to call it)
>>> rests on the idea of evolutionary epistemology, that we
>>> come to know things by our brains making conjectures about
>>> the nature of the outside world. Indeed, an organism itself
>>> can be viewed as a conjecture, or a guess, made by its genes
>>> in the same way that genetic algorithms operate by making
>>> guesses.
>>
>> Ah, look at the assumptions here. First of all, how do you know of
>> an outside world? How do you know conjectures match against it?

etc...etc..all fun stuff...kind of rang my bell!

I'm done. You may continue to 'ism me.

:-)

Colin



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:15:58 MST