Lines of thought and sentiments..cerebrations of a different list

[Guest guest]

This posting is taken from another list.It isn't in it's given

sequence, but cut and pasted for a more concise dileneation and I

don't think it suffers to much in terms of flow. I am posting it here

because I feel it may be of interest and pertain to some of the

threads and lines of thought that have been developing lately. It

also has some wonderful referents that can be followed up at will for

more enjoyment and edification.........bn

 

.......lest we would be forever copying and not ever

> seeing)

>

> then why can't seeing the world be of this sort?

>

>

>

> Kind of an infinite regress of representations? I

> suppose that some of

> this depends upon what is meant by " sense datum " . Is

> this the initial

> external physical stimulus that excites the

> receptor, or is the " data "

> carried by subsequent neural connections up to the

> relevant

> association cortex?

 

 

This presupposes some sort of meta structure - I suppose analogous to

Chomsky's competence - we are all born with some sort of ability to

parse

languages ( a sort of meta parser ) and a corresponding capability to

organise the world into objects, qualities and relationships -

jumping along

a bit we then can fit the world into our representation of it.

Russell (he

of Principia Mathematica fame ) defined an ontology of just this sort

to

support the evolving predicate calculus (around 1919). One

development of

this was Wittgensteins Picture theory that there is some

correspondence

between the structure of what is perceived and the corresponding

structure

of logic and language in which that perception is expressed (speaking

very

loosely). Wittgenstein claimed that this correspondence could be seen

but

only shown (illustrated) - One view is that he was trying to say that

you

cannot express meta propositions in an object language (Hence Russells

theory of types).

 

 

 

If you were trying to build some (not very) intelligent device to

understand

the world (a la Winograd's Shrdlu and similar) then I guess you would

want

to represent the facts of the domain in which you were interested in

some

sort of data structure (the relational model, prolog or lisp say) -

This

represention would have to be defined for the system in some sort of

schema.

The schema limits what the system can know - the (meta) limits of our

language also limit what we can know - This, I suppose, is Kant's

point,

also Wittgenstein's.

 

 

But we don't see sense data – we see the world. That

is how the term is used. But let's examine " sense

data " as you use it – essentially the same way as

" representation. " This is the same as " processing

information. " That is, according to this view the

brain is sort of complicated telephone. Energy of

some sort impinges on it, and is transformed and this

" information " travels on. Perhaps it is even

transformed again, and again. In the brain, this is

presumably a complex process, but the upshot is the

same; the world comes in in a patterned fashion, and

this pattern is transformed, maybe in extremely

complex ways, and then, literally, transported, maybe

to be further transformed. This seems obvious to us

when we examine the nervous system – each successive

location is simply some " computational "

transformation. But the patterned world produces

actions. Are the actions really just transformations

of the world that causes them at the end of some

causal chain? Is the vocal response " water " uttered

when, say, water is observed on a roadway, really some

representation of the water? What about salivating in

response to food? Is the saliva a representation of

the food? No, you say (I hope). But then how is this

done? Didn't we just observe how the nervous system

sends " messages " downstream? Everything downstream is

just some complexly-determined transformation of the

" input. " There must be some part of the brain that is

describable in some other fashion. If this were not

so, then salivation, (or the smooth muscle action that

produces it) would be merely a transformation of the

food, and " sunset " would simply be some transformation

of sunset, like a picture on a TV screen receiving an

image from a closed-circuit TV. Does that make sense?

Is behavior actually a representation of the

environment? So, there must be something that

" utilizes " the signal – that " operates " on the

" information like a person at the other end of a

telephone line and engages in a response that is not

some transformation of the pattern. A response that

may even be, from a physical standpoint, arbitrarily

related to the " input. " But if there must be a part

of the brain that doesn't simply transform, why can

that not be the function of the whole brain? After

all, the world itself is patterned, there is no need

to make copies of it.

 

absorbing eyeses and mentations to ponder..........bob