Re: Is consciousness a process?


"Stephen Harris" <cyberguard-1048@xxxxxxxxx> wrote in message
news:RoJYh.1659$tp5.1406@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Glen M. Sizemore wrote:
"Stephen Harris" <cyberguard-1048@xxxxxxxxx> wrote in message
news:a0GYh.3725$H_.2347@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Don Geddis wrote:
Stephen Harris <cyberguard-1048@xxxxxxxxx> wrote on Fri, 27 Apr 2007:


So in this sense I don't disagree. It is clearly not the case that
intelligent behavior _causes_ consciousness.

The real implication is the other way, as you suggest. _If_ an entity
demonstrates ("sufficiently") intelligent behavior, _then_ it must
actually
have consciousness inside.

But if it _doesn't_ happen to demonstrate intelligent behavior while
you're
watching, then you're on much weaker ground to conclude that it does
_not_
have consciousness.

(I made many of the same comments when we discussed the Turing Test.
"Failing" the Turing Test provides very little information about the
presence or absense of consciousness in a candidate entity.)

"I Am A Strange Loop" by Hofstadter page 257

"The cells inside a brain are not the bearers of its consciousness;
the bearers of its consciousness are _patterns_. The pattern of
organization is what matters, not the substance. ... And patterns
can be transferred from one medium to another, even between radically
different media. Such an act is called "transplantation" or, for
short, "translation"."

SH: I like this description. Earlier in the book he mentions that
infants have less consciousness than adults, so its something that
develops. The likely candidate for how it develops is experience,
which I guess can be broken down into behavior interacting with the
environment.

Would it not, then, behoove one to state, rather explicitly, how behavior
interacting with the environment produces the sorts of behavior that are
said to "show consciousness" etc., then to jump into the alleged
"patterns" that such an interaction leaves?


I snipped part of Don's post, IMO relevant to answering your question.

... DG: "You can take a human, and put them in a sensory deprivation
tank, and have them meditate for an hour, and you'll get precious
little behavior out of them to analyze for anything.

Yet we can all tell (by personal experience, if necessary), that the
human subject is just as conscious during that hour as at any other
time. You can work on math problems, compose music, construct a
grocery list, etc. Lots of internal mental behavior."

SH: So I accepted that thinking was a type of behavior because he
asserted it, although I was unsure about the status of thinking
as a type of behavior before then. Then my use of "experience"
included both exterior, observable to the outside, and interior,
not observable to the outside but to the self, types of behavior
according to Don's usage which I was accepting. Although a sensory
deprivation tank is not a very dynamic environment, it still seems
like an environment to me. I don't think this fits your assumptions.

You're wrong, of course. Talking to ourselves is behavior that generates the
same sort of stimuli as someone talking to us. In a SD tank, with little to
control perceptual responses, we begin to halucinate and what we see changes
our behavior. But why start with this? Why seize on the very hardest
problems when the simpler ones are nowhere near being solved? The
parsimonious view is simply that behavior that only the behaving person can
see is still behavior, and it can exert control over other responses, just
as someone elses behavior may exert control over our behavior.

I am pretty much a Physicalist. Hofstadter's patterns reminds me of

"A synaptic pattern is the unique configuration of connections
between neurons (synapses) and the strengths of these connections
in the brain. The term is usually used in the context of measuring
or recording the pattern.

None of this answers my question. Why do we say the things we do to
ourselves and why does it have the particular effect that it does? It is not
enough to say that you said something to yourself and it made you think of
something else etc. And here you are talking about these stupid "patterns"
again. You don't know how the animal/environment interaction results in the
behavior that we see, but its time to talk about the patterns that are the
"real" cause of the behavior. As I have said, ALL you are saying is that
physiology mediates behavioral function. That is ALL, I repeat, ALL you are
saying, and we have known that for three fucking hundred years.


Since memories, thought patterns, and aspects of personality are
encoded in the synaptic pattern, copying a synaptic pattern can be
used to duplicate aspects of a person."

Physiology mediates behavioral function.

SH: "copying a synaptic pattern" seems consistent with Hofstadter's
use of "transplantation" or "translation" quote above.

Getting around to your question ... I am engaged in another thread
which is similar to this one with Daryl. Daryl thinks that behavior
is by itself _sufficient_ to impute consciousness. I disagree with
him and contend that behavior is necessary but not sufficient, I
think the underlying structures are also necessary (the black box)
and by structures I don't mean just physical brain arrangement,
but conceptual structures, maybe synaptic patterns.

You have no idea what units of behavior are, and you don't understand that
identifying them is a matter of finding the independent variables that
produce behavior. That is why you, like Cybulski, can bring up something as
stupid as the "actor in 'pain'" argument, or your even stupider "my friend
mistook my answering maching for me." Your answering machine is not
behaving, at least not in the sense that an animal "behaves" and the "actor
in pain" emits behavior that is under control of different variables than a
person "in pain." Maybe someday paying attention to some sort of brain
measurement might allow you to decide what sort of behavior one is looking
at but, right now, it is behavior that gives meaning to physiological
observations. If someone published fMRI stuff and didn't tell you what the
person was at least doing in colloquial terms, what would you have? Nothing.
The only way to figure out "what the brain is doing" is to first have laws
that cover how an animal's interaction with its environment shapes and
maintains behavior, because that is what the brain mediates.


I tend to think all experiences which are generated by behavior
interacting with environment contribute to consciousness, but
maybe there are intensity thresholds or frequency considerations.

You can't even formulate a reasonable question. An animal's interaction with
its environment alters behavior. What are the laws that describe this?
Intensity thresholds" Frequency considerations? What are the fucking laws
that describe how an animal's inter-fucking-action with the environment
alters its behavior?


Like in the Turing Test, judges offer their expert, but nonetheless
subjective evaluations of program behavior reponses to determine
if the program "shows consciousness" which you emphasized above.
I don't think humans can scientifically establish some sort of
ideal, universally objective test whose questions provide great
certitude in serving as criteria for human-level intelligence
with a Computationalist assumption that a mind is instantiated
which usually also presumes consciousness, which most of us admit
we experience and is as real as any other indirect perception.

Then we are in a real pickle because you can't establish what "brain stuff"
causes which "behavior stuff" without having the behavior stuff in hand. You
can't very well look for the "neural correlates of consciousness" unless you
can identify that with which it supposed to be correlated. What is wrong
with you and most others on this NG? I should know better. It's just
cognitive "science" as usual. Its ackowledged that exposure to/interaction
with particular environments alters behavior, but the exact laws of these
phenomena are not sought.


Just as there is no sound word for word translation of say a
poem from French to English, the transplantation of patterns
from the wet brain substrate to a dry silicon substrate, though
possible, may not be a literal operation. So in this case I
think the interpretation would likely require a more explicit
understanding of how behaviors interacting with the environment
emerge as patterned contributions to consciousness.

What sort of gibberish is this? What in the *** are "patterned
contributions to consciousness"?


In another of your posts you said:

GA: "Cognitive psychology implies that it is because they possess
"a self-concept" but the question then needs to be asked "Where
did that come from?" This is always the problem with mentalistic
explanatory fictions - even if they are "real" (which they aren't)
we have to ask where they come from, and that leads back to the
selective environment either ontogenic or phylogenic."

SH: We will probably never agree on this, but "a self-concept"
which Hofstadter call I-ness, seems as real to me as my computer,
even though I cannot touch it, I experience it.

This doesn't answer the question.

There are people
who claim that consciousness is not amenable to the scientific
method... Computationalism has a goal which Behaviorism doesn't
consider a useful concept(instantiated mind/consciousness).

You are right that behaviorism says that the notion that we behave in a
conscious fashion because we possess consciousness is
drool-on-yourself-stupid. And I have told you many times what I think about
self-awareness. Someone saying "I did this" or "I did that" is no more
fucking mysterious than saying "Steven did this" or "Steven did that." I
have also told you a thousand times that behaviorism has no trouble with
"subjective phenomena" if that is part of what you mean above.

don't know how important such a controversy is to a result
since Comp. also uses behavior as a tool. Perhaps you can make
an argument for the superiority of your approach, but you can't
provide the convincing evidence of a human-level AI, thus the
intuitive perception of me having a mind, which Comp. also gives
some credibility to, doesn't face a lever with a large enough
fulcrum to dislodge it. Issues other than that remain unresolved
from my pov, since neither Behaviorists nor Computationalists
have used their relative approaches to produce a human-level
behavioral response program. That would be evidence, other
replies like, that isn't our focus may serve as valid excuses
but not amount to positive evidence for adopting that pov.

Why would there not be some other way to decide than a full-blown artificial
person? What about picking the view that isn't animistic fucking retarded
horse***? What are the laws that govern how the animal/environment
interaction shapes and maintains behavior? What sorts of history produces a
man that truthfully says "I'm in pain?" What sort of history produces a
person that fakes it? Only with these sorts of facts in hand can we begin to
talk about how physiology mediates such behavioral laws. AI might be
possible without this understanding (I sincerely doubt this, though - they
can save their "planes don't fly like birds" *** until after such a time
when what needs to occur can be given a description as clear as "flight")
but a physiological explanation of animal behavior cannot.


Regards,
Stephen


I'm not sure how some of the other terms fit in with
this base description of like mind, self-awareness, and I-ness, I
think Hofstadter uses them interchangeably, maybe because he thinks
they are all patterns (maybe with slightly different functions?).

It doesn't seem like his insight is sufficient to recognize an alien
consciousness, and also its behavior might not be evident. There are
some patterns which appear to have the same trajectories for quite
some time, but then radically diverge. I've seen this playing around
with fractals when they were popular. I'm thinking the trajectory
could be mapped or associated with behavior(s). That is my visual
notion of how a TT program stays within the boundaries of the space
of acceptable human-like conversational responses to questions.

Best regards,
Stephen


.

Quantcast