Mat's comment on my reply to the other MATT (was Laymans...).

Glen M. Sizemore gmsizemore2 at yahoo.com
Tue Apr 30 18:36:07 EST 2002


Because of some strange server problems, I occasionally have to reply via a
new thread.

Mat: I realise I'm not the 'Matt/Mat' to which this was addressed but
anyway... Computation in no way implies a homunculus, at the network or
behavioural level.

GS: I'm afraid it does. As I said, and you ignored, "computation" (like
"representation," incidently) is not a physical property of anything. We
identify something as a computation when a human can use it in a particular
way. You have not acknowledged that you understand this, or dealt with it in
any way. For example, computation is what my computer allegedly does when it
returns to me the square root of 63, it is allegedly what my kid does when
he follows the rules I taught him to ad 54 an16, and it is what the analog
device (built in the early 20th century) composed of a map with holes
drilled in it and weights on strings did when it located a local minimum for
the distance from all of the cities. What do all of these things have in
common? Think about it for a change. I say again, "computation" is
"computation" when a human can use the result in a certain way.

Mat: You admit it doesn't have the conotation at the network level, so form
where does it suddenly arise at the behavioural level.

GS: No. I did not say that it doesn't carry that connotation, I said the
connotation was relatively harmless. And I did add a caveat.

Mat: A computer (as in PC) could be described as having behaviours as
consequences of environmental stimuli but no one is saying that it has a
homunculus (and don't say the user watching the screen because that is not
the same thing). In fact, such a definition of behaviour is exactly what you
are trying for. Just becuase we describe the activity 'computation' from the
outside looking down on the cellular networks does not mean there has to be
a homunculus who normally watches the computation!

GS: I'm afraid it does. As I said, and you ignored, "computation" (like
"representation," incidently) is not a physical property of anything. We
identify something as a computation when a human can use it in a particular
way. You have not acknowledged that you understand this, or dealt with it in
any way. For example, computation is what my computer allegedly does when it
returns to me the square root of 63, it is allegedly what my kid does when
he follows the rules I taught him to ad 54 an16, and it is what the analog
device (built in the early 20th century) composed of a map with holes
drilled in it and weights on strings did when it located a local minimum for
the distance from all of the cities. What do all of these things have in
common? Think about it for a change. I say again, "computation" is
"computation" when a human can use the result in a certain way.

Mat: Computation can be implicit.

GS: What is computation?

Mat: Further to your replies to my posts: Why is delineating and
categorizing behaviour any more or any less valid than categorising
psychological process, which you seem to totally deride?

GS: Because the categorization of behavior that I support has rigorous
empirical support, AND THE CONCEPTS UTILIZED ARE NOT LAME, LIKE THOSE THAT
CHARACTERIZE COGNITIVE PSYCHOLOGY, BECAUSE THEY ARE BROUGHT INTO EXISTENCE
WITH THE GENERATION OF THE FUNCTIONAL RELATIONS. The issue is not whether or
not to use "behavioral" or "psychological" - elsewhere I would use them
somewhat interchangeably - the issue is what concepts are being used, and
how are they interrelated with theory and experiment.

Mat: Such distinctions will only ever be arbitrary[...]

GS: No. Such distinctions are not arbitrary. It is not arbitrary that the
consequences of behavior alter behavior. This is as profound as the early
recognition that it was the weight of a substance that was important and not
its color or taste, for example. Indeed, the facts delineated by the
experimental analysis of behavior are akin to the facts constituting the
periodic table.

Mat: [...] and based on preconceived ideas which do not take account of
neuroscience.

GS: They do not take account of neuroscience because they are behavioral
facts. And they are behavioral facts with which neuroscience must be
consistent.

Mat: Investigating how known network properties of neurons are involved in
generating actions (and thoughts) would be a much more profitable venture in
my belief.

GS: ?????? And how is this antithetical to what I've been saying? Indeed,
what I have been talking about IS how one identifies actions. What is an
"action?" I have an answer. Do you? And if you do not have a rigorous
definition, how in the hell do you have the potential to investigate "...
how known network properties of neurons are involved in generating
actions..."? Let alone "thoughts." I'm sure your analysis of "thoughts" will
be just as insightful as the rest of your analysis.

Cordially,

Glen






More information about the Neur-sci mailing list