Cybernetics of Epistemology - Understanding Understanding: Essays on Cybernetics and Cognition - page 232

say, “There stands a lectern.” The correlation between these three inde-
pendent observations permits me to say: “There stands a lectern.”
In the first case, we assume that with each new independent observation
we confirm the correct perception of the previous observation of the reality.
In the second case we use Ockham's Razor to shave off unnecessary
assumptions and allow a reality to emerge through the correlation of our
sensations.
In any case, I intend to use the indefinite article, as it represents the
general case: clearly the reality is a special case of a reality.
We have barely laid the indefinite article to rest when we are already con-
fronted with a new issue. What is “computing” supposed to mean? I could
by no means claim in all seriousness that the lectern, my wrist watch, or the
Andromeda Nebula is being computed by me. At the most, one could say
that a “description of reality” is computed, because with my verbal refer-
ences (“lectern”, “wristwatch”, “Andromeda”), I have just demonstrated
that certain sequences of motion of my body combined with certain hissing
and grunting sounds, permitted listeners to interpret these as a description.
Neurophysiologists may in this context object that incoming signals have
to undergo many steps of modification before we experience a verbal
message. First the retina provides a two-dimensional projection of the exte-
rior world which one may call a “description of the first order.” The next
post-retinal networks then offer to the ganglia cells a modified description
of this description; thus a “description of the second order.” And so it goes
on via the various stations of computing all the way to descriptions of higher
and highest orders. We can thus modify the second version of my proposi-
tion in the following way:
Cognition Æ computing of a description
This paraphrasing has two advantages: the contention “the reality” or “a
reality” has disappeared, as now there is no more talk about reality.
Secondly, we can use the insight that computing a description is nothing but
computing. This way we reach a final paraphrasing of the forever renewed
process of knowledge acquisition, i.e.:
Cognition Æ computations of
I, therefore, interpret knowledge, or the process of knowledge acquisition,
as recursive computations.
At this point I could follow two different paths: I could now talk about the
characteristics of recursive functions and of the characteristics of machinery
to compute these functions. In machine language, I would then talk about
cascades of compiler languages and about the theory of metaprograms. In
this case, naturally, the concept of the Touring Machine would be the ideal
conceptual tool. It becomes, for instance, very clear that the structure of the