Re: SUO: Re: Logic & Programming Languages
Leo,
I agree. There is a major difference between being able to do
something and being able to explain verbally how one does it.
A typical example is tying a shoestring.
Some comments:
Leo Obrst wrote:
>
> I would suggest that we make a distinction between consciousness (of
> inference) and inference. I am not convinced that what we
> introspect/reflect on as an inference process is the only kind. In fact,
> I doubt it; I think that most inference is in fact unconscious (but
> because "unconscious" is a loaded term, perhaps it's better to say
> "unaware of").
That is also true of computer systems. A theorem proving program,
for example, could certainly be called "logic based", but unless it
has been explicitly developed with an explanation facility, it cannot
comment on how it uses logic.
> I think in general that we (humans) are notoriously
> deficient at describing introspection our own mental processes except
> in very general terms. Many of us will say "That's not how I do it" when
> via reflection it doesn't seem that an inference process has taken
> place. We remember the original "pattern" and our
> action/response/conclusion. There is no formal sequence of modus ponens,
> we think.
Yes. Typical examples are the comments by fluent speakers of
highly inflected languages about the case systems. They certainly
don't think "now I am going to use the dative case" when they
use the dative case.
> In another realm (but related), witness language. As fluent speakers of
> a language, we think we understand language, but that is not generally
> true -- at least if you believe linguists. Our naive understanding of
> how we create and understand language is not necessarily very
> insightful. It's not usually the best theory.
Another typical example is to ask any native speaker of Chinese
about Chinese grammar. Unless they studied linguistics, the almost
immediate response is "Chinese doesn't have any grammar." What they
mean is that the words are not inflected. But since Chinese happens
to be an SVO (subject-verb-object) language, they can create sort of
intelligible English (i.e., Chinese-waiter English) just by doing
a word-for-word translation from Chinese.
> So then if logic and mathematics (and so, science) is always a
> "description" after the fact of real-time processes or reality, and are
> not "explanations", what constitutes an explanation or equivalently (I
> would say), a theory? [Now, I don't want to discuss here Goedel/Chaitlin
> foundations of math issues; they are important but in a different
> context.]
>
> What is the process the animal goes through? Is it hard-wired
> "instinct", whatever that is? And even if it is, what does that mean? If
> we say: oh, it's the firing of neurons in a certain way or flow across a
> finite state network in a certain way, and we leave it at that, we are
> not really interested in understanding what occurs. I would say, it *is*
> the firing of neurons in a certain way, that that certain way is
> responsive to the "truth" of real world situations, and that certain way
> can be described, one can propose a theory of that certain way which is
> better than other theories.
Yes. And that is why it is much more useful to use tools such as
formal grammar for describing syntax and logic for describing semantics
than to try to cook up some sort of "neural theory of language" (NTL).
I don't claim that humans and other animals use some version of those
tools in their heads -- it just means that the result of what they
do has the same patterning as that produced by the tools.
And speaking of NTL, I would like to mention my review of Lakoff and
Johnson's book _Philosophy in the Flesh_:
http://www.jfsowa.com/pubs/lakoff.htm
They make some useful points, but unfortunately, the book is more
dominated by Lakoff's sweeping-generalization style than by Johnson's
cautiously worded style. Following is the last paragraph of my review.
John Sowa
_________________________________________________________________________
In summary, this book makes an important contribution to the ongoing
debates about the roles of syntax, semantics, and world knowledge in
language understanding and their dependency on the physical world and
the human mechanisms for perceiving, interpreting, and interacting
with the world. Its major weakness is its tendency to exclude other
perspectives, such as Aristotle's, which can accommodate both formal
logic and a theory of embodied mind. Although the authors frequently
use the word neural, none of their discussion depends on the actual
structure or method of operation of a neuron. NTL could with equal
justification be considered an acronym for a Neoaristotelian Theory
of Language.