[Wittrs] Re: The CRA in Symbolic Form (According to Joe)
- From: "SWM" <SWMirsky@xxxxxxx>
- To: wittrsamr@xxxxxxxxxxxxx
- Date: Wed, 28 Apr 2010 17:29:27 -0000
--- In Wittrs@xxxxxxxxxxxxxxx, Justintruth <wittrsamr@...> wrote:
> > a running program is only using syntactic operations
>
> Can someone help me understand under what conditions a machine is
> "only using syntactic operations". Is Von Neumann architecture
> required? Parallel processing ok? A Turing machine required? Is
> storage of states and a clock required or is an analog device ok?
>
It's partly just a stipulation, i.e., that we agree to equate computational
processes running on computers with whatever we call "syntax" because, like the
rote (mechanical) behavior we usually have in mind when we speak of "syntax",
computer processes involve implementing various rules in a set way (according
to a prescribed format or set of steps, i.e., an algorithm) but without relying
on thinking about what one is doing in any explicit way and without penetrating
into the meanings of the symbols being manipulated.
Of course there are real differences between the syntax of a language as users
speak it, the syntax of the mechanisms of the users' capacity to actually
follow speak a language, the syntax of a machine operating according to pre-set
instructions by just implementing one event after another, etc. But for the
purposes of moving the discussion forward, it's probably not untoward to accept
the usage that holds that computational processes running on computers
(computer programs) are syntactical in nature.
For Searle, and his adherents, such computer behavior is called "syntax" and
insofar as all that's meant is that the computer's "behavior" is seen to be
totally mindless, thoughtless, intentionless, etc., etc., (even if a human
implementing syntactical operations may have accompanying thoughts, awareness,
etc., for his or her part), relatively little harm is done I think.
The problem arises, however, when, as seems to me unavoidable, different
concepts of "syntax" come into play in the argument and start to impinge on one
another.
> Surely "using syntactic operations" does not mean that "it is running
> a program" as the syntax of the program is just a function of the user
> interface for the compiler not a function of the target systems
> execution of the complied program.
Right, but that does become part of the confusion that you have rightly picked
up on here. Thus we humans may follow pre-set rules a la algorithms and, in
doing so, will have various levels of awareness of doing this, of the rules
themselves (their semantics or meaning), etc., while a machine running a
comparable set of algorithms is just a tool, a device with no intentional
awareness at all. Searle trades on this notion in making his case that the
"syntax" he says computer programs are can't do whateve it is that brains do.
Searle sometimes says this is because there is no understanding there to begin
with and, because there isn't, there is no possibility of that (his CRA) and
sometimes, because mere syntax is abstract, not, as Budd will tell us, a
"natural kind", case it cannot have any causal efficacy in the world at all
because what is purely abstract cannot cause events.
Searle's mistake here, in either case I think, lies in thinking that the
"syntax" he calls computer programs has no physical dimension. Whether he
denies the possibility of bringing consciousness about in the world because
syntax is utterly unlike anything we mean by "semantics" or because it is
thought to be so abstract as to be excluded from whatever physical brains can
do, Searle misstates what kind of "syntax" computational processes are. That
is, they are physical processes consisting of changing physical states and, in
this, they are no less physical or more abstract than what any normal and fully
operating human brain is.
> In fact different syntax can be
> used to achieve the same result at run time and the same syntax could
> be used to cause a completely different execution just by changing the
> compiler.
>
> If I have a simple button labeled "Start" then there is a syntax to
> the labeling of the button. That does not mean depressing the button
> is a "syntactical operation". Does it?
>
In a way the programmer or computer engineer is operating as syntactically as
the machine if he or she is not thinking about the meanings of what he or she
is doing which is perfectly conceivable. But just as often or even more often,
some degree of thought is accompanying the behaviors. And then we have a
difference in the application of the term "syntax". Searle will say, well, our
human subject "has semantics", the (or a) characteristic of having a mind. But
if the human mind is just the result of certain physical operations happening
in the brain, none of which are themselves self-aware, intentional, etc., then
the fact that the already conscious human has these is not an indication that
he or she doesn't have them because of lower level processes in his or her
brain which doesn't have them.
> So called "machine language" are typically just physical (usually
> electrical) states of a device which when "execution" is initiated, by
> pressing the run key for example, cause a cascade of physical
> processes. Sometimes those processes result in states that are
> interpretable according to the rules of some language (as in the CRA
> taken literally does) but not always. Is it required that the output
> be a language?
>
No, but it is supposed that the process is akin to a mindless manipulation of
symbols which the language speaker engages in more mindfully.
> Is a chimp operating syntactically? Only when he communicates with
> other chimps or when he reaches for a banana peels it and eats it
> while sitting alone?
>
I would say not but, certainly, on some interpretations, one might want to say
yes. This further points up the lack of an adequate explication of this term,
beyond the initial stipulations with which we started.
> I suspect that unraveling the CRA comes down to understanding what
> "using syntactic operations" means precisely. Or has the term been
> expanded so much that any device that is assumed to operate according
> to some formal rule or "physics" is syntactic?
>
I would say that Searle wants to dissociate formal rules from actual physical
operations in the world. In this I think he's dead wrong.
> At least what does it mean in this proof and in this conversation?
>
> ==========================================
>
That's the $64,000 question! But I've already given my answer. I think our
resident Searleans ought to have a go at it by giving theirs now.
SWM
=========================================
Need Something? Check here: http://ludwig.squarespace.com/wittrslinks/
Other related posts:
