Re: Are We Machines?

From: HARNAD Stevan (
Date: Mon Jun 03 1996 - 20:22:17 BST

> Date: Thu, 23 May 1996 12:30:36 GMT
> From: "Whitehead, Sonia" <>
> How can we define what a machine is? It is difficult to determine what
> is a machine and what isn't a machine. What categorisation mechanism
> can we use to split items such as programmed comoputers, typewriters,
> bows and arrows, plants, humans and robots into the two categories? Are
> people with artificial vital organs machines because they are owing
> their life to man-made mechanisms?
> Machines are programmed using a set of mechanical rules.

An electric tea-kettle is "programmed"? A locomotive?

> Humans are
> programmed in a similar way with their traits depending on the
> structure of our DNA.

So far this just suggests that some machines are programmed; but since
we don't yet know what is and isn't a machine, that hardly helps.

> Through reverse engineering Dennett has found
> that computers share the same experiences with humans.

Dennett has not found anything; he writes about reverse engineering; he
does not himself do it. But even those who do it have not and could not
have found that computers have experiences, much less the same ones
people do. The reason they could not find that out is called the "Other
Minds Problem."

> He looked into
> the mind of computers by looking at a computer chess game.

He didn't look into the computer's mind; even if the computer had a
mind, as we do, no one could look into it, because of the other-minds
problem. There is no "cerebroscope," and even if there were, you
couldn't know that it was reading someone else's experience rather than
your own.

All Dennett did was INTERPRET the computer chess programme as if it had
a mind: as if "it thought it should get its queen out early." That
helped him understand what the computer was doing, just as it helps us
understand what other people are doing. But having a mind is not the
same thing as being interpretable as if having a mind...

> He
> confronted the mind/body problem as how do we know what a computer is
> thinking?

That isn't the mind/body problem; it is a variant of the other-minds

> However the computer was originally made by mankind. Could it
> therefore be said that a machine is something which carries out
> functions and has been made by man?

What is "carrying out functions"? Isn't everything carrying out
functions (if only the function of sitting there)? And are you saying
that if typewriters grew on trees, they would not be machines? (Why not?
if they were otherwise identical to man-made ones?) And if someone
managed to build a person, then it would be a machine, even though it
was identical to a natural person? Why?

You were right the first time: No one knows how to tell apart machines
and non-machines.

> Also a machine could be defined as being something which cannot
> reproduce or survive independantly in the environment and is not able
> to look afer themselves successfully.

You could DEFINE it that way; but I could also define a machine as
anyone who can't make a living: Does that make a homeless person a

> A machine is merely a stand-in
> able to carry out specific mechnical functions.

A stand-in for what? Which mechanical functions? Why those? And what is
"mechanical" and "function"?

The take-home message about the "What Is a Machine" was that no one
knows, and all attempted definitions have been either arbitrary, or
obviously sorting things wrongly, or both. The only thing you can safely
say is that a machine is a system that is bound by the laws of physical
causation, but unfortunately, according to that definition, everything,
and every little part of everything, is a machine. And that definition
may even be correct! "Machine," "mechanical system," etc. may just boil
down to: causal system. The only thing that makes us think we are
something else is that we think our minds can cause things
"nonmechanically": Mind over matter?

> It is not conscious and
> probablly does not have feelings. Using this definition we are not
> machines.

But this definition, like all of them (including mine) is arbitrary.
(Mine also has the same problem as Universal Grammar: It has only
positive instances; everything is a machine; so nothing is NOT a
machine; with no negative instances, we don't really know what a mahcine
is after all.)

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:43 GMT