phl of mind paper
TRANSCRIPT
-
7/31/2019 Phl of Mind Paper
1/7
David Krajewski
Philosophy of Mind
Topic- Discourse V & Turings Test
Do only human beings have the capacity to think? Descartes concludes in Discourse V
that the answer is yes and machines do not have this capacity. Turing, in contrast, concludes that
a machine (a digital computer) does have the capacity to think. I will first present Descartes
argument and the criteria he uses for his determination. Then, I will present the criteria that
Turing uses to draw his conclusion, which is commonly called The Turing Test. Finally, I will
discuss an objection that Turing considers against The Turing Test and show how it is relevant
to Descartes argument.
Central to Descartes argument is the differentiation between a brute and a human being.
The differentiation has to do with the disparity in eachs capacity to respond to the external
world. A brute is any animal that is considered non-human, like a cat or a bear and so on. It is
something that is entirely deterministic and limited due to its biological program. In other
words, is limited to a set of pre-programmed responses. For instance, a cat is pre-programmed to
respond in a certain way to certain stimuli. If it feels dirty, then it will begin to lick itself clean.
If it feels physically threatened, then it will attempt to defend itself with its sharp claws. Try to
conceive of a cat that does not do those kinds of things in response to those kinds of stimuli. It is
difficult to do so for the cat is severely limited due to its biological program. Now consider how
a human responds to various things in the world. If she feels physically threatened, then she may
defend herself or run away. And she may defend herself with a gun, or a knife or her own fists.
If she feels dirty, then she may take a shower, a bath, or simply wash her hands. Notice how, in
contrast to the cat, the human can respond to similar situations in numerous ways for it is not
-
7/31/2019 Phl of Mind Paper
2/7
limited to the extent that the cat is limited. It is also possible for a human to respond absurdly to
any situation. If he feels physically threatened, he may jump out a window or run in front of a
car and choose not to defend himself. There are limitless possibilities for him whereas the cat
can only respond in ways that its pre-programmed responses dictate it can.
A human has limitless possibilities in so far as he can respond appropriately to any
given situation whereas a brute cannot. Descartes claims that it is conceivable to imagine a
machine that perfectly resembles a brute in all physical respects. In fact, it resembles a brute so
well that it is indistinguishable from the real brute. Descartes calls this machine an
automaton. Like a brute, he claims it is limited to pre-programmed responses. It cannot
respond in limitless ways like a human can. Suppose that a human were to meet this automaton.
Consider an automaton cat. The human would observe that the automaton cat would lick itself if
it were dirty or use its claws if it were threatened. The human would have no reason to believe
that the automaton is not a real cat for it mimics the real cat flawlessly. Suppose a real human
encountered an automaton human. The human begins to ask the automaton human basic
questions like What day is it? or How is the weather?. The automaton responds
appropriately. The human now asks the automaton more specific questions like What did you
do five minutes ago? or What do you think of the current economic crisis?. The automaton is
lost for words and cannot provide an appropriate response. It may be that whoever
programmed this automaton only programmed it to respond appropriately to very basic questions
and so it cannot respond to anything very specific. The real human, aware that human
automatons exist, concludes that it is an automaton and not an actual human.
The human can tell the difference between a real and automaton human whereas he
cannot tell the difference between the real and automaton cat. He cannot tell the difference
-
7/31/2019 Phl of Mind Paper
3/7
between the latter for a real cat does the same things that the automaton cat was doing. However
in the former case, a real human would respond appropriately to the very basic questions as well
as the specific questions. Descartes points to the expression of language as making the
difference. I may formulate an answer to the question about the economic crisis, however
inaccurate my assessment is. The automaton, though, simply cannot formulate an answer to the
question for it is not capable of producing limitless language. It is irrelevant whether my answer
is right or wrong in this context for at least I can produce an appropriate answer. An automaton
cannot produce an appropriate answer at all and perhaps its attempt is a jumbled mess of words
that do not make sense. As real humans, our responses are not a jumbled mess of words for
our acquisition of sophisticated language allows us to formulate coherent sentence after coherent
sentence. There is nothing restricting us in forming these sentences (we are not held down by
a program) therefore we can theoretically converse forever.
Might it be conceivable for a programmer to make an automaton that is capable of
answering highly specific questions in detail? I think yes. Eventually though, if the
conversation with the automaton were carried on long enough, I suspect that Descartes would
assume it would stumble for it would fail to respond appropriately at some point or another. If I
as a real human were in an endless conversation, I could produce appropriate response after
appropriate response without ever faltering. Eventually all real humans die and so it is
practically impossible to forever respond appropriately. However, it is logically possible for a
human to respond appropriately for an infinite amount of time. For Descartes, the capacity to
respond appropriately for an infinite amount of time equates to thought. Therefore, a human
can think and machines and brutes cannot.
-
7/31/2019 Phl of Mind Paper
4/7
Turing does not distinct between brutes and human beings. He instead presents The
Turing Test. The test involves a game that is to be played between two players and an
interrogator. Player A and the interrogator are human beings whereas Player B is a digital
computer. The objective of the game for the interrogator to successfully decide which player is
the human and which player is the machine after posing a series of questions to both players (he
knows that one player is the computer and one the machine going into the game). It is the
machines objective to fool the interrogator into getting the players mixed up. Turing supposes
that each player is unseen to the interrogator and the questions are entered in on a typewriter.
Suppose that the interrogator asks How hot is it in the room? or What move should I make in
chess under such and such circumstances?. Keep in mind that the computer is a sophisticated
machine and carefully programmed to respond appropriately to a great deal of questions. It does
so and the interrogator might think the answers are more detailed and articulate from Player B
(computer) than from Player A (human). He then might be lead to believe that the more detailed
and articulate answers came from a human. If after his questioning the interrogator decides that
Player A is the computer and Player B is the human being, then he has failed. If the interrogator
fails to discriminate between the human and computer player, then the computer passes The
Turing Test. Turing claims that if the computer passes his test then it should be concluded that
the computer has the capacity to think. For Turing, the computer does not have to respond
appropriately to an infinite amount of questions to have that capacity, unlike in the conclusion of
Descartes.
There is an objection that Turing considers that Descartes himself might pose. It is the
Lady Lovelace objection which claims that the computer cannot really do anything new; it can
only do what it is programmed to do. Descartes argument concludes that human beings can do
-
7/31/2019 Phl of Mind Paper
5/7
new things in the sense that they can produce appropriate responses in any given situation. It
follows from the above objection that computers are constrained to their computer program.
Computers cannot produce anything new for what they can produce is already ingrained into
them and limited by their programmers. This constraint is analogous to what brutes and
automatons are constrained by, which is their biological program. Brutes and automatons are
only capable of producing what is ingrained in them by their biology and it is their lack of human
language that makes this so. Turing may concede that brutes and automatons have this
constraint; however he does not believe that computers share it.
Turing would agree that brutes and automatons are limited by their biology. He would
respond to Descartes (and the above objection) by assuming that the entire world is determined
by physical laws and its inhabitants are determined by physical as well as biological laws. It
would follow that brutes, automatons and even human beings are constrained by these laws.
Brutes as well as humans would be constrained by their biological programs. So, just as humans
are programmed by biology (through genetics) a computer is programmed by its human
programmers. If a human is biologically programmed just as a computer is technologically
programmed, can a human do anything new? The Lady Lovelace objection seems to be
committed to making that claim for it implies that being programmed in any sense is to be
limited in what one can do. Genetics determine in part, personality and all of ones physical
appearance. If my personality is a result of my genetics, then how can I really do something
new? How can I break apart from my genetic makeup? Again, it seems that the objector has to
say that I am completely constrained by my biology and therefore cannot do anything new.
The above consequence of the objection is certainly undesirable for Descartes. He would
not want to conclude that human beings cannot do anything new due to their biological program
-
7/31/2019 Phl of Mind Paper
6/7
for it would be in direct contradiction to his argument. Instead, Descartes would point to how a
human uses language compared to how a computer uses language. Much like the human
automaton, the sophisticated computer can carry on a conversation with a human being. The
conservation might even go on for a substantial amount of time and the human would not realize
that she is talking to a machine (assuming the machine is behind a wall or unseen in some way).
The computer is very advanced and it can even carry on an intelligent philosophical debate.
However, just like the automaton, it would eventually run out of things to say. It would be
unable to formulate a coherent response to a question at some point for its program would have
pulled out all the stops. For Turing, this consequence is not important. We can imagine that
the computer in question would pass The Turing Test after a reasonable one hour conversation.
Merely passing The Turing Test would not satisfy Descartes. He would have to conclude that
the computer can go on forever in conversation or else he has to say that it does not have the
capacity to think. It turns out that it does not respond appropriately forever and so Descartes
concludes that the machine (however intelligent) cannot think.
Descartes has shown then that a computer cannot do anything new which was what the
Lady Lovelace objection was trying to establish. Turings reply shows that the objection leads
to an undesirable consequence. However, Descartes argument from Discourse V gets away
from that undesirable consequence by pointing to the differences in language present in the
machine and the human. Descartes might concede that biology limits human beings however it
does not limit them to the extent that it limits brutes and automatons. And, in response to
Turing, biology does not limit human beings to the extent that a computers program limits a
machine. A human can respond to an infinite set of questions whereas eventually the machine is
bound to fail.
-
7/31/2019 Phl of Mind Paper
7/7