|
|
|
Contents |
In the first presentations, I argued that in general terms the scientific perspective, which looks to causes, created difficulties for our belief in human freedom.
Now I want us to look at some particular examples of sciences, or self-styled sciences, which have set out to study human beings.
Our question again will be: what are their implications for our belief in human autonomy?
The first example I want us to think about is cognitive psychology.
Cognitive science is the strongest of these. It has really replaced behaviourism as the form taken by scientific psychology.
| Mandelbrot music |
It is the approach which, broadly, takes the brain to be an information-processing device - a computer if you like.
It is the embodiment in serious science of the picture of the human being that has hung over our culture for some time now.
It is the picture of the human being as essentially a robot - the flesh-and-blood body - controlled by a computer, the brain.
Perhaps this picture is especially plausible against the background of what we think of the human being's place in nature.
If we think of ourselves as a product of evolution, if we think that we have got the way we are through the workings of evolution, it is plausible to think of that survival value has attached to good information-handling, and that this is essentially what the brain is about.
It is our device for gathering information about the environment and processing it, so that we make sensible decisions - biologically sensible decisions.
It is plausible to think that this is why the human animal has been so successful. We have got better and better at information processing. The better and deeper our 'data', the quicker and more sophisticated our ways of analysing it - the better our decision-making would have been and the keener the competitive edge.
I'm just sketching a modern picture of the human being and how we might have arisen.
 |
| Courtesy Life Science Resources |
A modern picture, I mean, which many of us will share in broad outline, of the relation between ourselves and the rest of nature.
It is the picture that shows the human being as very much part of nature, a species among other species - more successful than many others, but a biological species nonetheless.
This is the picture that Darwin gave us in the 19th Century, exciting many of his contemporaries, outraging the rest.
It is not the only picture, but it is very much the one we have now.
It is sometimes hung, this picture, in a religious setting, with which I guess it is compatible. We may be the product of evolution, it may be said, - we may be a species among other species, but there is, behind evolution - behind the physical universe and its laws - a Being, a Being who made it all, perhaps sustains it all, perhaps cares for it all.
Either way, with God in the background or not, this Darwinian picture of the human being as the immediate product of natural evolutionary processes has a place for the brain.
It supports the idea that the brain has evolved as an information processor to bring ever more intelligent control over the body.
But you can see I think how this conception of the brain carries with it the threat of determinism for human beings.
What room is there in an information-processing device for autonomy or freewill?
You come at the computer-driven robot conception of the human being from all sorts of different directions today.
As you know, one problem that has continued to exercise philosophers throughout the Modern period - I mean since the 17th Century - is the nature of the mind.
If we are robots, even computer-driven ones, could we have minds? Could we be conscious?
This is a question you already have on your agenda. The question now, in the context of determinism and free-will, is whether a computer could have free-will.
Like the question of time-travel, the question of robots thinking for themselves appeals to the public imagination too. Though this is nothing terribly new.
Could we just remind ourselves of some characters in fiction or film that are presented as examples of a machine becoming autonomous?
There are interrelated issues, with films differing on their focus.
 |
Dr Frankenstein in his laboratory, still from Mary Shelley's Frankenstein, Director of Photography Roger Pratt. I got the pic from Resources for the study of Mary Shelley's Frankenstein by Martin Irvine which is now decamped. A more imaginative approach to resources anyway is by Costume Discounters, with whom I'm sorry to have no financial connection. Providing you with a Frankenstein costume is the least they will do. |
Mary Shelley's Frankenstein for example focuses on the hubris of science in daring to emulate God.
There is the question of whether robots could be got to feel - we think of Arnie in Terminator 2 (or at least I do!) coming to an understanding of sorrow, but knowing that he himself could never weep.
Star Trek is concerned with the emotionality of human beings, the Vulcan representing rationality which begins cold but acquires an emotional underpinning which is more and more apparent, and more and more impregnable as the stories develop.
There is the capacity to learn, emphasized by War Games, where the computer has to learn in order to give up his deadly game of Nuclear War.
And then there is autonomy itself. HAL in 2001 was thought by the astronauts to have developed a will of its own - though they were I think, according to the plot, mistaken about this. There is the intelligent nuclear bomb in Dark Star, which is led to see the force of the Cartesian argument I think therefore I am and is persuaded to stop the countdown which would annihilate it (and its astronaut companions) by the thought that its input may be false.
Summary: MIGHT MECHANISMS BEGIN TO DISPLAY AUTONOMY? TWO POSSIBILITIES PERHAPS
What you have in the imagination of a whole host of science fiction writers is the prospect of artificial, human-made, devices getting out of control.
Not in the way a lorry's brakes fail with it careering blindly down the hill but somehow escaping their programming, becoming independent, acquiring ideas of their own, acquiring autonomy and exercising it to pursue those ideas, perhaps at the expense of those who manufactured them.
How are we to think of this possibility - the possibility of mechanisms becoming autonomous?
The issues here are presumably the same as those involved in asking whether we ourselves might be mechanisms but at the same time autonomous.
Here is one line of thought.
As you program your robot in ever more sophisticated ways, programming it not just with oodles of information but with the capacity to learn, to acquire and systematize new information as it goes along, as the program gets bigger and bigger, more and more complex, what you will get is a robot which is gets more and more like human beings.
And there will come a point, some people predict, don't they, when you won't be able to tell the difference.
 |
|
Alan Turing, courtesy Alan Turing.net, an excellent site on Turing, thanks to Jack Copeland and Gordon Aston Great resource on the Turing
test, thanks to Ayse Pinar Saygin |
Alan Turing, a major 20th Century figure, rather neglected, who worked just down the road at Manchester, set this out as a formal test. Not for autonomy exactly, but for 'thinking'.
He said the moment you can't tell the difference between a machine and a human being you will have to say the same things about both.
If one is autonomous, you will have to grant the other must be.
If you say one is capable of thought, you will have to conceded the other is capable of thought as well.
Imagine you have a machine for which thought is claimed. Put it up against a human being and see if you can tell the difference. This was his proposal.
His idea was just to eliminate the distraction of the physical appearance and mode of communication of an artificial thinker.
Put him or her or it in a private room, he said, and let all communications between him/her/it and the examiners, who sit in a room by themselves, be done through a keyboard.
In another private room put a human being, also communicating via a keyboard.
Get the examiners to ask what questions they like through their keyboard. The responses of the human being and the computer come up on different screens.
If after the 'conversations' have gone on for some time the examiners can't tell which are the responses of the human being and which the computer, the computer is deemed to have passed the test.
What would be a reasonable period for this test, do you think?
(A) TRUE AUTONOMY TURNS OUT TO BE A MYTH AS HIGHLY COMPLEX MECHANISMS MIMIC IT SUCCESSFULLY.
(B) TRUE AUTONOMY 'EMERGES' FROM HIGH FULLY DETERMINED COMPLEXITY.
Once you have got a machine that passes the Turing Test, what moral will be there for autonomy?
Here is a mechanism, you may say, which is behaving exactly like a human being.
But we made it! We know it is a mechanism. We programmed it to work in a certain way. Because we can get a machine to copy human behaviour in this way, human freedom is a myth.
There is though another way you could point the gun - often in philosophy an argument can be pointed in more than one direction.
B. TRUE AUTONOMY 'EMERGES' FROM HIGH FULLY DETERMINED COMPLEXITY
| 'As we look at the continuous course of Evolution, we ought not to say: "After all, it all comes from a nebula with no life or meaning, so it, too, is without life or meaning." Rather we should say:"See what has come from that dead nebula; how full of potentiality it really was." 'William temple, The Nature of Personality, 1911, p. xxix |
You could say: here is what we thought was a mechanism - an elaborately programmed computer - but what we discover is that it can be got to mimic human behaviour. We know that human beings are autonomous. These things that we have made and have thought of as mechanisms are evidencing freewill. Therefore they cannot really be mechanisms at all. They may have begun as mechanisms, but somehow they must have 'taken off', as it were - turned into something of a different order - turned into system capable of free choice.
A term that is sometimes used in this connection is 'emergence'.
Something utterly new it is said 'emerges' from what began as mechanism.
There are precedents for this kind of thing, it is argued. At one time, so it is believed, there was nothing but gas. Different elements, but all in the gaseous state. At a certain temperature, and as conditions modified, some of the gases combined. Oxygen and hydrogen combined and produced something with a property that had not occurred before. What they produced was water and the innovative property was wetness.
Might not freedom be like that - something utterly new that emerges when mechanical contraptions get above a certain level of complexity?
Can you think of other examples of 'emergents' - really new things or really new properties that emerge from what went before.
Summary
Application of the scientific perspective to the human being
Cognitive Science - the brain as an information-processing device
A test for machines: the Turing test
Could autonomy emerge? - could what starts as a mechanism turn into something with autonomy
|
Revised 06:03:09| Prepared by VP Freewill and Determinism Home page A module of the BA Philosophy programme Center for Professional Ethics | University of Central Lancashire |
