2014-02-14

The Science Delusion (in defence of philosophy)

Darwin versus the philosophers

Don’t worry, I’m not going to go off on some sort of post-modernist relativist rant. I do think science is a “privileged” way of looking at the world. But I wish to defend philosophy.

I’m not entirely sure what Richard Dawkins had in mind when he wrote the above tweet – though he has provided some further clarification in subsequent tweets. My initial inclination was to respond with:

On this 127th anniversary of Borodin’s death (1887-02-15) it is a severe indictment of science that no chemist anticipated Prince Igor.

But then I remembered that Alexander Borodin invented a method for the identification of urea.

Anyway, the more general point I wished to make was that it seems to be a common belief amongt scientists that philosophy is basically a load of bollocks and that all real questions can be addressed by science. Actually, this is (sort of) the view of some philosophers (but let’s not go there) and is also (though I’m paraphrasing his actual remarks) the expressed view of another scientist for whom I have a great deal of respect: Prof Brian Cox (@ProfBrianCox).

The obvious problem with the “all real questions can be addressed by science” line is that someone only needs to retort: “oh no they can’t” (in a pantomime voice) and we’re off into discussing a question which can’t (on pain of circularity) be addressed by science.

I can’t possibly do this topic real justice in this humble blog post but let me focus on a specific example - which I hope may illustrate the role of philosophy – and let you all decide for yourselves whether you think it’s all bollocks:

The Turing Test:

Turing (most would argue – just in case you dispute what he himself actually meant) answered the question “Can machines think?” with (though I’m paraphrasing again) “yes, if they can pass the test of imitating a human so well that an interrogator of the machine (which would obviously have to be hidden from view – unless it were a very convincing robot) can’t tell whether it’s human or not”.

In other words, Turing answered the question “Can machines think?”, not (as scientists often do) in terms of underlying mechanisms but in terms of observable behaviour.

There are obvious parallels with Heisenberg's interpretation of quantum theory here

[....] man könnte zu der Vermutung verleitet werden, daß sich hinter der wahrgenommenen statistischen Welt noch eine „wirkliche” Welt verberge, in der das Kausalgesetz gilt. Aber solche Spekulationen scheinen uns, das betonen wir ausdrücklich, unfruchtbar und sinnlos. Die Physik soll nur den Zusammenhang der Wahrnehmungen formal beschreiben. Werner Heisenberg, Über den anschaulichen Inhalt der [Quantentheorie] [....], p 503.

Which translates as: "[....] it is possible to ask whether there is still concealed behind the statistical universe of perception a "true" universe in which the law of causality would be valid. But such speculation seems to us to be without value and meaningless, for physics must confine itself to the description of the relationships between perceptions." Translation (unattributed) in: Coley, N G & Stannard, R; Quantum Theory (The Bohr - Einstein Debate); p 109.

But, again, let’s not go there. (We can leave bringing up quantum mechanics to bamboozle your audience to the homeopaths and other quacks.)

I wish to argue that, while it’s perfectly reasonable for a scientist to design Turing Tests and to decide whether a candidate machine has passed any of them, you don’t need to believe in pixies or souls or ghosts in the machine to see that there are reasonable objections to Turing’s thesis.

Questions such as “Does passing the Turing Test really imply that machines can think?”, is, I hope you will agree, not one that could ever be decided by science. It is a question that might be refined or redefined by scientific discoveries (just as the question “Can machines think?” can be seen as a modern version of the Cartesian question as to whether animals have souls) but it is, I submit, an essentially philosophical question. Just because philosophers will never provide a definitive answer to this question (in the way that scientists have – pace the views of sundry mouth breathers - provided a definitive answer to the question of how humans appeared on our planet) does not mean that the question is not worthy of attention.

There is, I submit, a whole realm of perfectly rational intellectual activity that does not necessarily lead to the formulation of empirical tests.

Of course, one reason that philosophy is a worthwhile exercise is that philosophical speculation may help to clarify the thoughts of scientists as they try to compose testable theories about the world. In fact, all scientists must and do wax philosophical from time to time (see Prof Butterworth @jonmbutterworth for a recent example: How did I get here? ) though philosophical speculation is only a part of their day jobs.

But this is a bit like our disingenuous claims (especially in funding submissions) that the true value of scientific research lies in bagless vacuum cleaners, Teflon saucepans and the full body umbrella .

We all know that the real reason that we (those of us who do) love and pursue science is that we love the intellectual challenge of thinking about bigger questions about the world.

Similarly, philosophers love the intellectual challenge of thinking about the even bigger questions about the bigger questions about the world and it would be really mean and would diminish us all to prevent them from doing this!

2 comments:

  1. "Science can never answer whether passing the Turing test implies machines can think."
    Why not?
    Scientific study of brain processes seems to be progressing rather well. Computer systems sufficiently advanced to pass a Turing test could obviously also be studied in similar ways with less effort. Why would a comparison between the two types of systems be so difficult?
    It seems to me the author seeks comfort and wants to believe.

    ReplyDelete
  2. I'm afraid you are quite wrong there - I've obviously failed to make my point.

    If you want my personal opinion, I happen to "believe" Turing is right. One day scientist will build machines that pass the Turing test and we shall declare them to be sentient machines and I can't see any other viable criterion of sentience.

    I chose the Turing text deliberately as an example so I couldn't be accused of bias.

    My point, however, is that there are plenty of perfectly sensible people (see eg John Searle) who reject Turing's arguments.

    If I ever met John Searle, I would argue the toss with him but _ and here's the crux of th matter - our argument could never be settled by science.

    ReplyDelete

Comments are moderated, but you can leave them without registering.