Philosophy of the Sciences, 1997
lecturer: Professor Laurence Goldstein
Lecture 2: Thought Experiments I
1. A thought experiment in science is a creative processes in which
the scientist obtains a result not by the usual methods of constructing
apparatus, running tests, observing and measuring, but by thought alone.
2. There's a tendency to think of scientific practice as divided into
theorizing and experimenting. Until quite recently, philosophers of science
were mainly interested in the former, and experimental work was regarded
as a low status occupation. Some modern writers (e.g., Hacking, Galison)
have attempted to show that this is a misconception, and have given detailed
consideration to experimental work, not least because such work often proves
to be a stimulus to theorizing. In fact, one might argue that the division
between scientific theorizing and experimenting is artificial. Thought
experiments are an illustration of this point. Here experiments are not
conducted in a regular laboratory, but in the `laboratory of the mind';
as in the construction of theories, what is required is imagination and
pure thought. The apparatus for such experiments is created by the imagination.
3. Science and philosophy (particularly ethics) have been particularly
fertile fields for thought experiments, and this fact is quite easy to
explain.Sometimes, in science, we need to consider ideal situations
- frictionless planes, perfectly elastic bodies etc.; these don't exist
in the real world, but we can mentally create such fictions. Also, there
are experiments which it would be physically impossible for an experimenter
to perform - e.g. running along a light ray at the speed of light, and
observing it -- but such experiments can be done mentally.
4. Ethics deals with the moral dimension of practical situations, but
there are obvious ethical limitations to the interventions we can make
in actual situations to `see how things turn out'. But there is nothing
to prevent us creating morally problematic situations in the mind
and investigating the consequences of different kinds of imagined behaviour
of individuals in such situations.
5. A thought experiment on abortion (Judith Thomson's
violinist example).
First consider the conservative arguments against abortion. Now look
at Thomson's violinist example. Is her thought-experiment convincing ?
6. One question that we should keep in our minds as he discussion proceeds
is why thought experiments work. This is a particularly acute problem
in the field of scientific thought experiments. For here we seem to be
able to obtain firm empirical results -- scientific laws -- without having
to do any empirical work, without having to make observations or use measuring
instruments. It seems as if we are `getting something for nothing'.
7. Galileo: proof that bodies of unequal weight must fall at the same
rate. Assume pre-Galileo that heavier bodies fall faster. Now imagine a
falling heavy body catching up with a slower light one, and the two fusing.
The slower body must slow the faster one down (cf. grabbing the saddle
of a faster cyclist overtaking you) so the fused body now moves slower
than the heavy body; but the fused body, since it's heavier than
the heavy body must, by our starting assumption, move faster. These
last two conclusions are contradictory, therefore the original assumption
must be rejected (reductio).
8. Galileo: From the law of equal heights (an (ideal) ball on a curved
track will retrieve its original height) to a substantive conclusion refuting
Aristotle's theory of motion. Lengthen one side of the track -- ball returns,
on a longer path, to its original height. Now lengthen that side by laying
it horizontal to infinity. Since the law of equal heights says that the
ball must continue to roll until it regains its original height, we can
conclude that the ball will continue forever in a straight line. This result
destroys Aristotle's supposition that some external force is needed just
to keep things moving.
9. Leibniz' thought experiment for refuting Descartes' Law of Collision.
Descartes says that if a smaller object hits a larger one it rebounds with
equal speed; if a larger object hits a smaller one, they move together
in a way that conserves the total quantity of motion. Leibniz asks us to
imagine one collision in which A strikes an imperceptibly larger B (so
according to Descartes, A should bounce back with equal speed while B remains
stationary). But next suppose that A is slightly enlarged so that it is
imperceptiblly larger than B and again, we cause it to strike B. According
to Descartes now the two should move off together at a speed approx half
A's initial velocity. But it's absurd to suppose that such a tiny change
could have such a dramatic effect, ergo the Cartesian view is wrong.
10. Simon Stevin's thought experiment (1605)
to show that the force needed to stop an object sliding down a plane is
(as we should say) proportional to the sine of the incline.
11. A thought experiment on personal identity (Derek Parfit). I find
Parfit's thought experiments illuminating and capable of refinement (see
e.g., his `The Indeterminacy of Identity: A Reply to Brueckner', Philosophical
Studies 70 (1993), pp.23-33).
12. (Before next meeting) Consider the scientific thought-experiments
mentioned here, and any others you know. Is this a more or a less satisfactory
way of reaching a result than normal experimentation ? Are the results
really obtained just by thought alone ?
GO BACK TO COURSE OUTLINE
| GO TO MORE ABOUT THOUGHT EXPERIMENTS