Peter Bennett An Essay concerning Personal Identity 24
4 The questions in the previous section have defined the domain
for this one. The problematic thought experiments which arise in the
philosophy of personal identity are the ones which receive the
answers yes (or maybe) to Q1 and no to Q2, Q3 and Q4. It is such
thought experiments which are under discussion in this section: we
are now looking at thought experiments which appear to satisfy the
constraints required of real experiments, do not 'merely take place in
thought', are not 'merely imagined' and are not merely illustrative.
Given that one has answered no to Q3, the obvious question to
ask now is:
Q5: Why is the thought experiment impossible?
Parfit16 makes the distinction between deep and technical
impossibility. A thought experiment which yields negative answers to
Q2 and Q3 is deeply impossible if it violates a law of nature.
Otherwise it is technically impossible.
Similar to this is Wilkes's idea of background conditions. In
Chapter 1 of Real People, Wilkes is out to establish that in both
real and thought experiments, appeal is made to background
conditions. These may be explicitly stated, e.g. all the mice in the
experiment were of the same sex, or implicit, e.g. none of the mice
showed signs of poor health, or considered to be irrelevant, e.g.
some of the mice had longer whiskers than the others. Wilkes
claims that whilst in science there is a body of theory which
provides the background conditions and tells you what may or may
not be left out, in the philosophy of personal identity there is no
such body of theory.
Unlike Wilkes I do not see this as ruling thought experiments
out completely from discussions of personal identity; merely that
whilst in physics you may be able to broadcast your thought
experiment generally, in philosophy you have to say to whom within
the discipline you are addressing your thought experiment.
_
16 Reasons and Persons p. 219.
Peter Bennett An Essay concerning Personal Identity 25
Even within physics who your audience is makes a difference.
The discussion between Bohr and Einstein that led to the following
thought experiment could only have taken place between two
physicists who shared acceptance of general relativity and positivist
doctrines but disagreed about quantum mechanics. In the discussion
Bohr is out to show that the thought experiment which Einstein
claims to be a technical impossibility is in fact a deep impossibility.
TE5: Einstein's clock in a box (1930)17
The aim of this experiment was to come up with a situation in
which one had more information about a particle than was permitted
by quantum mechanics, thereby showing quantum mechanics to be
false. The objective is to measure the energy of a photon at a
particular time.
A box is suspended from a spring. On one side of the box is
a hole with a shutter over it. The shutter is linked to a clock in
the box. On the other side there is a pointer which points at a
ruler, so that the pointer and the ruler form a spring balance.
The shutter, which is controlled by the clock, opens for just
long enough to allow a single photon to escape. And as Wilkes tells
us 'by weighing the box both before and after the event, one could
measure the energy of the photon, by exploiting the famous formula E
= mc2. This would have been, as Bohr said, 'in definite contradiction
to the reciprocal indeterminacy of time and energy quantities in
quantum mechanics.'18
_
17 Real People pp. 8-9. See also N. Bohr, 'Discussion with
Einstein of Epistemological Problems in Atomic Physics'.
18 N. Bohr, 'Discussion with Einstein on Epistemological Problems in
Atomic Physics' p. 226.
Peter Bennett An Essay concerning Personal Identity 26
To see what the problem is with this, we have to think about
what is involved in weighing the box. The weighing operation at the
start of the experiment must not have a significant effect on the
momentum of the box. Therefore low-energy, i.e. long-wave,
radiation has to be used to measure the position of the pointer.
And long-wave radiation cannot give a very precise reading.
To compensate for this one might use a highly elastic spring,
one that will expand a long way for a small increase in the load
placed on it. The problem with this is that the clock in the box
would now be travelling at high speed in a gravitational field, and so
the considerations of general relativity become significant. This
means that there will be great uncertainty about the rate of the
clock. Hence, Bohr claims, the reciprocal indeterminacy of time and
energy quantities is preserved. To put this in a formula:
E.t > h
That is, the product of the uncertainty in energy and the uncertainty
in time is greater than the constant h.
'The general moral', as Wilkes says, 'is clear: just as with real
experiments, thought experiments presuppose that all the relevant
background conditions are included and specified.'19 In TE4,
apparently all the background conditions have been taken into
consideration - at least no-one has come up with an argument to
show they have not. In TE5, however, Bohr has come up with such
an argument.
Q5, 'why is the thought experiment impossible?' can therefore
be broken down into three interrelated questions:
Q6: Is the impossibility deep, or is it technical?
Q7: What are the background conditions?
Q8: Who is the thought experiment intended to persuade?
_
19 Real People p. 9.
Peter Bennett An Essay concerning Personal Identity 27
Q7 and Q8 can be combined to give the yes-or-no type
question:
Q9: Would the people the thought experiment is trying to
persuade accept the background conditions?
The claim of impossibility may come about in two ways. The
first is to say that the teletransporter cannot be built or the brain
cannot be successfully transplanted. This is a dangerous way to
argue since technology might prove you wrong. In the 17th century,
surrogate motherhood, which yields problems similar to those posed
by personal identity thought experiments, was a technical
impossibility. In the 23rd century teletransportation could be a
reality.
The second way to claim impossibility is to claim a logical
impossibility. This would straight away entail deep impossibility.
Some people believe that personal identity is organism identity. For
them, swopping large amounts of tissue from one operating trolley to
another would violate an important requirement for personal identity.
Teletransporters would in a similar way be ruled out by these people.
Impossibility does not of itself rule out the use of a thought
experiment. As Bohr remarked in his account of TE5:
The discussion, so illustrative of the power and
consistency of relativistic arguments, thus emphasized
once more the necessity of distinguishing, in the study of
atomic phenomena, between the proper measuring
instruments which serve to define the reference frame and
those parts which are to be regarded as objects under
investigation and in the account of which quantum effects
cannot be disregarded.20
_
20 N. Bohr, 'Discussion with Einstein on Epistemological Problems in
Atomic Physics' p. 228.
Peter Bennett An Essay concerning Personal Identity 28
We have to be clear about which parts of the thought experiment
have to be possible and about which parts of the thought experiment
are allowed to be impossible without admitting the possibility of the
pollution of the result. One thing one has to be very careful about
is the causal links. In this case Bohr has identified an unwanted
causal link between the energy required to measure the initial position
of the box and the momentum of the box. We have thus found a
confounded variable - the measuring device is showing the effect of
the energy of the light used to find its start position as well as the
effect of the momentum of the photon. This is highly relevant given
the shared background conditions of the participants as the thought
experiment is concerned with positions and energies.20
A common objection to thought experiments, the 'but that's
impossible!' objection with which I opened this chapter, can be put in
the context of quantum mechanics. Quantum indeterminacy implies
that it is physically impossible, therefore deeply impossible, to build
a machine which would create perfect replicas of a human being.
But such an objection misses the point if personal identity is taken
to be a moral concept, since if one refuses to derive ought from is,
then sub-microscopic physics should have nothing to say about
morality. For this reason, thought experiments which invoke magic at
suitable points are superior to thought experiments which indulge in
quasi-scientific excuse-making: they are being quite honest about
where the impossibilities take place, and do not waste time in
attempts to cover objections about irrelevant features being
impossible.
Thought experiments thus bear some resemblance to
phenomenological reductions: they bracket out certain features, in this
case certain physical laws, as being irrelevant. In The
Phenomenology of Perception Merleau-Ponty casts doubt on the
feasibility of doing phenomenological reductions arguing that they are
dualist. There is a difficulty for the monist who refuses to polarise
the physical and the mental: you cannot bracket one out and say that
there are no implications for the other. Objections to thought
experiments can be made on similar grounds, but only if one refuses
(and I do not) to subscribe to an is-ought dualism.
Peter Bennett An Essay concerning Personal Identity 29
It follows that is-ought monists, i.e. those who think that
what is and what ought are inextricably intertwined, should be much
more wary of fantastic thought experiments than is-ought dualists
such as myself. Very few thought experiments will get past Q9 if
they are addressed to an is-ought monist. I shall not deal further
with the can-one-derive-ought-from-is issue in this thesis.
If one is having problems in answering Q7 the next question
may help.
Q10: Is the thought experiment positive or negative?
By a positive thought experiment, I mean one which is arguing
for a specific conclusion, e.g. that memory is what is of paramount
importance for personal identity. TE1 is an example of this.
Negative thought experiments are experiments which do not give
specific conclusions, either because their whole duty is to show that
a theory is absurd, as was the case with TE4, or because their
purpose is to show that things which happen to be the case in this
world could be different: if we altered certain facts about what is
the case in this world, then certain other things would be likely to
follow. Negative thought experiments tend to require weaker
background conditions than positive ones.
The next experiment illustrates this. Wilkes here sets up a
thought experiment and then claims that the background conditions are
deficient. Whether the background conditions are deficient or not is
dependent on what the experiment is trying to show, but there is
nothing in Wilkes's text to say what this is.
TE6: Amoebae
Consider next one of the familiar thought
experiments to do with personal identity: that we might
all split like amoebae. It is obviously and _
Peter Bennett An Essay concerning Personal Identity 30
essentially relevant to the purposes of this thought
experiment to know such things as: how often? Is it
predictable? Or sometimes predictable and sometimes not,
like dying? Can it be induced, or prevented? Just as
obviously, the background society, against which we set
the phenomenon, is now mysterious. Does it have such
institutions as marriage? How would that work? Or
universities. It would be difficult, to say the least, if
universities doubled in size every few days, or weeks or
years. Are pregnant women debarred from splitting? The
entire background here is incomprehensible. When we ask
what we would say if this happened, who, now,
are 'we'?21
This kind of thought experiment sounds familiar, but it is a
hypothetical example. Since what Wilkes is trying to preach is that
we ought to be wary of hypothetical examples, perhaps she would
have done better to have picked a thought experiment from the
literature, as she did with the physical thought experiments, rather
than make up her own. Applying Q1 - does the thought experiment
observe the constraints by which actual experiments are constrained?
- we find the thought experiment deficient in that the thought
experiment lacks any clearly stated objective. Every experiment is
out to prove or refute some thesis: what is the thesis to be proved
or refuted here? If it is something specific, e.g. that there is or
there is not a problem with making the long term promises involved in
marriage vows, then the way in which marriage works or fails to
work in such a society would be highly relevant. But it could be
that the thought experimenter is merely trying to show that marriage,
universities and other institutions in the imaginary world described
would be very different - if indeed anything like them existed at all
- from the institutions we are familiar with in our society. If this
was what the experiment was out to show, then the social
institutions would not be possibly relevant but uncontrolled variables
- they would not be _
21 Real People p. 11.
Peter Bennett An Essay concerning Personal Identity 31
'background conditions' - rather they would be the independent
variables of the experiment. So in this example we see that positive
thought experiments, which make stronger conclusions, require
stronger premisses or background conditions than negative thought
experiments.
Another problem with TE6, though this has nothing to do with
the fact that it is a thought experiment, is the use of 'we'. It is
because 'we' is used in such a careless way at the start of the
experiment that Wilkes can ask the rhetorical question at the end.
This careless use of personal pronouns often happens in discussions
of personal identity. Since what is at stake in these discussions is
what the referents of these pronouns are, one has to be very
careful when using them. In TE6, 'we' is sometimes used to refer to
the observers and sometimes to the subject of the experiment.
Fortunately it is often possible to repair the argument by substituting
suitable expressions for the offending pronouns. What one loses in
elegance, one gains in precision. Our next thought experiment is a
case in point.
One final note about TE6: Wilkes's complaint about the amoeba
is not nearly as specific as Bohr's complaint about Einstein's clock-
in-a-box. Bohr says exactly how the results of the experiment are
polluted by the radiation required to read the position of the box.
Wilkes does not say exactly how the results of the amoeba
experiment are polluted, she merely points in the general direction of
some background conditions.
I have now completed my list of questions. Thought
experiments which are clear about what their purpose is, what they
assume and who they are trying to persuade can be a useful tool.
But the profligate and careless use of this kind of device are, just
like hard or misapplied words were to Locke, 'but the Covers of
Ignorance, and hindrance of true Knowledge.'22
_
22 Essay, Epistle to the Reader; Nidditch p. 10.
next
back
contents
Copyright Peter Bennett 1992