|
9 Instrumentalism and Interpretationism |
||
|
Instrumentalism Metaphysical and anti-metaphysical views of the mind F-ism and Identity theory and B-ism are metaphysical theories of Mind. They say what M states are. Instrumentalism
is an anti-metaphysical view. It
only talks about when we can say there are M states. Roughly, it turns out, we can say they are there when positing
them would fulfil some
purpose. Intentional
states Consider intentional states
or propositional attitudes,
characterized using indicative statements in ‘that’ clauses. E.g.
beliefs and desires. Stances and intentional systems The physical stance Given that the physical world is causally closed, we could (in
principle) predict anyone’s behaviour as an ensemble of physical parts.
This would be a prediction from the physical
stance. We can’t usually predict much about people’s behaviour from this
stance. Prediction from behavioural patterns We can make predictions of complex systems – including people – by
noting how they usually function, and their regularities. The intentional stance The most interesting and useful way to predictthe behaviour of people is
to adopt the intentional stance,
in which we assume that people operate according to a belief/desire/action (BDA)
psychology, applying practical
rationality. We don’t have to assume that the BDA reasoning is consciously reviewed
when we make a prediction in normal cases – but the expectations that we do
form would be justified in that way. We do need to understand how we come up with the hypotheses about what
beliefs and desires are being applied in each case. This would be a theory
of the interpretation of
behaviour. Charity as a starting place We’ll usually assume people believe what they should believe, given
their situation. A person we observe looking at a chair should be assumed to
be thinking ‘here is a chair’ in the first instance. But this assumption
will usually require revision. The assumption that the subject is right (or
agrees with the observer) is the principle
of charity. It is appealing since 1.
B are supposed to look like the real world. 2.
We have to have a first step, and we are good
examples. Humanity as a starting place Another approach is to apply the principle
of humanity: suppose that the subjects have the belief that we would
have if we were in their situation. All such principles try to capture the notion that we shoul suppose the
subject has the beliefs hat they ought
to have. If we make these assumptions and get the wrong answers, make poor
predictions, etc., then we need to modify the assumed B and D. It is
controversial how extensive these modifications can be in principle. The rational evolution of belief Similarly, we expect people’s B to evolve in the way that they ought
to evolve. They should be deductively
rational. E.g. If A believes ‘if X then Y’, and A then learns that
X. Then A should come to believe that Y. They should be inductively
rational. Eg. If A sees the statistical evidence that shows X, then A
should come to believe that X. We cannot assume ideal rationality; and there is a question of how
rational we have to be to be intentional. Intentional systems theory Such issues are the subject of Intentional Systems Theory. It aims to
find the principles of belief formation and revision suitable for predicting
the behaviour of intentional systems. It has 3 parts: 1.
Finding initial hypotheses (charity and humanity) 2.
Connecting current hypotheses of B/D with behaviour
(practical reason) 3.
Considering the nature of belief evolution given
unideal rationality IST should be considered as part of the theory of concepts for many
theories of Mind – CSF, for example. Instrumentalism and intentional systems theory Instrumentalism defined Instrumentalism: the doctrine that the predictive role of B/D is all
there is to them. To be a believer/desirer is to be something whose behaviour is
well-predicted by some form of IST. This is motivated by recognising the way our ideas about B/D are
conceptually linked to how we predict behaviours, and that we don’t need
to know what’s happening inside others to make these predictions.
Therefore don’t talk about those irrelevant things. There are internal
facts to be had, but they’re not part of the philosophy
of Mind. Instrumentalism
and outside control Consider the marionette example
by Peacocke. A puppet, M, is controlled by radio signals from Mars. Assuming
the controller is an intentional system, the I-ist will predict the
behaviour of the puppet successfully. However, statements that ‘M believes
X’ or ‘M desires Y’ are all false. An I-ist could fix this by saying that the causes of behaviour must be inside
the subject? It would require having to say something
about the inside of a subject; but not very much. Instrumentalism and behaviourism Like B-ism, I-ism denies that M states are inner causes. But it doesn’t have a problem matching dispositions to M states. I-ism
is mostly about what it is to have M states, and not much about what it is
to be in a M state. I-ism needs a rich story about a subject’s B and D in order to make
predictions. It also needs to acknowledge that B comes in degrees. The competition
between various B to justify an action depends on this. Similarly for D and
the competition to motivate action. The primacy of systems of belief and desire for instrumentalism This means I-ism doesn’t really consider individual beliefs. It would
say ‘A believes X’ only if that follows from the full story of the hypothesized B/D and observed actions. This raises several questions: 1.
Can this account for the way that we treat individual
beliefs as having distinctive causal roles? 2.
Can it explain failures of deductive rationality? E.g.
If A believes that triangles are equilateral only because it follows from
their total beliefs, then they must believe it is equiangular, since that
also follows. But we can fail to make all the right deductions. Are mental states in part determined by how things are with others? An objection to I-ism is that the attribution of M states to a subject
depends too much on how things are with others. The prediction of S’s
behaviour says something about S and also
something about the predictor. This
is not a problem. For it to be true that A believes/desires X it only needs
to be true that the attribution of X follows from an application of IST by perfect users of IST with all the relevant information. It does not
depend upon any particular hypothesizer, even though any particular
attribution would so depend. The Blockhead objection to instrumentalism It is true of me and of my Blockhead duplicate that behaviour is
well-predicted by adopting the intentional stance. And it is because of how
things are inside us that that is the case. However, attributions of B/D to
Blockhead are wrong, so the I-ist is wrong about what makes B/D ascriptions
true/false. Fictionalism So abandon the semantic claims and move to an eliminativist position
called fictionalism: pretending
that A has the B/D that X can be useful, but there is no claim that where
that is the case that’s what is meant
by saying that A believes or desires X. B/D ascriptions are all false, B/D
are ‘convenient fictions’. Centres of gravity Compare: we predict how bodies behave by reference to their centre of
gravity, but there is really no such object. Interpretationism Instrumentalism and interpretationism Interpretationism:
part of what makes it true that A has the M state X is that a suitable
placed expert would interpret A as having X. A suitably placed expert is the perfect user of IST mentioned above. So
Interpretationism (H-ism) is basically I-ism with explicit reference to the
perfect users of IST The
uncodifiability of rationality The reference to perfect (or other) users is in response to the fact
that rationality is an incompletely specified notion, and thus no IST can be
properly specified. In applying IST, we rely on an ability,
not an explicitly codifiable theory. H-ism is a version of I-ism that makes this explicit. Psychophysical laws and rationality Does the uncodifiability of rationality mean that there are not strict
(non-ceteris paribus) laws
connecting the physical to the psychological? If rationality is essential to
attribution of M states, and rationality is uncodifiable, then surely the
possession of M states is ‘uncodifiable’. Or: if physical facts settled
what M states existed, then they would settle what counted as rational; but
they can’t to the latter, so they don’t do the former. Abilities are not Miracles H-ists say we have the ability
to determine rationality; but this must be a physically explicable fact
about us (not magic!) The problem is just that we don’t know yet how we do this. So rationality can’t really be uncodifiable ‘in
principle’. The Blockhead objection to interpretationism Experts would make the same attributions to you and your Blockhead. But
they would be wrong about Blockhead at least, so being in a state to be
assigned M states by an expert is not
what it is to be in a M state. H-ists can do as the I-ists do; let their experts take account of what’s happening inside the subject, but that’s going to change H-ism into some other kind of theory of Mind – probably CSF-ism.
|