"People who tell you they’re not superstitious are lying."
Frankie Dettori, Jockey, Observer Magazine, 5 January 2002.

Since its birth in Ancient Greece, philosophy has sought the holy grail of
certain knowledge. In this respect, philosophy reflects a common human desire
to have things clear-cut. This desire can be satisfied – psychologically, if
not logically – through the adoption of beliefs which are immunised against
the very possibility of error.

Dettori’s assertion about superstition is a striking example. He believes that
everyone is superstitious. The problem is, of course, that some people claim
not to be. If, however, he adopts the maxim "People who tell you they’re
not superstitious are lying," then no such avowals count as evidence against
him. Accepting you are superstitious supports his thesis; denying that you are
simply shows you are a liar, and again fits the thesis.

Of course, it is possible to argue that someone might behave in such
as way as to show Dettori is wrong, but since Dettori’s claim is about what
we really believe deep down, the mere fact that someone doesn’t manifest
their superstition in behaviour should not count as counter-evidence.

Traditionally, this kind of claim would be called unfalsifiable, meaning that
nothing would count as evidence for its falsity. Jean-Paul Sartre seemed to
make a similarly unfalisifiable claim when he claimed that we all feel anguish,
and the reason why some people don’t appear to be anguished is because "they
are merely disguising their anguish or are in flight from it." What that
means is that no one can be held up as a counter-example to the thesis. A lack
of apparent anguish can always be explained away as the result of disguise or
flight.

It is not generally the case that people deliberately make unfalsifiable claims
as part of a conscious strategy to immunise themselves against error. On the
contrary, the popularity of such assertions is precisely their apparent certainty.
The fact that no evidence exists to counter a thesis is usually a good reason
to suppose it is true. But sometimes we fail to notice that the lack of counter-evidence
is due to the fact that its very possibility is ruled out by the claim being
made. It is like a court case where the only admissible witnesses are those
who support the prosecution. In such a rigged trial, it is not surprising that
all the evidence falls on one side.

The reason I prefer to talk about immunisation against error rather than falsifiability
is that the key structural point about this bad move is not the relation to
evidence but the way in which the assertion contains within itself the mechanism
to repeal all counter arguments. This helps avoid the red herring defence made
of many such claims: namely, that the more inconceivable it is that some evidence
could prove a thesis wrong, the more certain it is that the thesis is true.
Hence the most certain beliefs are precisely those for which nothing could conceivably
count as counter evidence.

I think this defence is flawed, but by dwelling on it we miss the crucial point,
which is not essentially about evidence. Rather it is that the apparent certainty
is not because the thesis captures a truth about the world; it is merely a result
of the way in which the thesis has been formulated. In other words, it cannot
be wrong because it has been (implicitly or covertly) stipulated that it cannot
be wrong, not because it is right.

Immunisation against error is most evident in conspiracy theories, since these
stipulate that any apparent counter-evidence is really evidence of the effectiveness
of the conspiracy. "They would say that," is how apparent counter-evidence
is rebutted. Yet even if we are not attracted to conspiracy theories, such ways
of thinking are quite common. Have you never, in an immune system-like response,
repelled a view that contradicts your own with a "they would say that"?
If you say you haven’t, it’s my bet you’re either lying or in denial. Touché.