Effect and Cause, Part 3
“Theory
without empirical evidence is mere speculation. Empirical evidence
without theory is simply a collection of facts. Theory supported by
empirical evidence is science” (From my book, Freedom or
Serfdom?”
We’ve discussed how there can be several possible causes for an
effect, several theories as it were about how it happened. That of
course raises the question, “How do we know which, if any, of those
theories is correct?” A good question, an important question, and
sadly a question with an unpleasant answer. The simple answer is
that, in most cases, the answer is beyond human ability. We cannot
know if some new theory will come along, better than what we have.
Nor can we know if new data will invalidate what now appears to be a
solid conclusion.
Perhaps the best example of this is the physics Isaac Newton gave us.
There was a saying that Newton was lucky, there was only one universe
and he got there first. At the start of the twentieth century it
looked like physics was complete, only a few minor questions to
answer and all would be known. Scientists regarded Newtonian physics
as essentially a perfect theory, well established and never to be
overturned. Then along came people like Einstein, Bohr, and Planck.
Today we know that Newton’s physics is only an approximation to
reality, a great approximation for most conditions, but an
approximation that breaks down at very high speeds or very small
sizes. Newton’s physics can still help us put a man on the moon,
but fails if that man travels at speeds approaching that of light,
and it fails when we try to describe the motion of an electron.
So what can we do with our theories? We really need some
understanding of cause and effect, so how do we decide which theories
to, at least provisionally, accept? Books have been written on the
problem so we will obviously not get a full answer here, but we can
make a start, and that start will suffice in most cases. There are at
least four characteristics we require:
First, the theory must be in agreement with data. Newton’s theories
met this requirement until new data showed them to apply only under
certain conditions.
Second, theory must in some sense be verifiable. Philosopher of
science Karl Popper said that a theory must be falsifiable. It must
be conceivable that experiment produces a result contrary to the
theory. A classic example is that some might claim that there are
invisible, undetectable elephants in the room. If they are
undetectable, the theory cannot be falsified. A more important
example is the claim that socialism is the best economic system, if
only the right leaders are in charge. That theory cannot be falsified
because any contrary data is explained away with the claim that it
wasn’t real socialism because the wrong people were in charge. Thus
the idea of science supporting socialism fails Popper’s test, those
claims are not science.
Third, though not a strict requirement, it is useful if the theory
predicts something verifiable but previously unknown and not
consistent with competing theories. Einstein’s general theory of
relativity did this with its prediction of how gravity affects light,
a prediction since verified.
Fourth, again not a strict requirement, but we want the theory to be
as simple as feasible. This is known as Occam’s Razor, we prefer
the simplest theory that fits the facts. That is a useful rule to
pick which theory we will find most easily used, but of course does
not preclude some more complicated theory working better when more is
known.
With all that, we have some indication of which theories we might
accept, but there is more. We must keep our minds open. A theory may
offend our sensibilities, but Nature cares not at all about what we
want to be true. We may want to think there is no difference between
ethnic groups, but Nature disagrees, at least in the sports world.
Blacks in the U.S. are mostly of West African extraction and they
dominate in the NFL and NBA, but are not known for distance running.
Meanwhile, Kenyans are not famous as sprinters, but they pretty much
own distance events like the Boston Marathon. No that is not
reason to discriminate, all should have equal opportunity to try. It
is, however, reason to expect that different groups will have
different outcomes.
One thing we should not require is that the theory make sense to the
human mind. In fact, any real advance in science tends to sound
weird, sometimes downright crazy. During one conference on quantum
physics, Niels Bohr was quoted as saying, “We here in the back are
agreed that your theory is crazy. The question which divides us is,
is it crazy enough?” Bohr understood that progress means change,
sometimes radical change. (However, that does not mean that change is
progress. Sometimes change can be detrimental. Let us not seek change
for the sake of change.)
If we can find theories that fit the facts, are verifiable, and guide
our actions appropriately, that is about all most of us can hope for.