4 Evolution and the Fossil Record
In that regard, the scientific method is similar to many other human endeavors, such as
mythology and folk medicine, which observe something and try to come up with a story
for it. But the big difference is that scientists must then test their hypotheses. They must try
to find some additional observations or experiments that shoot their idea down (falsify it)
or support it (corroborate it). If the observations falsify the hypothesis, then scientists must
start over again with a new hypothesis or recheck their observations and make sure that the
falsification is correct. If the observations are consistent with the hypothesis, then it is cor-
roborated, but it is not proven true. Instead, the scientific community must continue to keep
looking for more observations to test the hypothesis further (fig. 1.1).
This is where the public most misunderstands the scientific method. As many philoso-
phers of science (such as Karl Popper) have shown, this cycle of setting up, testing, and
falsifying hypotheses is unending. Scientific hypotheses must always be tentative and subject
to further testing and can never be regarded as finally true or proven. Science is not about
finding final truth, only about testing and refining better and better hypotheses so these
hypotheses approach what we think is true about the world. Anytime scientists stop testing
and trying to falsify their hypotheses, they also stop doing science.
One of the reasons for this is the nature of testing hypotheses. Lots of people think that
science is purely inductive, making observation after observation until some general scientific
law can be inferred. It is true that scientists must start with observations, but they do not
arrive at scientific principles from induction. As Charles Darwin himself put it in 1861,
About thirty years ago there was much talk that geologists ought only to observe and
not theorize; and I well remember someone saying that at this rate a man might as
well go into a gravel-pit and count the pebbles and describe the colours. How odd it
is that anyone should not see that all observation must be for or against some view if
it is to be of any service! (1903, 1:195)Darwin correctly points out that all useful observations must be made in the framework
of a hypothesis. What data are needed to test the hypothesis? How will they be useful in
falsifying or corroborating it? Instead of inducing general principles of nature, most science
is about deductive reasoning: we set up (deduce) a hypothesis, then try to test it. Philosophers
use the cumbersome term “hypothetico-deductive method” to describe this process, but it is
simple when you think about it.
The difference between inductive generalization and deductive hypothesis testing is
easy to illustrate. Suppose we make the inductive statement that “all swans are white.” We
could observe thousands of swans for many years but never prove that statement true. All
it takes is one nonwhite swan and we can easily falsify this hypothesis. Indeed, there are
black swans (fig. 1.2) in Australia and elsewhere, so the statement has been falsified. As Karl
Popper pointed out, there is an asymmetry between verification and falsification. It is easy to
falsify something; all we need is one well-supported observation that proves the hypothesis
wrong. But we can never prove something true (verify it). Additional corroborating observa-
tions may support the hypothesis but never finally prove it true. As Popper put it in the title
of a book, science is about conjectures and refutations.
Most people think that science is about finding the final truth about the world and are
surprised to find that science never proves something finally true. But that’s the way the
scientific method works, as philosophers of science have long ago demonstrated about the