The Science of Discworld IV

the same reason. So how can we possibly know how hot it is at the centre, when no person or instrument can be sent there to find out?
    We know such things because science is not limited to just
observing
the world. If it were, it would be firmly back in the human-centredrealm. Its power derives from the possibility of thinking about the world, as well as experiencing it. The main tool of science is logical inference:
deducing
features of the world from a combination of observation, experiment and theory. Mathematics has long played a key role here, being the best tool we currently have for making quantitative inferences.
    Most of us understand in broad terms what an observation is: you take a look at things, you measure some numbers. Theories are trickier. Confusingly, the word ‘theory’ has two distinct meanings. One is ‘an idea about the world that has been proposed, but has not yet been tested sufficiently for us to have much confidence that it is valid’. A lot of science consists of proposing theories in this sense, and then testing them over and over again in as many ways as possible. The other meaning is ‘an extensive, interconnected body of ideas that have survived countless independent attempts at disproof’. These are the theories that inform the scientific worldview. Anyone who tries to convince you that evolution is ‘only a theory’ is confusing the second use with the first, either through intention to mislead, or ignorance.
    There is a fancy word for the first meaning: ‘hypothesis’. Few people actually use this because it always sounds pedantic, although ‘hypothetical’ is familiar enough. The closest ordinary word to the second meaning is ‘fact’, but this has an air of finality that is at odds with how science works. In science, facts are always provisional. However, well-established facts – well-developed and well-supported theories – are not
very
provisional. It takes a lot of evidence to change them, and often a change is only a slight modification.
    Occasionally, however, there may be a genuine revolution, such as relativity or quantum theory. Even then, the previous theories often survive in a suitable domain, where they remain accurate and effective. NASA mostly uses Newton’s dynamics and his theory of gravity to compute the trajectories of spacecraft, not Einstein’s. An exception is the GPS system of navigational satellites, which has to take relativistic dynamics into account to compute accurate positions.
    Science is almost unique among human ways of thinking in not only permitting this kind of revisionism, but actively
encouraging
it. Science is consciously and deliberately universe-centred. That is what the ‘scientific method’ is about. It is like that because the pioneers of science understood the tricks that the human mind uses to convince itself that what it
wants
to be true
is
true – and took steps to combat them, rather than promoting them or exploiting them.
    There is a common misconception of the scientific method, in which it is argued that there is no such thing because specific scientists stuck to their guns despite apparent contrary evidence. So science is just another belief system, right?
    Not entirely. The mistake is to focus on the conservatism and arrogance of individuals, who often fail to conform to the scientific ideal. When they turn out to have been right all along, we hail them as maverick geniuses; when they don’t, we forget their views and move on. And that’s how the real scientific method works. All the other scientists keep the individuals in check.
    The beauty of this set-up is that it would work even if
no
individual operated according to the ideal model of dispassionate science. Each scientist could have personal biases – indeed, it seems likely that they do – and the scientific process would still follow a universe-centred trajectory. When a scientist proposes a new theory, a new idea, other scientists seldom rush to congratulate him or her for such a wonderful thought. Instead, they try very hard to shoot it down. Usually, the scientist proposing the idea has already done the same thing. It’s much better to catch the flaw yourself, before publication, than to risk public humiliation when someone else notices it.
    In short, you can be objective about what everyone
else
is doing, even if you are subjective about your own work. So it is not the actions of particular individuals that produce something close to the