Science is great, isn’t it?

Science has helped us control the world in ways few would have thought possible, created new forms of entertainment and new ways communicate with others, helped us live longer.

Science has become the best way to understand the world around us, and even the very nature of ourselves.

But how does science do this? And how does science keep doing this even when our theories may become obsolete?


These are age-old questions, but they became even more important to philosophers in Vienna and Berlin following the First World War. Many believed that Germans and Austrians had went to fight under very dubious banners: for honour, for the emperor, for the greatness of the nation, for a rightful place on the world stage, for glory.

Following the war, people asked themselves what these claims amounted to, and how we could stop such bloodbaths from happening again.


Their answer was to look to science, to meaningful claims about the world, as opposed to:

Metaphysics, or even poetry, which did not make such claims at all, and

Pseudo-science, which only appeared to make meaningful claims about the world.

What made science special was its method, which meant that even if our current theories are false, they would soon be replaced with something true. This position is known as ‘logical empiricism.’


For one logical empiricist, Carl Hempel, the scientific method worked because it made claims based on our observations and experiments with the world, and then used the rules of logic to build these into claims about the world we cannot observe and the laws which rule them.

But we can go further. From the knowledge that science was given us of the world – of observable entities, theoretical entities and laws of nature – we can give explanations and make predictions about the world around us. And all of these can be changed back into observable claims about the world that science can check.


It’s a nice picture, which seems to capture how we think science works logically. But is it true?


One problem has been known since the eighteenth century – the problem of induction – which the philosopher Karl Popper tried to solve.


But another two problems were raised by Carl Hempel himself.

The first is known as the Ravens’ Paradox, and questions whether we can ever make observations of the world without our theories getting muddled in.

Say we want to test the claim

All ravens are black.

But logically, this statement is the same as the claim

All non-black things are not ravens

Since finding a white raven, say, would prove both of them to be incorrect.

But this raises a problem for science: that observing anything that is not black proves the claim we wanted to test in the first place. A herring that is coloured red is proof that all ravens are black.

This seems to go against how we think science works, and yet it must be true.

So does science merely use the evidence of our senses?

What is the red herring here?


The second problem relates to scientific explanations, and questions whether our theories, laws and observations can ever exist in such a structure as Hempel gives it.

If we know enough about the laws of nature, and enough about the objects we want to study, then Hempel says we can explain or even predict an event.

For instance, knowing that metals expand when heated will tell us what will happen if we throw a metal coin into a fire.

But perhaps the example is not so simple. After all, what we’d like to say is that our scientific laws are somehow more fundamental to nature, that they somehow cause events to happen. So instead let’s consider these two examples:

Knowing the length of its shadow will tell us how high a flagpole is;

Knowing what some animals are behaving will tell us if a storm is coming.

Both these explanations are true, and yet it seems daft to claim that a shadow is causing the height of a flag pole or that cows have Thor-like properties.

So is explanation only in the eye of the beholder?


These are just two problems with our common-sense view of how science works. But all of them suggest that science is a much more complicated activity than we give it credit for. That science is not as close to logic as we’d expect.

And for some philosophers, including Thomas Kuhn and Paul Feyerabend, this is a good thing!