Science versus History

I've met a number of atheists over the years who claim they only believe what can be shown to be true by science. The evidence must be peer reviewed. They discount all other evidence. When challenged by the historical evidence for Jesus being better than that of Tiberius Caesar they declare that history is untrustworthy so they don't believe in Jesus and they don't believe in Caesar.

This position is obviously extreme and it represents only a tiny minority of atheists but it does reflect our current cultlure: Science is Top Dog. Science is the best way of determining the truth. Science trumps history.

But I think that this just isn't so. I think Science = History.

Let me explain.

How does science work?

In an ideal world:

Science is based on observations. To explain those observations someone creates a hypothesis. Experiments are then devised that can falsify the hypothesis. If the hypothesis is not falsified then increasing trust is placed in that explanation for the data until we might say we have a scientific theory e.g. of gravity.

Now the essence of good science is that the experiments are repeatable; someone else can repeat the experiment and get the same results. You can go check it for yourself.

So far so good. I can believe the theory is true because I can go out and test it. But here's the catch.

Do you test it?

For gravity I do. I'm always dropping stuff and it always falls to the ground with a reassuring clatter.

But what about theories of particle physics? Cosmology? Evolution?

I bet you don't.

So how do you know it to be true?

Trust.

It's been peer reviewed. Great. This means that some other group of scientists have repeated the experiments and found the result to be true. This might be true. But experiments are costly, require specialist equipment and large amounts of time and money. And some data – like a long running climate study - can never be repeated. So what's more likely is that they checked that the methodology sounded reasonable and that they too took the results on trust. And unless someone needs to replicate the results for their own work any fraud or error is unlikely to be discovered.

But surely a scientist wouldn't commit fraud would they? They're committed to the pursuit of truth...

and...

fame, research funding, getting published, promotion, job security, a pet theory, minimizing effort etc. ...

They're human. And if you not quite sure that's possible then read this and this.

So ultimately when we trust a peer review article we trust that the scientists who wrote and reviewed are reliable. That they are reliable eyewitnesses of the experiments they record. And on the whole that's a reasonable assumption to make.

So that's the first catch – or rather similarity with history. With history we have to trust that ancient writers are accurately reporting the events of their day. But this isn't done naively – the bias of the writer is considered and their purpose for writing is assessed. And when events they describe can be corroborated from other sources  greater confidence is placed in their accounts. On the whole we can decide if a source is a reliable account of an event.

Which leads us to catch number two.

Observations require interpretation. Just as an historian examines the data and forms a theory as to why an event happened, the scientist's hypothesis is ultimately an interpretation of the known observations. And more than one interpretation can fit the known data. So which interpretation are you going to choose? For the individual scientist that depends on what they've been taught, the theories they have emotionally invested in and what's currently fashionable (i.e. the ruling paradigm). None of which may be true. In other words scientist are biased – just like other people. But it doesn't stop there. The scientists who peer review your work are also biased and if they don't like your interpretation – even though it explains all the known observations – your work may never get published. And if you're not sure of that, here are three examples where  publishing against the paradigm is unlikely to suceed : climate change, evolution and the big bang.

Thus I think science and history are similar methods of knowing: both require trusting a source of observations and weighing up the explanatory power of any proposed interpretation.  Ultimately, a scientific paper is an historical record of observations along with an interpretation. Science is not the be all and end all of knowledge as many think. It's not a statement of ultimate truth.

 Science doesn't trump history.