Does science require beliefs?

I am not talking about a belief in God, or anything whatsoever to do with religion. All I am asking is whether the practice of science requires the holding of beliefs or not. If you ask most scientists, engineers or even random members of the public, you are likely to get a negative answer (for a vigorously atheist answer see here). Science, after all, is about knowledge and repeatable process. The scientific method delivers predictive, explanatory models of the universe, that are derived from, and agree with, hard facts – measured data. We know there is gravity from repeatable experiments and we have models that let us make predictions about how it acts. Therefore, we have no need for beliefs to explain that an object lifted off the ground drops when let go.

That does sounds pretty convincing. Given a law of nature we can explain how the entities and events it refers to interact and we can make predictions about how they will behave under some new, future conditions. This requires no beliefs.

Or does it? If you were to ask Max Planck (yes, the Nobel prize winning author of quantum theory and the guy after whom the Planck constant is named), he’d promptly admonish you as follows:

“We have no right to assume that any physical laws exist, or if they have existed up until now, that they will continue to exist in a similar manner in the future.”

No amount of past data is grounds for expecting the same, previously observed relationships to hold into the future (whether under past or new conditions), for which – by definition – we have no data. Making predictions fundamentally relies on the belief that the laws of Nature are constant and will persist as observed and deduced previously. Now, you might argue that this is a reasonable belief to hold, and I'd agree with you, but you'd be hard pressed not to concede that it is a belief rather than a (scientific) fact. Almost as an aside, there is some evidence though that puts a question mark over the belief that the laws of nature are constant (e.g., see this article in the journal Nature or reports like the one in ABC Science).

The above is just a specific application of the more general problem of induction, whereby we “[p]resuppos[e] that a sequence of events in the future will occur as it always has in the past.” A great, more recent thought experiment to illustrate the problem has been proposed by Goodman in the form of the purpose-built predicate “grue.” “Something is grue if and only if it has been observed to be green before a certain time or blue after that time.” Therefore all emeralds that have ever been observed are not only green, but also grue and we have no basis for assuming that after some future time T we will find green but not grue emeralds. Coming back with saying that emeralds have always been green is beside the point …

As you may have noticed, the above reasoning deliberately took shortcuts and did not explore other instances of belief in science, which may well be rectified in future blog posts (we just can't tell yet).

Finally, it is worth noting that the ideas presented above are in no way an attack on science! Science, on the basis of its underlying beliefs and assumptions, sheds light on how the world around us may work, allows us to make predictions (which for some phenomena have so far always come true), lets us harness the potential of materials around us for the benefit of humanity and dramatically demonstrates the advances that human intelligence is capable of. This makes science greatly valuable and something to be proud of, but let us not delude ourselves into thinking that it is devoid of belief.