Scientific Revolutions

Of course, scientists often put forward conflicting theories, so we need some way to test which theory is correct. One simple method is to use the theory to make predictions. Since an explanation depends on principles that are general, these principles have implications beyond the particular phe- nomenon that they were originally intended to explain. The theory thus pre- dicts what will happen in circumstances that the scientist has not yet observed. We can then test the theory by seeing whether these predictions hold true. For example, we can make spheres out of a wide variety of mate- rials, calculate their densities, and then see which ones float. If any sphere floats that is denser than water, then we have to give up our principle that a sphere floats in water only if it is less dense than water. If we find a sphere that is less dense than water but does not float, then we have to give up the principle that a sphere floats if it is less dense than water.

These methods help us rule out certain scientific principles, but the fact that a principle implies true predictions does not, by itself, prove that the principle is true. That argument would commit something like the fallacy of affirming the consequent. Nonetheless, we can still say that a theory is confirmed if it yields true predictions, and it is confirmed more strongly if it yields more, more varied, and more unexpected true predictions.

Scientific method is actually much more complex than this simple exam- ple suggests. This becomes apparent when we encounter anomalies. Suppose we have confirmed and explained the principle that a sphere floats in water if and only if it is less dense than water. Suppose also that another principle is well confirmed: A substance gets smaller and denser as it gets colder. Taken together, these principles predict that a sphere of ice should sink in water. Ice is colder than water, so, according to the second principle, ice should be denser than water, and that, given the first principle, means that it should not float in water. Of course, our prediction is wrong, since spheres of ice do float in water. What do we do now? The obvious solution is to modify the principle that a substance gets smaller and denser as it gets colder. This holds for most substances, but not for water. Water expands and thus gets less dense as it freezes.

We could have tried another solution. We could have denied the other principle: that a sphere floats in water if and only if it is less dense than wa- ter. Why do scientists not do this? One reason is that we have independent evidence that water expands when it freezes. That is why jars of water burst when they are left in a freezer. Another reason is that we could not give up this principle alone, because it follows from more basic principles about grav- ity and the mutual repulsion of molecules. Thus, many other areas of science would be affected if we gave up the principle that a sphere floats in water if and only if it is less dense than water. Scientific theories work together, so we cannot throw out one without undermining the others. That all these other scientific theories are not only well confirmed but useful is what makes sci- entists give up one principle rather than another when an anomaly arises. At this point we might seek an even deeper explanation and ask why wa- ter expands when it freezes. In fact, to this day, nobody seems to have a fully adequate explanation of this common phenomenon. There are various theo- ries but no agreement about how to explain the expansion of water. It has something to do with the way in which liquid water crystallizes to form ice, but nobody is quite sure why water crystallizes in that way. Does this show that certain phenomena are beyond scientific understanding? Probably not. But it does suggest that science may never be complete. More questions arise as science progresses, and there will always be questions that remain unan- swered. As scientists discover and explain more and more phenomena and see connections among principles in different areas, every new step gives rise to more questions that need to be answered. That is one way in which science makes progress.