It might be said in conversation someone had ‘a theory’ about something: why they weren’t feeling very well for example. However, to a scientist this would not constitute ‘a theory’, simply a ‘hypothesis’. So you could hypothesise an illness was caused by something you ate. This is a hypothesis because it makes a testable statement about reality. It is either the case that dinner was to blame or it wasn’t. If several other people ate the same then that becomes the test of the hypothesis. Are they ill or not? The hypothesis can be confirmed or rejected with this information.

A theory, however, requires more. It must be a descriptive and predictive model of the phenomenon. A theory could be that the problem with the food was that it was insufficiently cooked and hence contained still living E. coli bacteria. This would now be a description of the phenomenon. Next the scientific theory can be tested; this is to say that it is possible to attempt to falsify it. For this, volunteers could be provided with both cooked and uncooked food it could be observed whether any become ill. It would also be possible to test for E. coli in both portions. It may be found that the uncooked food contains living E. coli and the people who eat it become ill. Further, it may also be found that the well-cooked food contains dead E. coli and the people eating it do not become ill. In this case the data has not falsified the theory and as such it stands.

That’s basically how science is done. The scientific method can be roughly summarised like this: initially data is collected and then a model is sought which can describe these observations. Next predictions are made using that model and attempts to falsify the theory are made by testing these predictions. When it is known where a model is flawed it can be improved, increasing its descriptive power. As a result the theory can once again be modified to accommodate this new finding, and then tested again. In this way a scientific theory is open, in the sense that it can be improved and updated as more information is gathered. The process of testing the theory lets us do this.

Darwin’s theory of evolution by natural selection was a theory when it was proposed. And it still is. But it is also now considered by the vast majority of scientists to be a fact. So how does a theory become a fact? When a theory is supported by overwhelming evidence it can be accepted as a fact. Darwin’s ideas have been tested for 150 years now, by rigorous and originally very sceptical scientists. In the last 150 years no evidence has been found which refutes the theory. Instead evidence supporting the predictions of evolution has steadily accumulated.

The fossil record, once seen as one of the weakest sources of evidence for evolution, is now rich with examples of species changing over time, and whole clades of distantly related organisms can be linked by transitional fossils. We have many examples too of species evolving in real time, peppered moths, finches and fish for example. Recently the DNA record has revealed whole new depths of evidence. Scientists have also broken down the theory and tested it’s various parts; variation in species, inheritance through generations, competition between individuals for example are all well supported by evidence. When faced with such overwhelming evidence, evolution is as good a fact as you can find.

Intelligent Design is the latest in a string of anti-evolution movements. Originally things were kept nice and clear, anti-evolutionists were creationists favouring religious accounts of the history of life over scientific ones. Creationism was rejected as a scientific idea within Darwin’s own lifetime. In America, creationists tried to prevent the teaching of evolution in schools. When they failed they repackaged their ideas as “Scientific Creationism” a series of public trials however led to the failure of that movement too. Scientific Creationism was rejected as science because it failed to stand up to the scientific method: evidence such as the fossil record proved the antiquity of life and therefore disproved Scientific Creationism as a scientific theory. In addition, American courts also rejected Scientific Creationism from being included in science teaching because its primary effect was to advance Christian religion and had no secular purpose, as such it is unconstitutional.

To their credit, anti-evolutionists are, if nothing else, persistent. Out of the ashes of Scientific Creationism rose Intelligent Design. ID proposes that the idea of evolution is flawed and that the design and creation of life by an intelligent being (but not necessarily ‘God’ – they’re careful not to make that mistake twice) is a better way to understand the natural world.

With so much evidence supporting evolution, can ID hold up as an alternative? In science new theories arise all the time, often because new evidence has come to light. Has some new piece of the puzzle been found that means evolution doesn’t fit the evidence as well as it used to? No, the accumulation of evidence supporting evolution continues unabated.

Another scenario for the promotion of a new theory is when a rival theory can explain existing data just as well. Does ID do this? In the leading ID text book Of Pandas and People, the authors claim that things like the sudden appearance of animals during the Cambrian explosion, and the lack of transitional fossils, suggest evolution is not supported by the fossil record. Yet in all cases the authors do not present the full amount of data, they distort the evidence to fit their “theory”.

ID also claims some properties of living things are “irreducibly complex”, they could not have evolved, so must have been designed. Yet we know of many instances where evolution has produced systems or traits that are incredibly complex. When ID claims a particular trait is irreducibly complex, science can test this claim. We discuss two examples in a case study. In both cases the ID claims are falsified. The claim that these systems are too complex to evolve is shown to be untrue. In general however the claim that any system could be too complex to understand is not testable. To test this claim we would have to understand everything perfectly, we would in effect, half to travel back in time a watch the history of life unfurl!

This misrepresentation and cherry-picking of data is unscientific. The falsifiable claims of irreducible complexity of particular systems show that it is not a successful explanation. In addition ID fails to provide a testable method by which the ‘designer’ carries out its work. Evolution has natural selection, ID has… well, nothing. Without a proposed theory based on natural laws, which scientists could test against evidence, ID cannot be a science.

Science works by making observations, gathering facts, drawing a theory or model to explain those observations and facts and then testing that model. The model can be improved and added to if the evidence suggests it needs to be. Evolution by natural selection is a great example of a scientific theory which has been tested to the limits, and not found wanting. Intelligent Design does not fit this description so should not be considered, or taught, as a science.

Written by Andrew Maddox

The Blind Watchmaker
by Richard Dawkins, Penguin: 1986

Climbing Mount Improbable
by Richard Dawkins, Penguin: 1996

Living with Darwin: Evolution, Design, and the Future of Faith
by Philip Kitcher, OUP: 2007

Unweaving the Rainbow
by Richard Dawkins, Penguin: 1998