[I know this has been done to death, but I thought I’d write up my own take on it anyway.]
The most basic difference between science and religion, the one that makes them butt heads and complicates any discussion of the two, is their difference in epistemology.
Modern science starts with the assumption that we know nothing, and painstakingly have to establish every assertion as true. There have been studies done to confirm and quantify things that everyone takes for granted, such as that something that is farther away looks smaller than something that is closer. Scientists confirmed that this is true, and more importantly were able to describe the equivalence between distance and size, which has implications for eyewitness testimony. Ideally, research is conducted to confirm or deny and/or measure a specific point. Negative results can be just as informative as positive results. It is a flawed, human process, but one that is self-correcting and that builds on itself. It starts from nothing and, by collecting, investigating, and quantifying data, reaches conclusions. These conclusions are then tested by other scientists to see if they are accurate, and are modified in response to new data. Science works forward, the evidence pointing towards the conclusions.
Religion works in exactly the opposite way. It takes a given premise as a fact, and then looks for evidence to confirm the premise. Most importantly, disconfirming evidence is either devalued or ignored. Since the premise is known to be true, any evidence that shows it is false is de facto flawed. Religion works backwards, with the conclusions determining which evidence is correct.
This is a feature of not just religion but of many non-scientific beliefs, including pseudo-sciences. Disturbingly, I’ve found such belief systems in my own profession. Psychology is one of the soft sciences, and in some places it’s so soft it’s fluffy. Classic Freudian psychoanalysis, which has entered popular culture as the definitive form of psychology, is an example of this. Freud never performed research, and based all of his conclusions on a very limited number of case studies from his own practice. He assumed that he and his patients, all of whom were upper-middle-class Viennese, were representative of humanity. He came up with a lot of clever psychological mechanisms to explain their various disorders. When researchers tried to confirm his theories, they found that he had been right about a surprising number of things, but that he had mostly been wrong. Psychology changed from focusing on the subconscious to focusing on observable behavior and to describing cognitive processes. Yet it took decades for traditional psychoanalysis to be abandoned, and it continues to be used today in a modified form.
Another example I came across is William Glasser’s Choice Theory. I first heard of it in a counseling class in grad school, and I thought that it sounded very interesting. But when I got his books from the library and started reading them, I was dismayed to find that they were mostly conjecture. Worse, he claimed that if only everyone would ascribe to, study, and understand his theory, the world would become a Utopia. It sounded like he was trying to start a religion. I still think that the basic premise of his theory, that everything we do is a choice, can be a useful therapeutic tool. Many people feel that their problems are caused by forces they can’t control, and so rationalize not doing anything about them. I think that Impressing on them that their life is a result of their choices can be a powerful motivator for constructive changes. But I don’t think it is anything more than that.
Both classical Freudians and Glasser operate the same way religions do. They assume that a premise is true – neurosis are caused by subconscious pressures, everyone’s problems could be solved if only they would understand that it is their choices that cause them – and pick their evidence to conform to the premise.
On the other hand, there is research that makes me proud of my field, like forensic psychologists quantifying the size an object appears at a given distance, or findings in social psychology like the classic Milgram experiment. These studies produce evidence that can then be used to form theories. Sometimes they are wrong, but the faulty data is eventually weeded out and the theories are stronger for it.
I think that in religion, in science, in every field, we should try to move from evidence to conclusion rather than from conclusion to confirmation. It is the best way we have of knowing what is.