The Scientific Method

The scientific method is the systematic testing of coherent texts against publicly accessible experience within a community of research.

Induction is generalizing from specific cases. For example, we observe that all the swans we have seen are white and then conclude that all swans are white. Since there is always the possibility of encountering a black swan in the future, induction is inherently tentative.

Deduction is deriving new statements from existing statements using rules of inference. For example, if p implies q and p is true, then we can conclude that q is true. Deduction is more definite than induction because if the initial statements are true and the rules of inference are valid, then the deduction can be proven correct. But if the initial statements are in doubt, then we cannot be sure of our conclusions, and in a purely deductive system you must eventually start from axioms or assumptions, statements that you just assume to be true for the purposes of defining the deductive system. Another potential problem is that the rules of inference may not apply. For example, most rules of inference are based on two-valued logic, where every statement is either true or false. There are, however, other types of logic such as three-valued logic where each statement is true, false, or unknown or fuzzy logic where the truth values can range across a continuum from zero to one. So deduction is very useful for organizing our thoughts, but it does not in itself allow us to find truths about the world of experience.

One description that has been suggested for the scientific method is that it follows the inductive-deductive method. In this method you make observations, make a generalization from the observations to form a hypothesis, deduce from the hypothesis predictions of observations, and then compare the predictions to actual observations. If the predictions match, your consider your hypothesis to be supported. If they do not match, then you need to revise your hypothesis.

One of the problems with this definition is that it does not match much of scientific practice. Scientists take much more degrees of freedom in coming up with their hypotheses and theories, and in many cases, they are not just simple inductive generalizations.

Another description that has wider support is the falsification method. With this method it doesn't matter how the hypothesis is formed. Any creative means can be used. Induction may be used, but the scientists are certainly not limited to induction. Once the theory is created, the scientist attempts to falsify the theory through experiments or through showing that it contradicts other accepted theories. If the theory is falsified, then it needs to be adjusted or the related theories need to be adjusted until it is not longer falsified. If the theory is not falsified, then it can be provisionally accepted as a good theory.

Under this view the production of scientific texts is a creative activity drawing on observations, imagination, deductions or associations from other theories, or any other means. The theory then must be tested for coherence within itself and with other theories and with publicly accessible experience. As long as it passes the test, it can be tentatively accepted. If it fails the test, then something has to give. There is a lot of freedom, at least in principle, about what has to give. The whole body of scientific texts needs to be adjusted until it passes the tests or the tests themselves need to be invalidated.

This seems pretty good as far as it goes, but it leaves out a big factor, the community of scientists. Science is not done in isolation. It is done in the context of a community of research. New theories are proposed and new experimental results are reported to a community of other scientists, usually through peer reviewed journals or at conferences. In this process, there are gate keepers who decide what papers will be published or presented. These gate keepers are the reviewers, the editors of journals, and the organizers of conferences.

The broader community of scientists is further divided into various specialist communities. Within each community there are norms of research, what vocabulary should be used and what formal or experimental methods are appropriate. Innovations are possible, but the community must be convinced. Within this community there is competition for funding and recognition, hierarchies, authority figures, and the whole range of social interactions. Still, the central norm is to find theories that work, that allow us to better predict and control a certain region of experience.

Given all of this, it is hard to say that there is one scientific method. There are some general social norms of the research community, but the methods used are wide and varied.

Ultimately, though, the prestige and position of science in society as a whole depends on producing results, so this all works pretty well. However, there is no guarantee that some useful theory might not have been overlooked. As a matter of fact, there is a widely noted phenomenon called a paradigm shift. Scientists can be very reluctant to make radical changes to a successful theory. They would rather patch it up, make small adjustments. At times though, the patches may get to be too much and there may be too many open issues. Sometimes a theory, such as relativity theory, comes along that causes a major shift in theory and future research. The more plodding approach is called normal science and the big shift is called revolutionary science. How a particular line of research will progress depends on many factors, not the least of which are the creativity of the scientists and the complex social dynamics of the research community.

Science is divided into various specialties, each following its own methods and research programs. Typically scientists are not concerned with combining all of these results into one uniform picture. Also, scientists may or may not be concerned with interpreting their results for the general community.

From time to time though a scientist or a science writer may try to draw some general conclusions from science or form a philosophy of life based on science. This philosophizing needs to be distinguished from science itself. It is a legitimate intellectual activity, but it is not science.

One common example is the philosophy of scientific realism. This philosophy claims that the theoretical entities of science actually exist. Furthermore, it claims that these are all that exists. What we really have in science are scientific texts and experiences used to test those texts. Many fundamental scientific objects are not experienced directly. Their existence is included in theories because it helps in prediction and control. This is not to say that some of these things do not exist. For example, atoms probably do exist, but they are not necessarily totally described by atomic theory and in any case we do not experience them directly. We experience other things such as tracks in bubble chambers that help to confirm theories about their existence.

It could be that many scientists routinely assume the realist hypothesis in their work. This is how they think of it and it helps them come up with new theories or to test existing theories. But this realism is really their personal belief. It is not needed for the theory itself.

Some believers in scientific realism will associate the success of science with their philosophy and assume that this success proves scientific realism, but scientific realism is just a metaphysical theory. There are also scientific idealists who correspond scientific concepts to an ideal world of forms and say that the success of science is related to how close we get to the forms in our theories. There are other possible metaphysical theories of science. These theories cannot be tested scientifically and are not necessary for science, but they don't necessarily hurt it either.

This point is important because much of the conflict between science and other parts of human culture could really be conflicts with one of these interpretations of science and not with the practice of science itself.

In actual scientific papers scientists are much more careful. There is an effort in the scientific community to keep the actual science well grounded, but scientists are people too and they also operate in social situations outside of the scientific community. And they have their own inner lives and may, like others, engage in a little culture criticism of their scientific artifacts. It is only natural. But they should really be careful not to use the prestige and rigor of science to lend reflective glory to their private and often untestable personal theories.

This being said, it is still important for people to realize that theories that contradict well established scientific theories have a real problem. Science is very careful to test its theories, so if you advocate theories that are directly contradicted by scientific findings, it is difficult to claim truth for them. This does not mean that if you have a theory in areas that are not covered by science that you have a problem, just if science does cover that area. Science restricts it scope to get the most reliable theories. It really has nothing to say about things outside of its scope. Theories outside of the scope of science are probably less testable than scientific theories. If they could be thoroughly tested by publicly accessible experience, they would be a part of science.

There is a temptation on the part of scientifically minded people to dismiss other types of theories just because they do not fall into the realm of science. These other theories may need more caveats and may be less reliable than scientific theories, but that does mean that whole realms of human experience should be just dismissed out of hand.