Members of the modern skeptical movement frequently emphasize their use of the scientific method. Moreover, when discussing the scientific method and whether a given claim is scientific, members of the skeptical movement frequently emphasize falsifiability and experimentation. Both of these tendencies are problematic.
The organized skeptical movement has done a very good job over the last 30 years promoting skepticism and critical thinking. The movement has helped people understand the deep problems with many superstitions and pseudosciences. However, the skeptical movement’s characterizations both of how the movement functions in regard to science and how science functions are inaccurate. Moreover, the movement tends to exaggerate the degree to which falsifiability matters in science.
Let us first examine the claim that the skeptical movement uses the scientific method. No matter how one characterizes science and the scientific method, a central part of the method is experimentation and observation. However, the vast majority of skeptics will not engage in direct experimentation or observation of data nor in fact should they. For example, in an earlier blog post, I discussed
why claims of ghostly interference with electronics were extremely implausible. I didn’t need to engage in any experiments to reach that conclusion. I and the commentators in the thread discussed the matter based on what we reliably knew about the universe (especially how electronics work) and then made logical conclusions based on those results. No part of that process required any use of the scientific method.
If we wanted to investigate the claim in more detail, we might try to do actual experiments. But I can dismiss the claim of ghostly interference with electronics with a high degree of probability without experiments. Similarly, I can dismiss homeopathy without doing experiments or without such experiments having been done by anyone since the theory of homeopathy contradicts basic understanding of how the universe functions. The vast majority of skeptics will never actually use the scientific method, but rather rely on small bits of science and a lot of critical thinking. That’s fine. The movement is doing good work that way. But we need not pretend when talking to people that we’re engaging in science when we aren’t. This is all the more a concern because the emphasis on science distracts from the most important part of skepticism- careful critical thinking.
Skeptics often characterize science poorly. Skeptics frequently emphasize the need for claims to be falsifiable in order for them to be scientific. The philosopher
Karl Popper first proposed that the demarcation between science and non-science is falsifiability- the ability to falsify a claim. Thus, in a classical Popperian framework, claiming that there is an invisible, intangible dragon living in my bathroom is not scientific because the claim is not falsifiable.
However, naïve Popperism is not a good description of science as a whole. While falsifiability is a useful approximation of what is often scientific, there are many problems with it as a description of all forms of science. For example, as pointed out by Quine, people can add defensive hypotheses to defend an underlying hypothesis; it is far from clear when such defensive hypotheses are acceptable and when they are not. In Quine’s formulation, a defensive hypothesis is a hypothesis that is added to prevent the falsification of another hypothesis. Thus, for example, to return to the case of an invisible dragon in my bathroom, you could try to test for its presence by searching on spectra other than visible light (such as looking at infrared light). When no evidence is found using that method, I might add the defensive hypothesis that the dragon also doesn’t interact with infrared light. As I add more and more hypotheses to counter each experiment, I prevent my claim from being falsified, but, at any given point, the claim that a dragon is in my bathroom remains falsifiable – in theory at least. In portraying a scientific method which relies completely on falsifiability the skeptical movement ignores issues such as those raised by defensive hypotheses.
Unfortunately, the situation becomes even more complicated because sometimes one can add defensive hypotheses and still do good science. For example, consider the history of our understanding of the solar system. In the early 1600s, astronomers adopted Kepler’s model of the solar system in which planets orbit around the sun in ellipses. Subsequently this model was refined further by Newton whose mechanics gave orbits nearly identical to those of Kepler but slightly more accurate (essentially if there is a single planet around a star then Newtonian mechanics predicts an orbit that is a perfect ellipse. But in fact, gravitational attraction between planets makes the orbits slightly non-elliptical). However, with this very accurate, very precise model, new issues arose. The orbit of the planet Uranus seemed to be slightly off from what it should be. Thus,
Alexis Bouvard posited the existence of an as yet unobserved planet, which later became discovered and named Neptune. Bouvard’s hypothetical planet was a defensive hypothesis built to defend the more cherished hypothesis of Newtonian mechanics against being falsified by experimental data. Bouvard’s defensive hypothesis was good science; the defensive hypotheses about my dragon are not. It is not clear how one can easily distinguish between the good defensive hypotheses and the bad ones.
So how do we determine when defensive hypotheses are acceptable and when they are not?
Imre Lakatos suggested that we should look to whether a hypotheses is fruitful: hypotheses that lead to interesting predictions and new questions are fruitful; hypotheses that required unproductive defensive hypotheses should be rejected. While I am partial to Lakatos’s viewpoint, such an approach renders the line between science and non-science inherently subjective.
In addition to the problem of defensive hypotheses, there are many other problems with a falsifiability as the sole line between science and non-science. Naïve Popperism is insufficient to describe the borders of science. Skeptics need both a clearer understanding of the scientific method and a clearer understanding of how the skeptical approach relies on science