May 2, 2007

Science and Falsification

In a recent issue of eskeptic (, Robert Ehrlich contributed an interesting piece titled "Science Will Never Explain Everything: That is Why it is So Useful." In the article, Ehrlich wrote:

Science is powerful because it often provides explanations before
observations are made. As Yogi Berra once remarked, “prediction is hard —
especially about the future.” The nature of scientific theories is to make
surprising predictions — the more surprising, the more confident we can be
in the theory should the prediction be fulfilled.

On occasion religious figures also make predictions, most notably about the
end of the world, but I am unfamiliar with any example where the failure of the
world to end on schedule caused a reassessment of the religious leaders in their
fundamental beliefs. Instead, the holy man and the faithful sometimes make some
recalculations, and come up with a new date for the end of the world, or else
give up trying with their faith unshaken — for that is the nature of faith which
requires no evidence to justify it, yet somehow paradoxically it craves
confirming evidence when it can get it.

In this quote, Ehrlich is refers the idea that scientific theories usually have implications which can formulated into predictions about potential empirical observations. The idea, as Ehrlich envisions it, goes something like this: Imagine a scientific theory T. Scientists think hard about T and figure out that if T were actually true, then we ought to be able to see or observe in the world some other phenomenon O. In other words, theory T predicts observation O. So, the scientists go out into the world to look for O. If they find O, then that observation provides some confirming evidence for T. If they don't find O--or, more seriously, if they discover that not-O (i.e., O is false), then theory T is thereby falsified. And the scientist, being the rational creature that he is, will no longer entertain T as a scientific theory.

Because this is how Ehrlich characterizes science, he thinks that science is eminently more rational and useful than religion and religious people. Religious beliefs, he thinks, tend to be unfalsifiable--primarily, it seems, because religious people won't give up their beliefs when they encounter falsifying evidence. This is Ehrlich's point when he talks about the fundamentalist preachers who predict the end of the world. They predict that Armageddon will occur on a certain date, but when the prediction fails to come true, they find some way to go right on believing, simply making adjustments to their theories to cover up the unpleasant failed prediction.

Now I agree with Ehrlich that some Christians (and cultists) may hold on to implausible beliefs in the face of significant falsifying evidence. What I take issue with is his implicit assumption that scientists are immune from this kind of thing. In fact, Ehrlich (at least as far as this article would suggest) holds a pretty naive view of science and how it works. It is true that scientists make predictions from their theories and seek to test them in the way Ehrlich suggests. But, Ehrlich is mistaken in thinking that falsification works in the simple, straightforward way that he describes. Contemporary philosophers of science will tell you that it is virtually impossible to decisively falsify any but the simplest scientific theories. Why? Because--just like those fundamentalist preachers--a scientific theory can be adjusted and modified almost endlessly in minor and major ways to account for, incorporate, or avoid potentially falsifying evidence. And scientists do this kind of thing all the time!

I point this out not to call the rationality of science into question. I believe that it would be a mistake to let a plausible and promising scientific theory be rejected just because an initial prediction of the theory failed to materialize. If a small adjustment in some peripheral details of a theory can be made that leaves the major aspects of the theory intact, but also has the fortuitous consequence that the failed prediction is no longer a prediction of the theory at all (and thus no longer potentially falsifying), then by all means the scientist should make the adjustment and continue their research program. Such a practice is eminently rational. It is only when "anomalies" increase greatly and the adjustments to the theory become more and more ad hoc that scientists should abandon the theory. But, if this is rational for scientists to do, then one cannot automatically object when religious people do similar things. (I would argue that the problem with the Armageddon predictors is just that their adjustments are ad hoc.)

Let me end, however, with this observation. It seems to me that there is one scientific theory that, if we follow Ehrlich's view of science, ought to be discarded immediately. I speak of Darwinian evolution. Darwin's theory of evolution based on natural selection predicts that there ought to be millions and millions of transitional forms in the fossil record. However, to date, not a single undisputed example of a transitional form has been found--despite many decades of careful searching. On Ehrlich's criteria, the theory of evolution has been falsified. Will Erhlich follow his own advice and now repudiate the theory of evolution?

No comments: