PSEUDO-SCIENCE IS NOT THE SAME AS NON-SCIENCE

Popper’s idea (pseudo-science is untestable) is a marvel of simplicity. It explains why no matter how bad a pseudo-scientific doctrine is, its pro­ponent may regularly win debates.

According to Whewell, scientific theories must also have withstood test. Consequently, he viewed as pseudo-scientific those theories which falsely claim to have withstood test. This leaves unclassified those theories which are testable but have been obviously refuted. As Whewell con­sidered these to be neither scientific nor metaphysical, he confusedly im­plied that they are pseudo-scientific, especially when they are submitted to recurrent readjustment and retest. According to Popper such theories are scientific, for he only demands testability; according to Whewell they could not be considered scientific, and so he held them in contempt. He knew that Newton’s optics had been falsely held to have been verified. Yet he did not see that as long as verification was considered a hallmark of respectability, the immense respect for Newton gave these false claims an immense appeal.

Traditionally, a variety of characteristics have been attributed to sci­ence. Popper accepts some of these attributes, such as high explanatory power, high informative content, abstractness, generality, precision, and simplicity; he rejects others, such as obviousness and verifiability. He seems to have claimed in his Logic of Scientific Discovery that the charac­teristics in the first group are all reducible to one, to testability. This is his justification for requiring only this one characteristic of a theory before labelling it ‘scientific.’ I have little doubt that Popper will fully agree that the spurious simplicity of some monistic doctrines (such as Marxism or mechanism) rather than their spurious explanatory power has deluded some people into regarding them as scientific. Simplicity, however, is traditionally viewed (since Leibniz) as the paucity of assumptions relative to the amount of factual information they explain, so that there is no need to differentiate between simplicity and high explanatory power for the purpose of demarcation. And Popper would say the same concerning ex­planatory power, which, in his opinion, increases with refutability. For my part, I consider that the various characteristics of science are less often dependent on each other than Popper suggests. But I still side with Popper in viewing spurious refutability, rather than, say, spurious simplicity, as the chief characteristic of pseudo-science, and for two reasons. First, what­ever else may characterize a scientific theory, the very acceptance of the proposal that scientific theories are agenda to be tested renders Popper’s proposal to check whether a doctrine in question is testable or only spuri­ously testable a matter of supreme practical importance. Second, the claim of pseudo-science is the claim for empirical character. And empir­ical character is nothing else but empirical refutability, as I shall soon explain. Thus, Popper’s demarcation between science and pseudo-science does not require any amendment even on the assumption that he has erred in correlating the various characteristics of science. As to his characteriza­tion of science as such, it requires a reformulation if, as I think, his way of correlating the various characteristics of science is in error. I think we have to characterize scientific theories not only by their refutability, but also by their simplicity, high explanatory power, etc. This has an imme­diate bearing on the problem of selection of scientific problems and of scientific theories which is the topic of the present chapter. According to Popper we always look for the most easily refutable theory. In my opinion this is not the case.

V.