STUMBLING INTO AN INSIGHT:

According to the logical empiricists of the Vienna Circle, a theory is scientific if it is verified by empirical data. For Popper, this condition was grossly insufficient.

All demarcation criteria are designed to exclude something. What Popper really wanted to do was to show why psychoanalysis and Marxism were not scientific. Those theories had been widely understood as "scientific" in his Viennese milieu because of a logical empiricist theory called verificationism. According to this view, a theory is scientific if it is verified by empirical data.

For Popper, this condition was grossly insufficient. There was plenty of data that apparently confirmed psychoanalysis, he claimed. Every piece of data about personalities might be another brick in the confirmatory edifice for Freud, just as every event in politics or economics seemingly further confirmed Marxist theories such as the centrality of class conflict in history or the surplus value of labor. What this meant for Popper is that logical empiricists were looking at things the wrong way around. The issue was not whether a theory was confirmed--anything might be interpreted as confirming if you formulated the theory flexibly enough. The point was whether it was possible to falsify the theory. Was there any imaginable observation such that, should it be found, Freudians or Marxists would concede that their theories were false? If the answer was no, these were not sciences.

The appeal of falsificationism is obvious. It provides a bright line, and it rewards the boldness that we often like to see exemplified in science. How well does it work?

The short answer is: not very. Philosophers of science recognized this almost immediately, for two main reasons. First, it is difficult to determine whether you have actually falsified a theory. This is largely a restatement of one of Popper's own objections to verificationism. How do you determine that an observation actually constitutes a confirmation of a theory? Well, you interpret it within its framework, and sometimes those interpretations produce the lamentable distortions that Popper decried. But the same holds true for falsifying a theory, too. Suppose you did an experiment in your laboratory to test a theory, which predicts that under certain conditions your fact-o-meter should register a value of 32.8, and you got a result of 5.63. What do you do? Should you run to the journals and proclaim the death of that theory?

For Popper, the logical empiricists were looking at things the wrong way around. The issue was not whether a theory could be confirmed, but whether it could be falsified.

Not so fast. How do you know that your experimental result was accurate? Maybe the reason you did not get the value of 32.8 is that your fact-o-meter malfunctioned, or perhaps you did not perform the experiment under precisely the right conditions. In short, it is rare to have a thumbs-up/thumbs-down result as in the 1919 eclipse expedition. (As a matter of fact, the results of that expedition were more equivocal than Eddington made them seem. It was several years before absolutely incontrovertible results in support of general relativity were obtained, largely by observatories in California.) If any disconfirming result stood to invalidate a theory, then every tenet of modern science would have already been falsified by middle school science students failing to replicate utterly uncontroversial standard experiments. This is clearly nonsense. While it sounds like a good idea to insist on falsifying observations, it is far from straightforward to determine when precisely this has been done--and that defeats the purpose of having a bright-line standard.

The second problem with Popper's proposal has to do with the actual demarcations it gives us. The very minimum we should expect from a demarcation criterion is that it slices the sciences in the right places. We want our criterion to recognize as scientific those theories that are very generally accepted as hallmarks of contemporary science, such as quantum physics, natural selection, and plate tectonics. At the same time, we want our criterion to rule out doctrines such as astrology and dowsing. Popper's falsifiability standard is not especially helpful in this regard. For starters, it is difficult to present the "historical" natural sciences, such as evolutionary biology, geology, or cosmology--those fields where we cannot "run the tape again" in the laboratory--exclusively in terms of falsifiable claims. Those sciences provide persuasive explanations of nature through the totality of a narrative chain of causal inference rather than a series of empirical yes-no votes. Popper thus inadvertently excludes important domains of contemporary science.

Tweet   @brothersjudd