Lessons from Karl Popper
Referring to "Exchange with Mark Notturno on Karl Popper and F.A. Hayek" by Kelley L. Ross:
The Friesian school is obviously very friendly to Karl Popper. I remember thinking when I first learned about falsificationism that there was something very upside-down about a theory of science where progress was achieved by falsification alone. I think this also misses the ways normal science progresses: determination of significant fact, matching of facts with theory, and articulation of theory, as Kuhn recognised. But maybe my understanding of Karl Popper is simply incomplete. In particular, I think that it would here be fruitful to see what Ross, as a Popperian, thinks of Thomas Kuhn, so that we might acheive a useful synthesis of the two approaches.
The top Google-ranked page on the Friesian website for "Kuhn" was "Foundationalism and Hermeneutics" by Kelley L. Ross. It is truly appropriate. Ross regards Kuhn as a hermeneutical theorist of science, and he considers that many of Kuhn's claims are implicitly refuted in Martin J.S. Rudwick's book The Great Devonian Controversy. (University of Chicago Press, 1985.) He also argues that the approach taken by the geologists in the book to the mapping and interpretation of the geological strata in Devonshire in the 1830s was conformable to Karl Popper's view of falsification. This sounds like it would be a good book to read after I finish The Structure of Scientific Revolutions.
However, Ross does adumbrate his basic view of science in the course of the essay, which helps. On the one hand, Kuhn is right to note that there is a cycle of hermeneutics that is present in science. On the other hand, he does not appreciate that the hermeneutical cycles get smaller and smaller the more evidence comes to light. The more observations are made, the greater are the constraints upon the theories that can be produced. Ross does acknowledge that it is possible in principle to construct an infinitude of different theories for any given set of observations. But he rightly notes that in practice, theoreticians have difficulty proposing even one new theory that fits all the observations. "[U]sually exercises of the sort must grotesquely overlook some of the most elementary considerations." Occam's Razor, for example, would preclude most of the theoretical possibilities from being explored as unnecessarily complex. It be easy to construct an inelegant collection of ad hoc propositions to cover the facts, but this would never gain general acceptance.
Is it right to require that a paradigm be falsifiable? There is, after all, a priori no reason to suppose that this must be the case, as according to Kuhn, science progresses just fine without a paradigm's being falsifiable. Supporters of the old paradigms almost never get converted to the new paradigms because they can always devise solutions that are consistent with the old paradigms. That is why the paradigms are incommensurable with each other. Yet something about this claim just doesn't ring true, because if it were impossible to commensurate two paradigms, then why would one paradigm ever get chosen over another? Even the claim that one paradigm admits of fewer anomalies is already a method of commensuration. So it seems like somewhere along the way, Kuhn is going to get involved in a contradiction.
Nevertheless, it is still possible to make a distinction between a hypothesis and a theory in this respect. Hypotheses must always be falsifiable, according to either Kuhn or Popper. If they are not falsifiable, then it would be impossible to construct a meaningful experiment to successfully confirm or disconfirm a hypothesis. But Kuhn's claim is that the rules of confirmation and disconfirmation of scientific hypothesis themselves come from one specific paradigm or another. This paradigm, therefore, is the part that becomes unfalsifiable.
On the other hand, Popper can explain how one paradigm gets chosen over another. Kelley L. Ross notes this in "Founationalism and Hermeneutics". The refutation of Ptolemy by Galileo really came down to one critical observation: The phases and the change in apparent size of Venus were not explained by Ptolemaic astronomy, but they were by Copernicus's theory. He does note that others like Tycho Brahe were able to propose geocentric theories that would take the new observations into account. But these approaches involved ad hoc modifications to the old paradigm. They were either unnecessarily complex, or else they had little plausibility independently of the fact that they happened to solve the old problem. The basic principle of falsificationism still seemed to have been what caused the paradigm shift in the first place. It was not a subjective matter which paradigm got chosen in the end.
This does not mean that I need to drop the vast majority of my previous approach to ethics, however. The Kuhnian analogies were indeed proving most fruitful. However, what I now have is some extra principles drawn from falsificationism that sharpen certain areas of Kuhn's own explanations of the progress of science. For example, an account of Popperian falsification explains why an old paradigm is never returned to. It has indeed been conclusively falsified by the additional observations that have come to light as science progresses.
Nevertheless, it does remain to be seen whether a moral principle can be falsified. Logical Positivists like Stevenson typically assume that it cannot, and therefore is more expressive of emotion than reason. Even if a moral principle can be falsified, it is a separate issue entirely whether it ought to be. Popper himself argued that the meaningfulness of a proposition and its falsifiability were two different things. These are issues that I shall have to explore in detail in the next few sections.
And many who demand logical consistency would deny that scientific theories are or should be falsifiable.
The Friesian school is obviously very friendly to Karl Popper. I remember thinking when I first learned about falsificationism that there was something very upside-down about a theory of science where progress was achieved by falsification alone. I think this also misses the ways normal science progresses: determination of significant fact, matching of facts with theory, and articulation of theory, as Kuhn recognised. But maybe my understanding of Karl Popper is simply incomplete. In particular, I think that it would here be fruitful to see what Ross, as a Popperian, thinks of Thomas Kuhn, so that we might acheive a useful synthesis of the two approaches.
The top Google-ranked page on the Friesian website for "Kuhn" was "Foundationalism and Hermeneutics" by Kelley L. Ross. It is truly appropriate. Ross regards Kuhn as a hermeneutical theorist of science, and he considers that many of Kuhn's claims are implicitly refuted in Martin J.S. Rudwick's book The Great Devonian Controversy. (University of Chicago Press, 1985.) He also argues that the approach taken by the geologists in the book to the mapping and interpretation of the geological strata in Devonshire in the 1830s was conformable to Karl Popper's view of falsification. This sounds like it would be a good book to read after I finish The Structure of Scientific Revolutions.
However, Ross does adumbrate his basic view of science in the course of the essay, which helps. On the one hand, Kuhn is right to note that there is a cycle of hermeneutics that is present in science. On the other hand, he does not appreciate that the hermeneutical cycles get smaller and smaller the more evidence comes to light. The more observations are made, the greater are the constraints upon the theories that can be produced. Ross does acknowledge that it is possible in principle to construct an infinitude of different theories for any given set of observations. But he rightly notes that in practice, theoreticians have difficulty proposing even one new theory that fits all the observations. "[U]sually exercises of the sort must grotesquely overlook some of the most elementary considerations." Occam's Razor, for example, would preclude most of the theoretical possibilities from being explored as unnecessarily complex. It be easy to construct an inelegant collection of ad hoc propositions to cover the facts, but this would never gain general acceptance.
Is it right to require that a paradigm be falsifiable? There is, after all, a priori no reason to suppose that this must be the case, as according to Kuhn, science progresses just fine without a paradigm's being falsifiable. Supporters of the old paradigms almost never get converted to the new paradigms because they can always devise solutions that are consistent with the old paradigms. That is why the paradigms are incommensurable with each other. Yet something about this claim just doesn't ring true, because if it were impossible to commensurate two paradigms, then why would one paradigm ever get chosen over another? Even the claim that one paradigm admits of fewer anomalies is already a method of commensuration. So it seems like somewhere along the way, Kuhn is going to get involved in a contradiction.
Nevertheless, it is still possible to make a distinction between a hypothesis and a theory in this respect. Hypotheses must always be falsifiable, according to either Kuhn or Popper. If they are not falsifiable, then it would be impossible to construct a meaningful experiment to successfully confirm or disconfirm a hypothesis. But Kuhn's claim is that the rules of confirmation and disconfirmation of scientific hypothesis themselves come from one specific paradigm or another. This paradigm, therefore, is the part that becomes unfalsifiable.
On the other hand, Popper can explain how one paradigm gets chosen over another. Kelley L. Ross notes this in "Founationalism and Hermeneutics". The refutation of Ptolemy by Galileo really came down to one critical observation: The phases and the change in apparent size of Venus were not explained by Ptolemaic astronomy, but they were by Copernicus's theory. He does note that others like Tycho Brahe were able to propose geocentric theories that would take the new observations into account. But these approaches involved ad hoc modifications to the old paradigm. They were either unnecessarily complex, or else they had little plausibility independently of the fact that they happened to solve the old problem. The basic principle of falsificationism still seemed to have been what caused the paradigm shift in the first place. It was not a subjective matter which paradigm got chosen in the end.
This does not mean that I need to drop the vast majority of my previous approach to ethics, however. The Kuhnian analogies were indeed proving most fruitful. However, what I now have is some extra principles drawn from falsificationism that sharpen certain areas of Kuhn's own explanations of the progress of science. For example, an account of Popperian falsification explains why an old paradigm is never returned to. It has indeed been conclusively falsified by the additional observations that have come to light as science progresses.
Nevertheless, it does remain to be seen whether a moral principle can be falsified. Logical Positivists like Stevenson typically assume that it cannot, and therefore is more expressive of emotion than reason. Even if a moral principle can be falsified, it is a separate issue entirely whether it ought to be. Popper himself argued that the meaningfulness of a proposition and its falsifiability were two different things. These are issues that I shall have to explore in detail in the next few sections.
Comments