In a Thought Leader piece in the February 2009 issue of Inside Knowledge, Neil Olonoff made a case for the importance of recognizing that all our knowledge is uncertain, that we in Knowledge Management should have no hesitation in admitting uncertainty, and since we “live in a world of uncertainty, we should use that truth to find our way.” (p. 5)
I very much agree with this main point of Neil’s argument, since I’ve been a dyed in the wool fallibilist for many, many, years. However, in arguing for fallibilism, he stated:
”. . . The famed philosopher Karl Popper showed that scientific statements are not truly verifiable. What makes them scientific, in fact, is not their truth or verifiability. It is, paradoxically, their falsifiability.”
However, even though things are sometimes stated this way, this is not what Popper showed, nor is it exactly what he proposed as a demarcation criterion between science and other forms of inquiry. First, he didn’t think that scientific statements could be either verified or falsified as individual statements. Instead, he showed that such statements could only be tested in conjunction with auxiliary statements and assumed background knowledge as part of a system, which is what was falsifiable. Basic statements (assertions describing observations) can conflict with deductions from such a system also purporting to describe those observations, thus creating a logical inconsistency that has to be resolved by deciding which statements in the system being tested will be taken as false (i.e. are “falsified”).
Second, Popper showed that, as a matter of logic, singular pure existential statements such as “There exists at least one white raven,” can be verified, but not falsified. However, he also proposed his falsifiability demarcation criterion for scientific systems, which implied that pure existential statements were part of “metaphysical” systems and not “science.” Since scientists use pure existential statements frequently, this has always been one of the weak points of the falsifiability demarcation criterion.
Third, however, it’s also important in considering these things to recognize what Popper meant when he said that a statement was verifiable, and this will get us back to Neil’s main point about uncertainty. For Popper, neither verifiability nor falsifiability of systems of statements is tantamount to proof in the sense of demonstrating with certainty that a statement, or system of statements, is true or false. Both verification and falsification are matters of what we decide in response to our experience. So, for example, a spatio-temporally specific statement such as “The Ice in the beaker on the table in the laboratory at 223 41st street in New York City, melted at 1:32 PM on April 4, 2009,” can, unlike a universal generalization, be verified or falsified by decisions made in response to singular basic statements that conflict with or corroborate such a statement. However, for Popper such verifications or falsifications are not certain. They may always be challenged, because they are always fallible, and new evidence may always cause us to reconsider our decisions. Even in the case of “pure existential statements,” where falsification is impossible because without a spatio-temporal specification, they force us to search throughout the universe and throughout space and time to falsify them, verification of one of them can be challenged by criticisms showing that the verifying statement was in error. So such a specific verification is falsifiable, even though the pure existential statement that was verified by it, is not.
Finally, even though Neil’s characterization of Popper’s position is, with the qualifications I’ve given, close to what he thought, I think his view on the importance of falsifiability in science is more a consequence of his prior acceptance of fallibilism and the idea that no statement can be demonstrated to be true, than his acceptance of fallibilism is a consequence of his holding his views on falsifiability. In particular, I think that Popper’s falsificationism is a response to how seriously he takes the Friesian Trilemma. In 1830 Fries, a Kantian, pointed out that when someone asks the classical “how do you know” question of epistemology with respect to any statement, calling for a justification in terms of another true statement and a valid logical argument demonstrating that the first statement followed from the second, he or she was led to one of three results. Either a foundational statement that was accepted as true based on authority (dogmatism), or a foundational statement accepted as true based on intuition (psychologism), or an infinite regress. The implication of the Friesian Trilemma is that no statement can be justified by another statement, even though the classical demand of Cartesian Epistemology was that “rationality” requires such a justification.
What Popper saw was that to escape from the Friesian Trilemma one had to reject the idea that “rationality” and knowledge require justification and that epistemology was the study of conditions fulfilling the need for justification. Instead, he accepted the idea that no statement can be justified, that every statement and every system of statements is fallible, that there is no certainty in life or science; and his response was to offer an epistemology and a philosophy of science which rejected the need for justification, which based the growth of knowledge on testability, falsifiability, falsification, and, in the end, (in his evolutionary epistemology phase) on even broader notions of critical evaluation and rationality, which emphasized critical comparison of alternatives and survival of ideas that have withstood the best critical evaluation we can muster. His theory of evaluation, thus, is entirely negative, and always assumes that none of our knowledge about the world is certain.