Muddying the Mud: Complicating Lakatos’s Synthesis of Popper and Kuhn

by Shaun Terry

It seems that part of what Lakatos is doing is to synthesize Popper’s and Kuhn’s theories in order to save them from each other. Lakatos reframes Kuhn’s paradigms as scientific research programmes, by which theories can survive substantive challenges. He retains Popper’s linearity, but with the caveat that falsification is not as simple as refuting a theory once and that paradigm shifts require greater empirical power for superseding theories to displace previously dominant ones. All this said, I propose that Lakataos’s position has shortcomings in the form of arbitrary and vague definitions, as well as in the form of linearity that Lakatos proposes (with which Kuhn seems to deal more accurately).

Lakatos seems to be saying that one cannot falsify a theory without calling into question the tools and methods being applied, including “interpretative” theories. In essence, one can never be sure what it is that we are falsifying. Here, I agree with Lakatos.

However, I believe that Lakatos’s conception of Popper2 is problematic in that he describes Popper seeing progress as linear while Lakatos describes the problem of not being able to tell whether inconsistencies come from interpretative or explanatory theories. His position seems to be that, by showing preference to theories that expand our knowledge and/or ability to predict, we are always moving forward in linear fashion, but I am not convinced. If we accept an idea as fact, only to later reject it, only to reaccept it, then on what bases are we accepting or rejecting?

What it seems to me is that inconsistencies show problems with compatibilities of the theories with which we operate. We do not seem to have a way to parse out which theory(ies) might or might not be correct, as we cannot prove anything. If we (tentatively) accept the set of theories used in an experiment, any inconsistency only demonstrates that some theory necessary to the expected outcome is incorrect. It seems that we can at no point accept that a particular theory is right or wrong. Instead, we try to make the best guesses we can, based on deductive reasoning and the extrapolation/hope of sense (empirical) observation.

Lakatos states that Kuhn’s paradigm shifts are not logical——that they are merely “mob psychology.” I think that he is right to point this out, but I also do not think that Kuhn is wrong. Lakatos and Kuhn seem to recognize that paradigm shifts coincide with the ability of new paradigms/research programmes to answer the most troubling contemporary scientific questions, but Kuhn emphasizes the arbitrariness of these shifts in that new paradigms do not always answer all the questions that old ones do, while Lakatos seems to emphasize that there are always competing research programmes.

I do not think that we should be so sure about these shifts, though. Refutations cannot give us sufficient cause to eliminate a theory. A series of refutations, preferably in varying contexts, has historically suggested the rejection of theories, but such cases are still reliant on inductive reasoning. We might conduct an incredible number of experiments, all of which rely on a specific theory. If all such experiments fail to produce outcomes consistent with the theory, this does not mean that we should reject the theory. It may be that, in each case, another theory on which the experiment relied was the cause for the inconsistency.

Lakatos describes his negative heuristic and intermittent success of theories, but why should we accept that these are necessarily successes of Lakatos’s hard core and not just coincidences? Further, Lakatos does not seem to sufficiently define this intermittency.

Acceptance and rejection can only ever be tentative as, I argue, acceptance and rejection are unreasonable. The only reason to tentatively accept or reject anything is on a pragmatic basis. Otherwise, doing so relies on factors outside of the validity of the theory, including that the associated theories interact with the subject theory to produce reliable outcomes, that new cases produce outcomes consistent with old cases, and that there is no shifts of fundamental reality. Lakatos attempts to explain why theories are sometimes not rejected despite refutations, but does not sufficiently explain why logic would dictate that any theory should be accepted or rejected.

Perhaps Lakatos makes an attempt at speaking to pragmatism, but he is not so clear. Lakatos points out, “Popper1 argued that one had to choose between naïve falsificationism and irrationalism.” But is it the case that we have two choices and only two choices? Popper seems to point out that arbitrarily selecting what experimental aspects to reject is no better than irrationalism, but does falsificationism necessarily do better?

Lakatos asks, “If the criticisability of the empirical basis is to be taken seriously, do we not ‘throw empiricism overboard’?” I would simply say “No.” Again, this seems like a false dichotomy. There may be infinite other choices. Moreover, Lakatos seems to (dangerously) suggest that we should not take seriously the criticizability of empiricism.

For Lakatos, the value in a theory rests in whether or not it increases knowledge. Lakatos takes the position that theories that increase knowledge are scientific, whereas other proposals are pseudo-scientific. However, this begs the question of what constitutes knowledge. This question is not only reliant on the nature of knowledge but also on the efficacy of science——and its constituent aspects——to produce knowledge. To be clear, I agree that science produces means of framing the world, but to describe what science produces as “knowledge” seems to go farther than I would argue is reasonable.

To that point, when Lakatos quotes Popper’s position on “accepting the majority verdict of the scientists’ jury,” this seems to me like nothing more than an appeal to the majority of experts. I can see the attractiveness in such an appeal, but it is not a logical argument for the acceptance in anything.

To sum, Lakatos makes some interesting points, but I do not feel that he sufficiently defines his terms and I do not feel that he sufficiently answers all the questions that arise from Popper’s and Kuhn’s proposals.