I’ve been having an ongoing exchange with Ryan Murphy and Gene Callahan about Popper, the problem of induction, and falsifiability. What follows is brief summary of our inconsequential little argument to provide context. If you don’t care about all that and just want the meat of the post, then just skip the next four paragraphs.
The exchange got started after I shared Chapter 12 of Bartley’s Unfathomed Knowledge with Ryan. He soon published a post stating: ‘[Bartley] really helped me gain some perspective on Wittegenstein, twentieth century philosophers, and everything that is wrong with them.’ A brief exchange followed in the comments involving Ryan and me, where Gene Callahan objected to Bartley’s interpretation of Wittgenstein and, among other things, maintained that Popper was a positivist. Gene then posted a couple of dismissive quotes on his own blog.
That might of been the end of it, but Gene later decided to drag up old arguments accusing Popper’s method of conjecture and refutation of implicitly depending on induction, specifically referring to the so-called ‘principle of conservative induction‘. These arguments are greatly flawed, and I set out to explain exactly how in a guest post on Ryan’s blog, since that was the forum of our previous disagreement.
Gene responded directly. While he clarified and elaborated on his previous argument, he appears to have entirely ignored my counterargument on the matter, instead choosing to focus on a rather banal logical point to the detriment of more relevant arguments. I originally attempted to explain and defend myself in the comments on his blog, but Gene refuses to approve them because ‘it would lead to what might appear to be a discussion’.
My intent was to leave the matter there, but Ryan shared a thoughtful appraisal of the discussion that Gene was afraid it might appear we were having. Ryan attempted to explain how me and Gene were, for the most part, talking past each other, and he draws and important distinction between induction as the source of our conjectures and induction as the justifier of our beliefs.
This distinction that Ryan raises is very important point that is rarely made explicit. I actually alluded to it in one of my comments on Gene’s blog, a comment which he has declined to publish. The distinction is sometimes referred to as induction in context of discovery and induction in the context of justification, and it strikes to the heart of a lot of confusion.
In any case, I want to clarify my position here, because this distinction Ryan mentions was not something I had ignored. I wish to make it clear that Gene is wrong, or at least misleading, even in the narrower sense in which Ryan interpreted his arguments (which I think are pretty close to what Gene had in mind).
In the context of justification, we are more concerned with the logic of induction, and whether it is possible to derive, by some transformational rule, scientific hypotheses from sensory experiences. Moreover, if such a rule can be formulated, what does it transmit from the premises to conclusion? Truth? Probability? Scientific status? In the context of justification, we pay more heed to purely logical questions. We’re not attempting to give a psychological account of how hypotheses are discovered, but rather trying to address if or how beliefs in hypotheses are justified inductively.
In the context of discovery, induction mutates from a inferential rule of logic to a description of cognitive processes. The concept of logical consequence is tacitly substituted for psychological cause and effect; abstract propositions are exchanged for psychological categories like experiences, thoughts, and beliefs. In the context of discovery, the ‘premises’ of an inductive argument are the sensory experiences which cause, physically or mentally, their ‘conclusion’, a novel thought of some scientific hypothesis.
Hume’s famous attack on induction can be largely understood in the context of justification. That is, he accepted that people thought inductively and would continue to, but he could just find no epistemological justification for them doing so.
When Gene says ‘of course, no one could make it through the day without making continual inductive inferences,’ he’s talking in the context of discovery rather than the context of justification. He is describing causal relations between thoughts rather than logical relations between propositions. This is why he makes statements like, ‘so of course Popperians make [inductive inferences] all of the time.’ This is not about the justification of beliefs, since Popperians typically hold that such justification is neither possible nor desirable, but rather it is about the causal relation between their thoughts and experiences.
In a very narrow sense of the term ‘induction’, Gene is quite right–words, after all, can mean anything one wants them to mean. Personally, I prefer so say something like ‘I was inspired by my experiences to conjecture a new hypothesis’ rather than ‘I induced the hypothesis from my experiences’, because the former is more explicitly psychological, about the causal relations between experiences and thoughts, whereas the latter is easily confused for some kind of logical procedure called ‘induction’.
In this psychological sense, we “do” induction all of the time. Indeed, our thoughts rarely proceed in a sequence corresponding to anything like a deductive argument. On the contrary, they bounce around erratically, a mix of tacit reactions or impressions and a smattering of slow, purposeful ratiocination. From the perspective of a logician, our thoughts might appear to be little more than a long sequence of logical fallacies (e.g. denying the antecedent, category error, false analogy, ad hominem, middleground fallacy, etc.)
But any such “fallacious” thinking, inductive or otherwise, could lead to the discovery of some new and interesting hypothesis. Should we attempt to construct special logics to describe each of them? In fact, in this sense, ‘one could not make it through the day without making continual invalid inferences’ and ‘Popperians make invalid inferences all of the time.’ But what does this say besides that our thoughts do not always relate in a way analogous to the premises and conclusion of a deductive argument? But whoever claimed that? Not critical rationalists.
While the creative act might be psychologically explicable as a physical or mental process, the abstract content exemplified by one thought need not be logically derivable from some previous thought or experience. This is merely the nature of creative guesswork or conjecture, the source of new hypotheses, whether inspired by previous experience, as with ‘induction’, or the result of abduction, head trauma, or anything else.
In his Logic of Scientific Discovery, Popper described this confusion of logical and psychological issues as ‘psychologism’ and warned against it. He explicitly denied to be trying to describe how scientists think, but rather his concern was with the objective growth of knowledge. That his critics still fail to even appreciate, much less recognise, these distinctions, just demonstrates how poorly equipped they are to understand the critical rationalist position on these matters.