Contingencies announces an exciting new series of posts on Karl Popper, and Popper's mistakes, with assistance from one Nicholas Dykes, an English writer and philosopher of science (that does not mean, except to the dysfunctional, that we completely buy libertarianism, or the Objectivism that Dykes sometimes leans towards (as in Ayn-Randian Ob.). Regardless, Dykes brings up some important points criticizing Popper's "Critical Rationalism", as the following passages indicate:
""""Popper built his philosophy on foundations borrowed from Hume and Kant. His first premise was wholehearted acceptance of Hume's attack on induction. The second, to be addressed in the next section, was agreement with Kant's view that it is our ideas which give form to reality, not reality which gives form to our ideas.
Hume, whom Popper called "one of the most rational minds of all ages" [PKP2 1019], is renowned for elaborating the 'problem of induction' - a supposedly logical proof that generalisations from observation are invalid. Most later philosophers have accepted Hume's arguments, and libraries have been filled with attempts to solve his 'problem.'
Popper thought he had the answer. "I believed I had solved the problem of induction by the simple discovery that induction by repetition did not exist" [UNQ 52; c.f. OKN 1ff & PKP2 1115]. What really took place, according to Popper, was CR, knowledge advancing by means of conjecture and refutation: "... in my view here is no such thing as induction" [LSCD 40]; "what characterises the empirical method is its manner of exposing to falsification, in every conceivable way, the system to be tested" [LSCD 42].""""
Good, if a somewhat naive reading of Hume's points on induction. Not so naive readings would consist of awareness of undecidability of various issues, whether in sciences or social sciences. The current problems with identifying all the causal factors associated with global warming (ie CO2: culprit or benign? most likely benign) and atmospheric physics somewhat Humean. Few humans would care to wager on weather predictions for say October, or the stock market, or World Series). At the same time, unpredictable is not the same as undetermined (a point Bricmont makes, rather forcefully, and with boo-coo partial derivatives).
Dykes does not address falsification in much depth either; we might also question whether Hume's skepticism regarding induction entails Popper's falsification ("what characterises the empirical method is its manner of exposing to falsification, in every conceivable way, the system to be tested"). Popper tends to be a bit "conclusionary" himself; at the same time, much empirical research (whether in sciences, economics, social science, history, etc) does often relate to Popperian themes (even statisticians are ad hoc Popperians. Confidence intervals, margins of error, sampling problems---not to say reliability of any data. That is Popperian (or perhaps Kuhnian, who updates KP)).
""""Hume, said Popper, had shown that: "there is no argument of reason which permits an inference from one case to another... and I completely agree" [OKN 96]. Elsewhere he referred to induction as "a myth" which had been "exploded" by Hume [UNQ 80]. He further asserted that "There is no rule of inductive inference - inference leading to theories or universal laws - ever proposed which can be taken seriously even for a minute" [UNQ 146-7; see also RASC 31].
The Problem with 'The Problem'
Popper's solution was certainly correct in one respect. The problem of induction would indeed vanish if there were no such thing as induction. However, the issue would be resolved much more positively were it to turn out that Hume had been wrong, and that there never had been any problem with induction in the first place. And, in point of fact, this is the case. Despite his great skill as a thinker and writer, Hume missed the point. Induction does not depend for its validity on observation, but on the Law of Identity.
Hume stated, in essence, that since all ideas are derived from experience we cannot have any valid ideas about future events - which have yet to be experienced. He therefore denied that the past can give us any information about the future. He further denied that there is any necessary connection between cause and effect. We experience only repeated instances, we cannot experience any "power" that actually causes events to take place. Events are entirely "loose and separate.... conjoined but never connected."8
According to Hume, then, one has no guarantee that the hawthorn in an English hedge will not bear grapes next autumn, nor that the thistles in a nearby field won't produce figs. The expectation that the thorn will produce red berries, and the thistles purple flowers, is merely the result of "regular conjunction" which induces an "inference of the understanding."9 In Hume's view, there is no such thing as objective identity, there is only subjective "custom" or "habit."
However, Hume also wrote: "When any opinion leads to absurdities, it is certainly false"10 and the idea that one might gather grapes of thorns or figs of thistles is surely absurd enough to qualify. And false is what Hume's opinions most certainly are. Left standing, they lead to what he himself called "the flattest of all contradictions, viz. that it is possible for the same thing both to be and not to be.""""""
OK, but at the same time, the law of identity is a logical law, and analytically true: 5 squared will always equal 25. The pythagorean theorem (identities regarding the relationship of a right triangle) holds whereever you are, as does modus ponens or the law of non-contradiction and other logical givens (some freaks might dispute the LONC: they are mistaken). But a rose might be red this year, white next year, or non-existent. The naive reader of Hume (or Popper) utters "the bastard's saying inductive knowledge, indeed science, is impossible"! Yet the real point concerns epistemology: whatever sort of knowledge induction is, it is NOT logical and mathematical. Yes, ceteris paribus as they say, the roses will probably be red, and our observations about regular order to nature will be confirmed, and we will grant, reluctantly, that Hume (and his disciple Popper) may have overstated the case for subjectivity. But that "that bush with red roses will, nearly with certainty, have red roses next year" is not necessary in the sense of 5 squared equals 25; a botanist knowledgable in Mendelian genetics could hardly predict the flower's exact shape or form or color.
Dykes errors at least slightly here: evolution itself shows adaptationism (as Popper, often somewhat Darwinian, realized). What appear to be regularities often are not (especially over centuries); probability always enters the picture with inductive knowledge, and in many cases, even with hard sciences (say medicine) things may remain unresolved; Wittgenstein in the Tractatus often seemed somewhat Humean: "We cannot infer the events of the future from those of the present. Superstition is nothing but belief in the causal nexus." (TLP, 5.1361). At the very least inductive knowledge is NOT the same as the axiomatic knowledge of logic and mathematics, whether one holds that axiomatic knolwedge to be a priori or not.
"""""The crux of the case against Hume was stated in 1916 by H.W.B. Joseph in An Introduction to Logic: "A thing, to be at all, must be something, and can only be what it is. To assert a causal connexion between a and x implies that a acts as it does because it is what it is; because, in fact, it is a. So long therefore as it is a, it must act thus; and to assert that it may act otherwise on a subsequent occasion is to assert that what is a is something else than the a which it is declared to be."12 Hume's whole argument - persuasive though it may be - is, to borrow Joseph's words, "in flat conflict with the Law of Identity."13
Existence implies identity. It is not possible to exist without being something, and a thing can only be what it is: A is A. Any actions of that thing form part of its identity: "the way in which it acts must be regarded as a partial expression of what it is."14 Thus to deny any connection between a thing, its actions, and their consequences, is to assert that the thing is not what it is; it is to defy the Law of Identity.
It is not necessary to prolong this discussion. Entities exist. They possess identity. By careful observation - free from preconception - we are able to discover the identities of the entities we observe. Thereafter, we are fully entitled to assume that like entities will cause like events, the form of inference we call induction. And, because it rests on the axiom of the Law of Identity, correct induction - free from contradiction - is a valid route to knowledge. The first premise of CR is therefore false."""""
That "law" works with equations, functions, numbers, argument forms. Does it work with roses or roosters? Not quite. There's an egg. Safe bet that if a chick hatches it will have one head. Then after a few years, a two-headed bird appears! Sacre bleu! St. Hume at work again.
"""""There is nonetheless a substantial grain of truth in Hume's position, or few philosophers would have followed him. The grain lies in the precision of our knowledge of future events. Hume denied all knowledge of the future because we can have no experience of it. As we have seen, this is not true, it overlooks the Law of Identity. What is true, is that our prediction of events is limited by the unforeseeable. An 'O' ring may fail and destroy an otherwise reliable spacecraft; an icy road surface may cause a pristine Rolls-Royce to crash. For, no matter how sound our judgement nor wide our experience, we cannot possibly have complete, certain and absolute knowledge of future events. We are not omniscient: all kinds of unforeseen happenings may intervene to spoil even the best laid of our plans. Further, new information about old subjects continuously comes to light and, over time, things can evolve or change. Nonetheless, armed with the Law of Identity, there is no reason to allow the unforeseeable to turn us into sceptics. The universe is not a series of "loose and separate events" any more than time is a series of discrete, unrelated segments of duration.
It should also be noted that, in fact, all knowledge of entities, and all knowledge of language, is acquired inductively. A child's knowledge of apples, for example, is based on a very limited sampling. A student's knowledge of the word 'inference' is founded on a similarly narrow acquaintance. If it were true that induction is a myth, then all knowledge of external reality, all language, and all human thought - which depends on knowledge of reality and on language - would be myths as well, including, of course, CR...."
That's the naive reading of Hume again. Hume read his Newton fairly closely (supposedly), and does not really suggest radical skepticism: he's saying inductive knowledge, whatever it is, does not operate like deductive knowledge operates. Yes, the cannonball trajectory, if we know all the parameters and initial conditions (another problem), will nearly with certainty follow a predictable parabolic path. Yet whatever it is that gives that regularity is not at all the same as the pythagorean theorem. It's damn near certain, but not necessary (that's not to suggest that we take a naive reading of quantum physics and uphold "non-locality" either; locality appears to be the default position (as Bricmont also argues). Hume arguably does not deny determinism, and is not as subjective as Dykes believes, nor is Popper: Hume's doing epistemology, not physics. If determinism does hold for physical reality and our knowledge of the physical world, it's a quite a different type of thinking than the axiomatic thinking of logic and mathematics (though that's not to make claims for any quasi-platonic dualism either).