A Refutation of Nicholas Dykes on Karl Popper

Many years ago, Nicholas Dykes wrote a criticism of the philosophy of Karl Popper. His essay not only fails to refute Popper, it also seems to reflect an inability on Dykes’ part even to state Popper’s positions accurately. I will go through the arguments section by section, labelling what section each argument comes from. I have left out references and footnotes in my quotes, the reader can get them from the original article. Where I have put in references I refer to Section numbers rather than page numbers since different versions of Popper’s books have different page numbering.

Section 1

Dykes tries to describe Popper’s position:

CR was originally developed by Popper to demarcate science from non-science. He stated that for scientific knowledge to be considered knowledge it had to be refutable: “‘In so far as scientific statements refer to the world of experience, they must be refutable … in so far as they are irrefutable, they do not refer to the world of experience'” [OSE2 13].

The quote is misleading. Popper didn’t develop CR to demarcate science from non-science, he did it to explain the growth of knowledge: “to analyse the method of the empirical sciences” (LScD, Chapter I). A fuller quote from the source Dykes used makes Popper’s actual position clearer:

In the empirical sciences, which alone can furnish us with information about the world we live in, proofs do not occur, if we mean by ‘proof’ an argument which establishes once and for ever the truth of a theory. (What may occur, however, are refutations of scientific theories.) On the other hand, pure mathematics and logic, which permit of proofs, give us no information about the world, but only develop the means of describing it. Thus we could say (as I have pointed out elsewhere): ‘In so far as scientific statements refer to the world of experience, they must be refutable; and, in so far as they are irrefutable, they do not refer to the world of experience.’ But although proof does not play any part in the empirical sciences, argument still does; indeed, its part is at least as important as that played by observation and experiment.

According to Popper, knowledge that is not empirically testable is not scientific, but it is still knowledge.

More from Dykes:

It follows that we can never attain certainty: “The quest for certainty… is mistaken…. though we may seek for truth… we can never be quite certain that we have found it” [OSE2 375]. “No particular theory may ever be regarded as absolutely certain…. No scientific theory is sacrosanct…” [OKN 360]. “Precision and certainty are false ideals. They are impossible to attain and therefore dangerously misleading…” [UNQ 24]. He summed up with an oft-repeated aphorism: “We never know what we are talking about” [UNQ 27].

Accordingly, Popper refused to grant any philosophical value to definitions: “Definitions do not play any very important part in science…. Our ‘scientific knowledge’… remains entirely unaffected if we eliminate all definitions” [OSE2 14]. “Definitions never give any factual knowledge about ‘nature’ or about the ‘nature of things'” [C&R 20-21]. “Definitions…. are never really needed, and rarely of any use” [RASC xxxvi].

This is a false description of Popper’s position on definitions, see Section 7 of Unended Quest. His actual position is that words can’t be defined with perfect accuracy since all definitions have to employ undefined words. Suppose I say that a tiger is “a big cat”. If all the terms are defined then there is a definition of “a” and “big” and “cat” and these definitions refer to other words and if those definitions are defined then they lead to still more definitions with still more words… So either we use undefined words or we have an infinite regress. As a result of this problem the habit of trying to understand the world by coming up with the right definitions, which Popper calls “methodological essentialism”, is untenable.

A word is shorthand for an idea not for some imaginary perfect definition. As such, we should be willing to change terminology to talk in ways that other people understand. That is, we should be willing to use their definitions. Furthermore, we should never try to be more precise than is necessary to address the problem we are dealing with since this will lead to loss of clarity. The best summary of his position on this issue is (Unended Quest, p. 24):

Every increase in clarity is an intellectual value in itself; an increase in precision or exactness has only a pragmatic value as a means to some definite end – where the end is usually an increase in testability or criticizability demanded by the problem situation…

Dykes never states let alone answers Popper’s argument on definitions.

Section 2

Dykes tries to refute Popper’s arguments against inductivism. Inductivism is the belief that there is a method called induction by which people can get new ideas from sense information and then confirm those ideas.

Popper built his philosophy on foundations borrowed from Hume and Kant.

No he didn’t. Popper argued that knowledge doesn’t have foundations (see in particular “Realism and the Aim of Science” Chapter I), and Dykes hasn’t refuted that argument, nor does he do so anywhere in the essay.

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].”

There then follows a lot of stuff on Hume’s position on induction, not Popper’s. Popper had many arguments against induction that Hume didn’t give him and took those arguments more seriously than Hume. Dykes does not address most of the arguments Popper provided.

What argument does Dykes attempt to address?

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.”

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.” In Hume’s view, there is no such thing as objective identity, there is only subjective “custom” or “habit.

Dykes then goes on to argue as follows:

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.” Hume’s whole argument – persuasive though it may be – is, to borrow Joseph’s words, “in flat conflict with the Law of Identity.”

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.” 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.

This argument doesn’t solve the problem at all. The problem of induction as stated by Hume is that our expectations of the future don’t follow from what we have observed in the past. To see why let’s take Dykes’ example of the hawthorn, which he claims will not produce grapes. How does he know it won’t produce grapes? Perhaps some scientist will genetically engineer hawthorns to produce grapes. And even if he doesn’t the fact that it won’t produce grapes doesn’t follow merely from the fact that it hasn’t in the past.

To put this in Dykes’ language, if we were to accept that existence implies identity that would not tell us the identity of any specific entity. And indeed characterising the issue as being about the identity of the object in question is a bad way to think about it. Whatever the thing in question is we need an explanation of how it works to say what it will do next and why. And we won’t be able to tell what we can predict about the entity in question without such an account. Why do hawthorn bushes not produce grapes? That has to do with a complicated set of circumstances in its evolutionary past that selected against hawthorns producing relatively large fleshy fruit and refers to lots of things that are not hawthorn bushes, like human beings who did not selectively breed hawthorn bushes to get them to grow grapes. Stating this theory in terms of definitions would make it less clear because the explanation involves tying together many different entities and so the whole explanation would have to be repeated many times in slightly different ways.

Note also that Dykes’ approach to creating knowledge amounts to defining terms in the right way: that is, to the idea of methodological essentialism that I criticised in my comments on Section 1.

Section 3

Dykes then discusses the idea that all observations are theory-laden, which he describes as Popper’s “Kantian premise”:

[I]f it is true that our senses are pre-programmed; if it is true that “there is no sense organ in which anticipatory theories are not genetically incorporated” [OKN 72]; then what flows into our minds is determined and what flows out of them is subjective. If our senses are not neutral, if they organise incoming data using pre-set theories built into them by evolution, then they do not provide us with unalloyed information, but only with prescriptions, the content of which is determined by our genetic make up. Whatever is thereafter produced inside our heads – cut off as it is from any objective contact with reality – must be subjective.

It is not true that the fact that our senses are theory-laden implies that we are cut off from objective reality. More broadly the problem is this: all sources of knowledge are flawed in some respects, so how can we ever learn anything? First, knowledge can have implications beyond the problem that it was invented to solve. Our eyes can be used to do things they did not evolve to do like looking at readings on scientific instruments: they are limited by the content of the knowledge instantiated in them not by the problem they originally evolved to solve. Second, what counts is not whether any particular piece of information we get is flawed. Rather, what matters is whether we can correct those flaws. We can do this by comparing different sources and trying to come up with an explanation of the underlying objective reality that explains what all of the sources say. They will not all be flawed in the same way and so we can try to work out which ones are giving us bad information on any particular issue. (See “On the Sources of Knowledge and of Ignorance” Section XIII onward in “Conjectures and Refutations” and Chapter 1 of Objective Knowledge). Dykes never mentions this argument let alone addresses it.

Section 4

In this section, Dykes plays a sort of verbal game saying that conjecture is not knowledge because conjecture means “an opinion formed on slight or defective evidence or none: an opinion without proof: a guess.” Popper does think all of our knowledge consists of unproven guesses. Dykes is not satisfied with this but logically it is no different than the problem that all sources of information are flawed and so it does not need a distinct answer from the one given in the previous section.

Section 5

Dykes then claims that Popper does not give a way for us to select what conjecture we should consider next. But why would this be needed? If your propose a conjecture and it fails tests you just ditch it and try another. Any rule for picking the next conjecture would imply that there are some ideas you can’t question: the rules for selecting the conjectures. If those rules are flawed in any way then you might miss some flaws in your ideas that you would otherwise see. Dykes raises yet another variant of the problem that all of our knowledge is flawed, which he still has not addressed.

Section 6

Dykes attempts to criticise the idea that science can be distinguished from non-science by scientific theories being refutable by observation or experiment.

Relatively few philosophers have embraced it however. Tom Settle, a major contributor to P.A. Schilpp’s massive festschrift, The Philosophy of Karl Popper, stated firmly in 1970: “As a criterion of demarcation between science and nonscience, Popper’s ‘falsifiability’-plus-a-critical-policy does not work” [PKP2 719]. Other contributors evidently agreed; among them A.J. Ayer, William C. Kneale, Imre Lakatos, Grover Maxwell, and Hilary Putnam.

Dykes doesn’t discuss the content of their criticisms, except for one mention of Maxwell.

Next Dykes tries to produce an actual criticism:

One can understand the importance of the distinction to the young Popper. Fascinated by science, he was surrounded by true-believing Marxists and Freudians all of whom claimed science on their side while espousing doctrines which seemed to Popper obviously false. Nonetheless, ‘refutability’ seems to miss the mark. The ideas of Marx or Freud stand or fall on their conformity to logic and the available evidence – in exactly the same way as the ideas of Newton or Einstein. Marxism and Freudianism failed to survive as viable theories due to myopic concentration on a narrow range of data, false interpretations of evidence, and logical inconsistency. They never were ‘irrefutable.’ They failed precisely because they could be, and were, refuted; either by contrary evidence, by exposure of contradictions, or by the resolute refusal of reality to conform to their predictions. It wasn’t refutability which made them unscientific, it was inaccuracy and/or illogicality.

Marxism and psychoanalysis can’t be refuted by experiment because they either started out as too vague to make any specific claims about what would happen or were changed to make them vague. If a theory doesn’t make specific claims it can’t be refuted by observation. As such observational or experimental evidence are not relevant to judging them. They may fail in other ways but they don’t make the grade for being scientific theories for that reason alone even if they had no other flaws. Note also that Popper didn’t disqualify using other kinds of criticism to eliminate scientific theories so Marxism and psychoanalysis could have been eliminated by finding logical flaws even if they were not empirically testable.

Dykes continues:

Science is distinguished by its strict adherence to physical evidence. Non-science, on the other hand, is invariably characterised by preconception, followed by a cavalier disregard for, or rationalisation of, anything that doesn’t fit into the preconceived schema. In one sense, this is what Popper was saying. But, due perhaps to his dislike of definitions, he homed in on the wrong identifying characteristic.

Whether an idea can be tested experimentally is important since there is no point in trying to test an idea experimentally if no such test is possible. If Dykes just wants to call all good ideas scientific then he can do that. However, we would then need a new term for experimentally testable scientific ideas. If Dykes were to say, as David Deutsch has, that a scientific theory should be a good explanation in addition to being experimentally testable this might be a good idea, but Dykes doesn’t make this point.

Grover Maxwell pointed out that ‘All men are mortal’ is a perfectly sound scientific statement which is not falsifiable [PKP1 292]. Popper defended himself robustly [PKP2 1037ff], but Maxwell seemed to have the stronger case.

What was Popper’s counterargument? Dykes doesn’t say, which means the reader has nothing to go on when trying to make a decision between Maxwell’s position and Popper’s since Popper’s position hasn’t even been stated. What was Popper’s reply? If you try to refute a theory that implies that all men are mortal and you fail to do so, then “all men are mortal” has also passed a test by virtue of the fact that it is a consequence of a theory that has passed that test. An example of such a theory might be that biology might limit the extent to which cells can be repaired and as a consequence all men die before they get to 150, say. No such theory has actually been proposed so I would actually class “all men are mortal” as a bad example cooked up by philosophers.

Next, we get another argument:

Maxwell might also have taxed Popper about mathematics. The axioms of mathematics cannot be refuted. According to the demarcation theory, therefore, mathematics is not a science. But physics is inseparable from mathematics. Quantum mechanics, for example, could hardly be expressed without it. So physics cannot be a science either. Much the same could be said about logic. The Law of Contradiction, etc, cannot be refuted, so logic is not a science.

By irrefutable I presume Dykes means “not experimentally testable” since mathematical and logical theories can be refuted by non-experimental arguments. Maths is not a science, nor is logic. Rather, they are ideas we use to help us express other ideas more clearly. This does not mean that the ideas thus expressed are untestable. Quantum physics claims that particular mathematical structures provide an accurate depiction of how the world works and that is the testable claim made by quantum physics. It is a claim not primarily about maths but about the laws of physics. If a particular calculation is done badly it can be tested in non-experimental ways, e.g. – redoing the calculation or doing it using a different technique, so you can’t wiggle out of experimental tests just by saying some piece of maths might be wrong. Physics also uses other untestable ideas, e.g. – philosophical ideas like realism.

Dykes then tries to poke holes in Popper’s position on Marxism:

Marxism was one of the theories which led Popper to develop his conception of demarcation in the first place: “I had been shocked by the fact that the Marxists… were able to interpret any conceivable event as a verification of their theories” [UNQ 41-2]. Yet in “Replies to my Critics” Popper changed his tune: “Marxism was once a scientific theory”; “Marxism was once a science” [PKP2 984-5]. No doubt Popper would have swamped this objection with distinctions between Marx and Marxism, but the notion that Marxism could both be and not be a science does little to inspire confidence in Popper’s theory of demarcation.

A fuller quote explains what Popper was actually saying and it is not contradictory from The Philosophy of Karl Popper v1, v2, edited by P.A. Schilpp, pp. 984-5:

Marxism was once a scientific theory: it predicted that capitalism would lead to increasing misery and, through a more or less mild revolution, to socialism; it predicted that this would happen first in technically highest developed countries; and it predicted that the technical evolution of the “means of production” would lead to social, political, and ideological developments rather than the other way round.

Popper then describes a series of developments that refuted all these implications of Marxism and so refuted the theory itself. He then adds:

However, Marxism is no longer a science; for it broke the methodological rule that we must accept falsification and it immunized itself against the most blatant refutations of its predictions.

The version of Marxism that is not scientific is a modified version that has been stripped of its empirical predictions according to Popper. Dykes has attacked a position Popper didn’t hold, namely that a particular idea can be both scientific and unscientific.

Section 7

Dykes attacks Popper for only allowing negative statements about what does not exist:

The scholastic principle, “we cannot know what God is, but rather what He is not” is remarkably similar to Popper’s assertion that “natural laws…. do not assert that something exists or is the case; they deny it” [LSCD 69].

Popper’s point is that the testable content of a theory consists of what it excludes not in existential statements. For if I just say “neutrinos exist” I could always wiggle out of a purported refutation by saying you haven’t looked for them in the right place. But if I say that it is impossible to violate the conservation of energy in a nuclear reaction and predict the existence of a neutral particle with a particular momentum distribution then I have made a testable statement that implies the existence of the neutrino.

More from Dykes:

However, recalling Popper’s Kantian premise, one might reasonably enquire at this point: if all observations are theory-laden, and thereby suspect, what justifies our placing any confidence in negative observations? The procedure of observation is identical whether one is seeking evidence in favour of a theory, or testing for evidence against it. If our senses are automatically suspect, as Popper maintained, negative or falsifying instances deserve no more credibility than positive or confirming ones.

Dykes is assuming that we need to have an argument showing an observation is true or probably true to use it, but this is simply not so. We can instead treat observations as testable conjectures. If we have made a mistake about some observations then when we find out about the mistake we can reconsider the theories that were supposedly refuted by them. We should treat the creation of knowledge as an error-prone process and look for problems. Dykes fails throughout the entire essay even to mention this point let alone argue with it.

Further, remembering Popper’s Humian premise, one immediately wants to ask: If we are not allowed to argue from positive instances to true laws, why are we allowed to argue from counterinstances to negative laws (we were told above that “natural laws… deny”). The reasoning process is the same. Collecting disconfirmations and arguing negatively scarcely differs from collecting confirmations and arguing positively. Both are inductive procedures and, as such, have been disallowed in advance by Popper’s rejection of induction.

If a theory claims that every time X happens Y will also happen, then if X happens even once and Y doesn’t happen that theory is false. The mere fact that this particular theory is false does not imply that any other particular theory is true. There is no question of arguing from an observation to a negative law. In every case, we start with a proposed law, look for instances in which it is false and if we find such an instance we look for a new law.

There are no new points in the rest of the section so let’s go on to the next section.

Section 8

Dykes claims that CR has a problem with facts:

CR states that for knowledge to be regarded as scientific it must be falsifiable. Plainly then, if an item of ‘knowledge’ is falsified, it can no longer be regarded as a fact. In Popper’s own words, a false conjecture “contradicts some real state of affairs;” “falsifications… indicate the points where we have touched reality” [C&R 116]. What we are left with are conjectures which have not yet been falsified. But a yet-to-be-falsified conjecture can hardly be called a fact, ‘a real state of things.’ It is rather ‘a mere statement or belief’ from which facts are to be distinguished.

So any statement of fact should be considered a conjecture about what is the case in reality. That conjecture may be false. The solution to that problem is to look for problems with our ideas about reality, including our ideas about observations. This doesn’t imply that we have no knowledge about the world: it implies that our knowledge is conjectural. Dykes provides no argument about why this is unacceptable.

Section 9 is just a rehash of Dykes’ methodological essentialism. I criticised this position in my comments on Section 2, so I will move on.

Section 10

Dykes attacks Popper’s theory of the three worlds. World 1 is the world of physical objects, world 2 is the world of psychological states and world 3 is the world of the objective content of our knowledge. How does Dykes argue against this?

First, there seems little conjectural about the theory of worlds 1, 2, & 3. In none of Popper’s several presentations is the theory offered as an hypothesis. Rather, it is laid out as a discovery, as what Popper thought the facts to be.

Popper doesn’t precede every sentence with “this is a conjecture”. So what?

Second, the idea of objective knowledge appears directly to contradict CR. If knowledge can exist objectively, it is not clear how it remains at the same time conjectural. The exercise of studying Popper, for instance, depends on the existence of a dozen or so world 3 objects – his books. Now, either those books exist and say what they say or they don’t, there is simply no room for conjecture.

Either the books exist or they don’t. And for any particular statement either they make that statement or they don’t. They have content independently of the content people think they have. For example, the fact that Dykes has ignored a lot of Popper’s arguments doesn’t mean that Popper didn’t make them. And Dykes may think that he has answered Popper satisfactorily but that doesn’t mean that he has. The fact that we have to make conjectures about the meaning of a particular passage and that we may be wrong does not mean the book doesn’t have objective content. Indeed, we could not be mistaken unless they did have objective content about which we could be mistaken.

The existence of objective world 3 ideas also seems to conflict with Popper’s rejection of ‘essentialism’ – the real existence of concepts – which formed an integral part of his notorious attack on Aristotle and underlay his dislike of definitions. Surely it is unreasonable on the one hand to lambaste essentialism – the idea that concepts are, or have, real ‘essences,’ which exist in our own reality or in another dimension – while claiming on the other hand that concepts have a separate existence in world.

Dykes here is saying that Popper is attacking essentialism without paying any attention to what Popper said about it. Popper argues against methodological essentialism: the idea that the way to understand things is to produce precise definitions. Popper interprets Plato’s theory of forms as an attempt to explain why different objects may be similar in some respects The Open Society and Its Enemies, Vol. 1 Chapter 3 Section V:

If things are similar because of some virtue or property which they share, for instance, whiteness, or hardness, or goodness, then this virtue or property must be one and the same in all of them; otherwise it would not make them similar. According to Plato, they all participate in the one Form or Idea of whiteness, if they are white; of hardness, if they are hard. They participate in the sense in which children participate in their father’s possessions and gifts; just as the many particular reproductions of an etching which are all impressions from one and the same plate, and hence similar to one another, may participate in the beauty of the original.

Popper’s solution to this problem is that different objects obey the same universal laws discovered by science, which are “descriptions of the structural properties of nature.” (See Realism and the Aim of Science, Section 15.) Popper thinks we should discuss the way the world works in terms of theory, truth and arguments rather than definitions and concepts since the latter is methodologically bad for reasons I discussed earlier, see my criticism of Section 1. Popper’s world 3 is about the content of theory and arguments  that change over time as people create new knowledge not concepts and definitions, still less is it about the unchanging divine world of forms envisioned by Plato, see Section 5.1 of “Objective Knowledge”.

Finally, the ‘autonomy’ of man-made, objective knowledge shows a marked kinship to Aristotle’s concept of potentiality. Popper often used number theory to explain world 3: “natural numbers are the work of men,” he stated. However “unexpected new problems arise as an unintended by-product of the sequence of natural numbers…. These problems are clearly autonomous. They are in no sense made by us; rather, they are discovered by us; and in this sense they exist, undiscovered, before their discovery” [OKN 160-1]. That is fair enough, but is it not merely another way of saying that the future is not actual but potential; that unknown future advances do not actually exist, yet must exist as potential in the known?

It is not true that unknown future advances exist as potential in the known since, for example, Newtonian gravity did not in any way imply the future advances in our understanding of gravity: Einstein’s general theory of relativity. And even in maths it is an inaccurate way of discussing mathematical problems. Since we have no infallible way of telling which results can and cannot be derived from previous mathematical ideas we have to propose bold new conjectures and look at their consequences to try to figure out the truth.

Dykes likens the belief in world 3 to idealism:

For instance, in one of his several discussions of worlds 1, 2 & 3, he wrote: “I regard world 3 as being essentially the product of the human mind. It is we who create world 3 objects…. these objects have their own inherent or autonomous laws which create unintended and unforeseeable consequences…. [these] repercussions on us are as great as, or greater than, those of our physical environment” [UNQ 186]. Elsewhere he wrote of “the ‘objective mind’ or ‘spirit'” [OKN 149]; and that “the third world is… superhuman”, it “transcends its makers” [OKN 159]. But surely the notion of a transcendent mind or spirit which effects human beings more than their physical environment is a straightforward depiction of idealism?

This is wrong. The reason why world 3 is important is that it provides the knowledge that enables us to solve problems with our environment. The idea that surgeons can help prevent sepsis by washing their hands before surgery is more important then the mere fact that bacteria exist since the former allows us to overcome the latter. These ideas also create new problems that become more significant than the problem they were intended to solve. Most people don’t experience much direct physical violence in their daily lives and the government has helped solve that problem but it has created other problems, like economic problems caused by taxation and subsidies. Our attention is now focused on such problems that many of our ancestors could not even have imagined.

More from Dykes:

In The Open Universe, the idealist element seems even plainer: “we ought to admit the existence of an autonomous part of World 3; a part which consists of objective thought contents which are independent of, and clearly distinct from, the subjective or personal thought processes by which they are grasped, and whose grasp they can causally influence. I thus assert that there exist autonomous World 3 objects which have not yet taken up either World 1 shape or World 2 shape, but which, nevertheless, interact with our thought processes” [TOU 119-20]. It would be hard to describe ‘independent, autonomous, objective thought contents which influence human thought processes’ in other than idealist terms.

I think this is a worthy attempt by Popper to draw attention to the problem of how we should think about abstractions. The best account of them I have read is given in Chapter 5 of David Deutsch’s book The Beginning of Infinity.

In The Self and Its Brain Popper’s idealism becomes explicit. The thesis of the work, a joint effort by Popper and neuroscientist Sir John Eccles, consists of a revival of Cartesian dualism.

“The Self and Its Brain” is a terrible book but substance dualism is not idealism. David Deutsch has described the problems associated with how to explain thought, creativity and other problems in the philosophy of mind in Chapter 7 of “The Beginning of Infinity”.

Section 11

In this section, Dykes maintains that some established theories are not conjectural:

Yet theories do exist which, in fact, are positively confirmed, as Grover Maxwell has pointed out [PKP1 292ff]. Copernicus’s heliocentric theory, for example, was indeed hypothetical in 1543 because the instruments did not then exist with which to prove it. But now that huge telescopes and space probes have eliminated any rational doubt that the earth revolves around the sun, it would seem bizarre to maintain that heliocentricity remains conjectural.

I can’t think of any problem with the heliocentric theory at present. That doesn’t change its logical status: it is a well tested conjecture. And in some respects the heliocentric theory has been superseded because we don’t think of the sun as being the centre of the universe, only of the solar system, which is on the edge of a galaxy of billions of stars, which is part of a cluster of galaxies and so on. The best a theory can hope for is to preserved as an approximation to a better theory in the future.

Other similar issues are pointed out but I won’t go through them all. There is just one more issue to consider in this section:

Popper’s problem was of course that the theory of evolution is just about as inductive as one can get, yet he wanted us to believe that induction is a myth. He found no way out of this impasse, and in the end decided that the only solution was to evade the issue: “I have come to the conclusion that Darwinism is not a testable scientific theory, but a metaphysical research programme” [UNQ 168].

The theory of biological evolution is not inductive. It contradicts induction. Biological evolution proceeds by the production of variations on current genes and selection among them just as human knowledge creation proceeds by producing variations on current knowledge and selecting among those variations. Biological evolution does not consist of animals or plants somehow condensing observations into genes, which would have to be the case if biological evolution was inductive: that is the discredited idea of Lamarckism. One of many problems with Lamarckism is that it doesn’t explain what changes in an organism’s body get incorporated into genes and which do not just as inductivism does not explain how we pick what observations to do. See Chapter 4 of “The Beginning of Infinity”, which also explains how the biological evolution of life on Earth can be tested in principle. If you found that some organism only underwent favourable mutations then the postulate that mutations are random would be refuted and a new explanation would be needed. Deutsch also provides other examples.

Dykes also does not quote the full context of Popper’s remark that evolution is a metaphysical research programme:

From this point of view the question of the scientific status of Darwinian theory – in the widest sense, the theory of trial and error elimination – becomes an interesting one. I have come to the conclusion that Darwinism is not a testable scientific theory but a metaphysical research programme – a possible framework for testable scientific theories.

In other words, Popper was considering both biological evolution and human knowledge creation as being instances of a wider evolutionary explanation of knowledge creation – all knowledge must be created by trial and error elimination. He thought that wider explanation was not testable. I don’t know if he was right or wrong about that but Dykes gives no sign that he even understands the issue involved.

Section 12

A lot of the points in this section rehash stuff tackled above so I’m not going to bother addressing all of it. I will consider only a few points:

Popper’s philosophical premises also led him into more serious confusions. For example, he explicitly rejected as “utterly naïve and completely mistaken” what he called “the bucket theory of the mind” [OKN 61], the idea that “before we can know or say anything about the world, we must first have had perceptions – sense experiences” [OKN 341]. Yet earlier he had stated: “I readily admit that only observation can give us ‘knowledge concerning facts’, and that we can… become aware of facts only by observation” [LSCD 98].

This is a confusion on Dykes’ part not Popper’s. First, an observation is different from a sense perception. To find facts relevant to testing a theory we have to do observations, and we do this by developing conjectural knowledge about how to observe the relevant phenomenon and then using that knowledge to do the observations. Second, we do have knowledge about the world prior to observing it. Babies are born with the knowledge required to do certain things like breathe: they didn’t get that knowledge by observation.

Popper’s attitude to ‘the laws of nature’ was just as perplexing. In Open Society he called natural law a “a strict unvarying regularity…. A law of nature is unalterable; there are no exceptions to it…. laws of nature… can be neither broken nor enforced. They are beyond human control…” [OSE1 57-58, c.f. OKN 196]. But such absolutist claims are difficult to reconcile with the actual discovery of natural laws when, according to Popper: “There can be no valid reasoning from singular observation statements to universal laws of nature” [RASC 32, c.f. OKN 359].

This is not perplexing at all. The fact that laws of nature are beyond our control is one of the reasons why our knowledge of them is conjectural: we have to go out and find out what they are by a process of proposing and testing explanations.

Another point:

In like vein, Popper’s use of illustrations often involved disregard of his own dicta. In Realism and the Aim of Science, when once again attacking induction, he told us that “mere supporting instances are as a rule too cheap… they cannot carry any weight” [RASC 130]; and that, “confirming instances are not worth having” [RASC 256]. However, when he had earlier sought to demonstrate the case that “practically every… ‘chance observation’ is an example of the refutation of some conjecture or assumption or expectation,” he unhesitatingly drew attention to scientific discoveries by Pasteur, Roentgen, Crookes, Becquerel, Poincaré and Fleming to reinforce his point [RASC 40].

Dykes doesn’t seem to grasp the distinction between illustrating a point with concrete examples to help people understand it and using those examples as confirmation.

Another accusation from Dykes:

The trait of employing what he sought to deny can be found throughout Popper’s work. Take his critique of Plato’s politics. In Volume 1 of Open Society Popper went through the Republic, Laws, etc, with a sort of remorseless philosophical laser. Yet not once did he give any hint that he regarded the object of his study as conjectural. His method was purely and simply inductive. He took Plato’s dialogues as fact, examined them line by line in search of evidence, and generalised his (very firm) conclusions.

Popper said that all knowledge is conjecture several times in that book alone. And as for the idea that his method was inductive, if that were true, then none of the people who had previously read Plato would have been at all surprised by anything Popper had to say, nor would any of them have found it worthwhile since they would all have inductively reached the same conclusion having read the same book.

Dykes claims that Popper didn’t respond to a criticism:

Another failing was Popper’s occasional lack of response to important criticisms of his philosophy. As a critical rationalist, to whom criticism was “the lifeblood of all rational thought,” this was serious indeed. There was, for example, the incisive refutation of the falsification principle published by the famous American philosopher Brand Blanshard. Blanshard noted that particular propositions such as ‘some swans are white’ can only be falsified by showing that ‘no swans are white.’ Since the latter would be self-evidently untrue, ‘some swans are white’ is a perfectly valid scientific statement which cannot be falsified.

“Some swans are white” is equivalent to “some white swans exist”. These statements are not directly testable but they may be consequences of testable scientific theories and so may be indirectly tested. So to the extent that “some swans are white” is scientific it is because it is related to testable theories. See Popper’s reply to Maxwell which was referenced by Dykes but not refuted by him. Why Popper should offer yet another reply to the same old stuff that he refuted before is beyond me. If people aren’t willing to make any effort to think you can’t force them.

Dykes also criticises Popper’s leniency toward Marx. His criticisms on that point are correct. However, that is an argument against ignoring criticism, not against Popper’s epistemology, which implies that we shouldn’t ignore criticism.

Dykes’ criticisms of Popper seldom get close even to stating Popper’s position never mind criticising it successfully. If somebody wants to understand Popper’s epistemology the best account by Popper is Chapter I of “Realism and the Aim of Science”.

UPDATE: I have edited this essay slightly to make it clearer and eliminate points that did not refute the substance of Dykes’ position.