*  ESSAY *




Brill, 1993

"This book offers a straightforward account of Sir Karl Popper's views on scientific methodology ranging from Logik der Forschung in 1934 to A World of Propensities in 1990. Part I covers his treatment of the interrelations between metaphysics and science, the fallacies of induction, the method of conjectures and refutations, evolutionary epistemology, the propensity theory of probability, and the interpretation of quantum mechanics. Part II considers the problems of the social sciences, his critiques of historicism and holistic planning, his defence of piecemeal planning on both scientific and humanist grounds, his method of situational logic based on models that use a `rationality principle', and the roles of institutions, traditions and history. The book is addressed to those who are interested in general problems of scientific method but find it difficult to get a clear or connected view of Popper's important contributions because these have been published over long intervals and have been subject to misinterpretations."

Amazon Product Description

During the 1980s as a result of making contact with Bill Bartley I had the good fortune to discover that Colin Simkin, Popper’s lifelong friend from New Zealand, lived only a mile away. We met almost weekly for a decade or so until he died in 1998. During this time he exchanged many letters with Popper and he also made several visits to Britain where he met Popper and was even allowed to smoke while they walked together in the garden. During one of  these trips Popper died (not from lung cancer).

We discussed Popper’s ideas exhaustively, along with many other interests that we shared, not including sport. We also became concerned about the general lack of reception of Popper’s ideas, despite the release of the long-awaited Postscript to The Logic of Scientific Discovery in 1982 and 1983.

We were especially concerned with the state of economics and the social sciences where Popper’s ideas had made even less impact than was the case in the natural sciences, where many leading figures had heartily endorsed his contribution. The same could not be said in economics or sociology or the humanities.

Colin conceived the idea of a collection of Popper readings in the social sciences, following the example of the Pocket Popper produced by David Miller with extracts and short forms of his most important papers. He identified suitable readings in four categories and wrote an introduction to each section. One of the sections was on metaphysical research programs, a major interest of mine, and Colin invited me to write the first draft of the introduction to that section.

After some consultation, possibly including David Miller, Colin decided that this plan was not feasible due to the difficulty of getting the permissions from several publishers to print the extracts. So he decided to write a book himself, with a general introduction to the body of Popper’s work in epistemology and methodology followed by several chapters crafted especially for economists and other social scientists.

After the usual trials and tribulations of publishing, the volume appeared from Brill in 1993.

It is available from Amazon, regrettably priced over $100 which will presumably ensure that many people who would benefit will not buy it (the same applies to another recent introduction to Popper by Phil Parvin). Joe Agassi wrote a very positive review, with some critical comments, as befits a Popperian reviewer!

Some commentary on this excellent book, with extracts, may encourage people to chase it up.



1. Metaphysics and Science 11
2. Growth of Knowledge 21
3. Deductive Knowledge 27
4. Justification 31
5. Pragmatism 38
6. Inductive Probability 44
7. The Conjectural Method 51
8. Objectivity and Truth 58
9. Causal Laws, Probabilistic Laws and Models 64
10. A World of Propensities 70
11. Metaphysical Research Programmes 78
12. Evolutionary Epistemology 84
13. A Case for Indeterminism 92
14. Critique of Quantum Mechanics 99

15. Critique of Historicist Views 109
16. Holistic Planning 117
17. Situational Logic 122
18. Piecemeal Planning 132
19. Individualism 138
20. Models and Individualism 144
21. Institutions and Traditions 149
22. The Role of History 155
23. An Application to Economics 160
24. Critical Economists 165
25. Relevance to Economics 172
Conclusion 180
I. The Birth of The Open Society: a Personal Reminiscence 183
II. Popper and Hayek on Piecemeal Social Engineering 191
Ill. Advice to Russian Readers of The Open Society 198


In this book I try to give a straightforward, connected and up-to-date account of Sir Karl Popper's leading ideas about scientific method, paying special attention to their relevance to social theory. That emphasis is natural for one who is an economist rather than
a philosopher.

My reason for writing it arises from a conviction that his original ideas are still not widely understood, still less properly appreciated, in spite of his presentations of them, over almost six decades, in many articles and books that are remarkable for vigour and clarity of writing. One might expect that his fellow philosophers would have read them properly but that is seldom the case, as becomes evident from reading his 'Replies to My Critics' in The Philosophy of Karl Popper; see, for example, his remarks on 'the Popper legend'. I share the impression of Bartley and Medawar that opinions about Popper's methodology have too often been formed not so much from his own writings as from incomplete and partial expositions of his ideas by critics or revisionists.

There has been neglect rather than misunderstanding of Popper's work by natural scientists, although he has been primarily interested in the growth of scientific knowledge in physics and biology. Bartley laments the lack of fruitful dialogue between Popper and physicists, notwithstanding some notable exceptions, which include Einstein, Schrôdinger, Bondi and Margenau.

Popper has had more luck with biologists, including the physiologist Eccles and the medical scientist Medawar, both Nobel prize winners, the biochemist Wãchtershãuser, and the psychologist Campbell. His general contribution to scientific thinking, moreover, has been acknowledged by election to the Royal Society, as well as to a number of  foreign academies of science, and by the award of prestigious international prizes.

Nor have his ideas been adequately appreciated or understood by social theorists. His main concern, admittedly, has been with physics and biology, and he has confessed that the only social science which had ever interested him was economics. But his general work on methodology has considerable relevance to the problems of social analysis, and he has written two important books which deal exclusively with social science. The Open Society and Its Enemies was first published in 1945 and The Poverty of Historicism in 1957, following its appearance as three articles in Economica, 1944-45. A few articles bearing on social science were published later, the last in 1967, after which he appears to have lost interest in its problems.

Popper's first book, the Logik der Forschung, had appeared in 1935 and immediately attracted a good deal of attention, so that he received, and accepted, foreign invitations to speak about it. A few economists also read it. Terence Hutchison did while studying at Bonn, and used it for his own first book, The Significance and Basic Postulates of Economic Theory (1938). Gottfried Haberler also read it, at Harvard, and recommended it to Hayek, who then invited Popper to address his seminar at the London School of Economics. There he spoke formally about methodological problems in social science. This lecture was developed into The Poverty of Historicism in New Zealand, where Popper taught from 1937 to 1945. Influenced by the Logik and by this lecture, Hayek wrote an important article, 'Economics and Knowledge' in Economica (1937), which Hutchison claimed, perhaps exaggeratedly, is a vital turning point in Hayek's ideas about epistemology.

In New Zealand, Popper also wrote The Open Society, and it was published in England shortly before his arrival at the LSE to take up a teaching post which he held until his retirement in 1969. This book came out a year after Hayek's The Road to Serfdom, and both were widely read and discussed. Independently written, they attracted attention because of different yet complementary exposures of the intellectual roots and direful consequences of totalitarian influences, powerful even after the collapse of Nazi Germany and Fascist Italy.

Social thinkers thus became aware of Popper's emphatic rejection of ideas that there are inevitable historical laws which determine social developments, which could help us to predict such developments, and which it is the task of social science to discover. They would also have become aware of his advocacy of 'piecemeal social engineering' as against holistic or collectivist central planning.

But his constructive attacks on historicism and collectivism were strongly resented by various Platonists, Hegelians, Marxists, sociologists of knowledge, and wholesale planners, so that there were polemical arguments and misunderstandings which have long obscured or distorted the transmission of his ideas. Similar troubles arose from his vigorous exposures of the poverty of the logical positivist and linguistic schools of philosophy, then dominant and especially in Britain.

It was not until 1959 that The Logic of Scientfic Discovery was published as an English translation and extension of the Logik der Forschung. The twenty-four year delay was unfortunate in that it gave time for much indirect and garbled reporting of Popper's basic ideas among those who could not read German or obtain a copy of the now very scarce Logik. Still more unfortunate was an even longer delay in publishing a most important Postscript to the Logic of Scientific Discovery. This had been sent to a publisher in 1956, but serious eye trouble prevented Popper from completing proof reading. The galley proofs circulated among his colleagues at the LSE, who made some use of them; but the
Postscript did not appear until Bartley published it as three volumes. The Open Universe and Quantum Theory and the Schism in Physics both appeared in 1982, and Realism and the Aim of Science a year later. They surpass even the Logic in philosophical or scientific interest, and include an exposition of his original idea about metaphysical research programmes. But they came much too late to prevent widespread acceptance of misleading versions or critiques of his thought.

Between the Logic and the Postscript Popper published three other books. Conjectures and Refutations (1963) and Objective Knowledge (1972) are collections of essays, written at various dates and containing further developments of his epistemological and methodological themes. Of particular interest to social theorists are articles in the former on tradition (Ch. 4), social prediction (Ch. 16), liberalism (Ch. 17), and humanism and reason (Ch. 20). The last chapter of Objective Knowledge is a 'Realist View of Logic, Physics and History'. In 1977 he collaborated with Sir John  Eccles, the eminent physiologist, in writing The Self and Its Brain, which thoroughly explores the body-mind problem, an old puzzle in philosophy and one relevant to psychology. Finally there is a paper which Popper read to Haberler's seminar in Harvard during 1963, 'Models, Instruments and Truth', a condensation of which was prepared by a colleague, Alan Musgrave, and published later in a book of essays honouring Jacques Rueff. An English translation of this French article appears under the title of 'The Rationality Principle' in David Miller's excellent book of selections, A Pocket Popper. It is an illuminating and suggestive account of the kind of scientific theorizing that does not use specific causal laws.

In the last few years Popper has taken, and given addresses on, what he calls 'a new view of causality' based on his propensity theory of probability. It has led him to the metaphysical view of an indeterminate and evolving cosmos, subject generally to only 'weak causality'. These ideas about probability and indeterminacy, first developed in connection with physics, have obvious relevance to biological and social sciences, relying as they do on statistical methods.

All this provides a wide range of insights for social theorists, yet they have seldom shown much comprehension of the earlier works, as L.A. Boland has often observed, and they have hardly noticed the later ones. Mark Blaug, for example, in his The Methodology of Economics, devotes seventeen pages to a sympathetic account of Popper's views on falsification, induction and corroboration but ignores other contributions, more relevant to social theory, such as historicism, piecemeal engineering, theoretical models and metaphysical research programmes. Nor, in discussing other methodological positions, does he notice Popper's more effective criticisms of them. Blaug also makes the revealing claim that although Popper has had a great influence on modern economists few of them have read him, but gained such understanding as they have of his ideas indirectly and from Milton Friedman's Essays in Positive Economics. If so, they have gained it from one who shows imperfect knowledge, or else limited acceptance, of Popper's  methodology.

Bruce Caldwell's Beyond Positivism (1982) provides another example. He devotes more space to Popper's ideas than Blaug does, but covers no more of them. His understanding of them is not enough to stop him from advocating a confused 'methodological pluralism' that denies the objectivity of economic analysis. Neither does his discussion of Kuhn and Lakatos notice Popper's criticisms of their arguments, nor recognize that Lakatos' 'scientific research programmes' are a distorted derivation from Popper's metaphysical research programmes. But then his extensive bibliography did not list The Philosophy of Karl Popper, which he showed no sign of having read.

[Note added in proofs] Since then he appears to have read it as it appears in the bibliography to his article 'Clarifying Popper', which was published in the Journal of Economic Literature in March 1991. In this article he claims that, following the Nafplion conference sponsored by the Latsis Foundation in 1974, Friedman's instrumentalism dropped out of discussions about economic methodology and interest developed in Lakatos' scientific research programmes, especially in those aspects which most separate his thought from Popper's, Blaug playing a major role in this development (pp. 10-12).

Not that these two examples are by any means the worst. They are given only to illustrate the need for a better and fuller explanation of Popper's ideas to social scientists. A further reason for the need is that these ideas and their developments are spread over a range of writings separated by wide time intervals, and are sometimes partly obscured by distracting polemical or historical material.

It has seemed worthwhile, therefore, to attempt bringing them together in a systematic way and as simply as accuracy permits. Some attention is paid to predecessors who had similar ideas, or who influenced Popper's thinking, positively or negatively.

Part One offers an account of his writings on general scientific method, concluding with a fuller account of his pioneering work on evolutionary epistemology, his view of  indeterminism, and his important critique of quantum theory which has been the major influence on his views about probability and indeterminism.

Part Two covers what he has written about the application of his general methodology to social theory, and about the distinctive problems of analyzing social phenomena. It pays more attention to anticipations and criticisms of Popper's ideas, although still trying not to let their discussion become unduly distracting. The penultimate section considers criticisms that economists have made of Popper's ideas in a recent seminar which was held exclusively for that purpose, and it concludes with my own evaluation of what economists may learn from Popper. Two appendices have been added about the genesis of The Open Society and about the relation between his and Hayek's views on 'piecemeal social engineering'; the first is an article which I wrote for the Australian quarterly, Quadrant, and the second is a shortened version of a paper given to the Christchurch Meeting of the Mont Pelerin Society in 1989 and later also published in Quadrant.

There is also a third appendix summarising advice Popper has recently given to readers of the Russian edition of The Open Society (to be published in 1993).

The main results of the book might be very summarily indicated by twelve theses.

(i) Science has developed from metaphysics and has become increasingly, different from it by putting theories into a logical form that allows them to be empirically tested. Yet metaphysical elements can never be completely purged from scientific theories, and some metaphysical ideas have often usefully guided scientific research.

(ii) There is neither a deductive nor an inductive path to scientific understanding of phenomena, natural or social.

(iii) The only sound way towards such an understanding is by bold conjectures about problem situations, and severe testing of these conjectures, logically and empirically.

(iv) Scientific theories are thus always provisional, liable to replacement by more informative theories which survive, for a while, rigorous tests.

(v) It is exceptional for these theories to be exact causal laws; they are rather probabilistic.

(vi) Probability is not a reflection of human ignorance but a propensity of objective situations to generate frequency distributions of events generated by those situations.

(vii) The universe is not fully deterministic but is evolving to create new situations and, in that sense, continually opens up new possibilities and so changes propensities.

(viii) Social scientists have no hope of finding historical laws of development, nor of providing any rational basis for comprehensive social planning.

(ix) They should give up attempts to emulate the physical sciences by searching for timeless causal laws.

(x) If there are social laws, these must be probabilistic, but they cannot be established by any appeal to so-called inductive probability.

(xi) Nevertheless we can reach scientific explanations of social phenomena by using models of social situations together with a very weak rationality principle which avoids the ambiguities of psychological theorising.

(xii) Such explanations can be greatly helped by piecemeal social engineering that addresses practical social problems in a scientific way.

Although conscious of personal deficiencies for attempting this kind of exposition, I have had the advantage of a close personal friendship with Karl Popper since 1939 and, particularly in more recent years, have had valuable discussions with him on most of the problems tackled here. I am also grateful to a younger friend, Rafe Champion, whose great interest in, and understanding of, Popper's work have been very helpful to me throughout the writing of this book.


Simkin pointed out  that Karl Popper's strongest intellectual interests follow those of the pre-Socratic Greeks who he admired to much for their invention of the critical method and their interest in cosmology.

“They offered, instead, disinterested arguments that were based solely on objective experience. Their daringly imaginative speculations about the cosmos and life thus pointed the way to scientific thinking, although they gave no laws for predicting future events and so fell short of modern scientific standards. Nevertheless they foreshadowed some of the ideas of Newton, Dalton and Darwin, and Popper has admired the 'simple straightforward For example, Anaximander's idea that the earth is suspended in space because he speculated that it was equidistant from celestial bodies which whirled in circles around it; Democritus' view of endless change from the aggregation and disaggregation of infinite, indivisible and indestructible atoms; and Anaximander's conclusion that man must have descended from fish through animals because his long infancy would not have allowed survival otherwise.”

Cosmology and epistemology can be considered as parts of metaphysics, that is, of attempts to explore reality and explain it in an intellectually satisfying way. However at least since the time of Hume there has been a huge bias against metaphysics in the main stream of western empiricism which found its full flowering in the logical positivists of the 20th century.  Another trend was the apparent shrinking of the subject matter of philosophy which began as the science of everything (natural philosophy) but lost one science after another, as Simkin wrote:

“As knowledge grew, one science after another split off from metaphysics, first astronomy and mechanics then, chemistry, geology, biology, medicine and psychology, leaving philosophy with the task of combining scientific results with cosmological speculation in the hope of obtaining a coherent intellectual framework for understanding the universe and for further research into its mysteries.”

However some sociological aspects of the philosophical profession distracted most philosophers from working on the framework. Most British philosophers came to the topic from the study of classics and ancient languages (the Greats of Oxbridge) and were intimidated by the demands of science, especially in the 20th century. It was too tempting for them to follow the trend to linguistic philosophy in its various forms and leave the philosophy of science to be dominated by anti-metaphysical positivists.

This chapter shows how Popper was too interested in science and cosmology to follow either the linguistic or the positivist roads. Simulated by Agassi’s work on metaphysics and the importance of metaphysical programs as the framework of scientific investigation, Popper considered Kant's problem of distinguishing science from metaphysics. Simkin explains:

“The problem arose from Hume's dictum that, because a metaphysical theory is devoid of mathematical reasoning or of experimental reasoning about fact, 'it can contain nothing but sophistry and  confusion'. Kant, however, held that there was another class of meaningful statements or judgements than the analytic and empirical ones stressed by Hume. This was the class of synthetic a priori statements which are neither tautological nor dependent on sense experience, in particular, the presuppositions of traditional logic, arithmetic, Euclidean geometry and Newtonian physics. They included what Kant called the 'principle of universal causation'. All, he held, are involved in our thinking about the world, and especially in scientific thinking.”

As Bryan Magee pointed out, Popper emerged from his study of Kant as a “modified Kantian”, or a “conjectural apriorist”, with the doctrine that we create theories about the world which are in advance of experience but can be tested against it. Hence the importance of formulating theories in a form that can be tested and the importance of methodological conventions to promote critical discussion of theories, including tests.

“They must, that is, be falsifiable in the sense of having logical consequences which could be contradicted by factual statements about experiments or observations. This had become clear to Popper after hearing a lecture by Einstein in which he listed some possible experimental results that could falsify his new theory of relativity. Falsification of a theory occurs if scientists come to accept the validity of statements which contradict some of its consequences. While falsifiability is a purely logical requirement, falsification is always problematic because there can never be a logical compulsion to accept the truth of a falsifying statement; it can be questioned or even denied without involving any purely logical error.”

Simkin noted some of the less helpful criticisms of Popper’s views on testing and demarcation and then proceeded to a sketch of one of Popper’s most interesting and little appreciated innovations, the theory of  objective knowledge.

Objective Knowledge and the “Three Worlds”

“Popper distinguishes different aspects of the reality which these theories try to explain by speaking of three 'worlds'. World 1 comprises physical things, states or processes; it has been the earliest and main concern of philosophical cosmology and later of the natural sciences. World 2 comprises mental states and behavioural dispositions; it has always been the main concern of epistemology and, of course, of the psychological sciences which sprang from epistemology.”

“Nearly all epistemology, ancient and modern, has failed to distinguish clearly between subjective and objective knowledge, and most of it has concentrated on subjective knowledge, and so on World 2. Descartes, for example, took knowledge to be an activity that requires a knowing subject, and the British Empiricists, beginning with Locke and Berkeley, took sense experiences to be the basis of all our knowledge. Others, such as Plato, Aristotle and Kant, found an important role for intuition, the ability of the mind to grasp basic aspects of reality. Popper's epistemology is very different. It focuses on his World 3 of objective contents of thought, 'the content of scientific and poetic thoughts and great works of art'. A book, for example, is itself a physical object and so part of World 1; what puts it into World 3 is its content, which remains the same through various printings and editions. As a World 3 object the content of the book may induce some people to produce other books and thus to act upon World 1 which is a compelling argument for regarding the content as real. Even more compelling is the argument that a scientific theory can lead to great changes in World 1; for example, the theory of heat led to railways and steamships, and quantum mechanics led to the atomic and hydrogen bombs.”

Simkin provides a very helpful summary of Popper’s arguments in favour of the autonomy of the third world, for example in number theory and physics. Another good introduction to these ideas about the third world can be found here, provided by Dr Roger James in another excellent introduction to Popper’s ideas. 


This is a short chapter on the relationship between subjective (personal) knowledge and objective (public) knowledge, starting with Popper’s critique of  the ideas of Kulpe and Buhler which were so important in saving him from positivism.

“Popper early came to see (while studying educational psychology at a teachers' training institute) that the growth of knowledge had to relate primarily to objective knowledge, because of the 'priority of the study of logic over the study of subjective thought processes'.' He came to this view by discerning two mistakes in logic which had led to mistakes in psychology. The first mistake was made by Kulpe who, through confusing logical and material implication, held that arguments were complex judgements and thus denied a difference between arguing and judging. The second mistake was made by his teacher, Bühler, who had not seen that the descriptive function of language could be distinguished from its argumentative function. Subjective knowledge, moreover, develops as we try to deal with objective problem situations, and it throws less light upon the growth of objective knowledge than this throws upon the growth of subjective knowledge. As Popper put it: 'An objectivist epistemology which studies the third world can help to throw an immense amount of light upon the second world of subjective consciousness, especially upon the subjectivist thought processes of scientists; but the converse is not true.'  As bodies we are part of World 1, and as minds part of World 2, but not the whole of it because Popper's theory of 'evolutionary epistemology' finds a place there for Other forms of life. All organisms, he thinks, even plants, are born with certain innate dispositions and expectations arising from their biochemical make up.”

Popper drew comparisons between the innate expectations of animals which may be modified by (bitter) experience and the theoretical expectations of scientists which can be systematically tested.

“Popper does not believe that intuition is an infallible or certain guide to truth, nor that scientific knowledge, as distinct from mathematical knowledge, can grow by a process of pure deduction. Nor does he believe that there is anything direct, immediate or certain in our sense experiences. Learning about ourselves and the world around us 'is all decoding, or interpretation', and that does not function faultlessly, especially during a learning process or if unusual situations arise.”

Simkin unveils Popper’s four-stage scheme to explain the problem solving activities of all organisms from the amoeba to Einstein.

“Scientific knowledge is objective knowledge par excellence, so that Popper takes it to be the clearest case for studying growth of knowledge. The starting point is always a  problem, one which arises as we try to understand or use the world better by finding a satisfactory explanation for some group of phenomena that interests us. A theory about these phenomena is proposed tentatively, and it is then tested in various ways. The testing may well lead to the discovery, often unexpected, of errors, new facts or other difficulties in the tentative explanation. A different problem situation then arises, and is dealt with, sooner or later, by proposing a modification to the original theory or else a new one. There will then be further testing, discovery of new difficulties, the conjecture of still another theory to deal with them, and so the process goes on.”

P   ->  TS  ->  EE ->  P

Here P is a problem, TS a tentative solution, and EE error elimination through testing, testing the theory's logical consistency, its explanatory power as against that of rival theories (particularly in regard to novel applications), and its correspondence with fact. Not only can the process lead to new and better theories, it also promotes the discovery of new facts and new problems. Knowledge, accordingly, is doubly enlarged.

The creative function of criticism is not a point made strongly by Popper or Simkin but it adds weight to the critical method by adding a constructive dimension to the more common perception of error-elimination by criticism and testing.


"I hoped sooner or later to arrive at a perfected mathematics which should leave no room for doubts, and bit by bit to extend the sphere of certainty from mathematics to other sciences."
Bertrand Russell 

This is a very short chapter (only 3 pages) pointing out the limits of the axiomatic approach outside logic and mathematics.

“Hilbert succeeded in giving an axiomatic formulation, not only to a wider geometry than Euclid's, but also to some branches of theoretical physics. Woodger attempted the same thing for biology. In the social sciences, Debreu has won a Nobel Prize in Economics for work which most notably included an axiomatic analysis of economic equilibrium.”

“Popper recognizes that putting a scientific theory into axiomatic form facilitates severe testing, and has the further merit of making every new assumption, especially an ad hoc one, recognizable as a change to the system. But he points out that, because scientific knowledge is always changing, and may well require changes in associated theories, only some branches of science can be given an axiomatic formulation, and then only temporarily.”

Popper also pointed out the danger of excessive formalism (at the expense of imaginative criticism) and the need to avoid striving to achieve a degree of precision that is in excess of the requirements at the time.

Godel handed out a final blow to the hopes of the most optimistic adherents to formalization and axiomatisation.

“The hopes of Russell, Hilbert and others for certainty in mathematics were dashed in 1931 by Gôdel's demonstration that, within a deductive system which can include arithmetic, there must be some statements that are undecidable in the sense that they can neither be proved nor disproved. Nevertheless much can be, and is, proved within a deductive science, and Gödel's result is wrongly interpreted if it is taken to justify relativist claims that truth and objective knowledge are illusory.”

“Inescapable as that conclusion is, it has been strenuously resisted on the ground that it undermines science, if not reason itself. It seems to leave science without any firm logical basis, a situation which is unacceptable to those philosophers and others who have sought justification for their beliefs in the hope of reaching some certainty about a puzzling world. The next section, therefore, takes up this question of justification, which is the supposed basis for many and repetitive attacks on Popper's methodological views.”


"To justify the original conclusions one must stop at something not open to question for which one does not have to provide satisfactory reasons."
W.W. Bartley III

“The commonsense theory of knowledge is subjectivist. It regards knowledge as a state of mind about a belief or an opinion which, in everyday life, guides our actions; and it seeks justification or sufficient reasons for the certain truth of such beliefs. Popper has no quarrel with the ideas that subjective knowledge is a state of mind, and that we [often] have feelings of subjective certainty.”

“Popper, however, strongly dissents from the associated view that we could ever find reasons for certain belief in regard to objective or demonstrable knowledge, beyond that part of it which is confined to logic and mathematics. The commonsense theory has ignored World 3 and, worse, fails to see that scientific knowledge, the most important part of objective knowledge, is thoroughly conjectural.”

Simkin referred to Aristotle, the inventor of syllogistic logic, and his belief in the ability of the human mind to reach distinct ideas about the basic nature of things (the essences of reality), using intuition, after reason and experience had cleared the way.

The Rationalists or Intellectualists who followed Descartes also put their faith in the clear and distinct perception of ideas by intellectual intuition.

“A very different variant of justificationism is associated with the British school of empiricist philosophers, notably Bacon, Locke, Berkeley and Hume. Rejecting the Cartesian idea of 'natural light', they argued that our knowledge of reality can derive only from the observations of our senses. These thinkers appealed, that is, not to the authority of revelation or reason, given by God, but to the trustworthiness of our senses, given by Nature, for justification of belief. Yet Hume's influence was subversive. He argued that sense impressions, although the only basis for our knowledge about the world, cannot provide a sure basis; they can only throw light on the psychology of our beliefs. Empiricism was thus reduced to solipsism - the theory that nothing exists independently of the person having sensory experiences.”

The problem with all forms of justificationism is the infinite regress or the choice of a stopping place to avoid the regress.

“Statements grasped by intuition, or 'the faculty of reason', are the halting place for essentialists and intellectualists. Intuition, no doubt, has its mysterious uses, particularly in inventing theories, but it can never be a sure guide to truth. For there can be no guarantee that different people's intuitions will coincide, nor that these (or commonsense) will remain constant in the light of new experiences or discoveries. We are then faced with the need to choose between different intuitions, and that requires critical examination of their consequences, not further justificatory intuitions which would present the same problem of choice.”

On the empiricists side the problem of justification came down to the problem of induction, the justification of beliefs on the basis of empirical evidence and the problem raised by Hume (and later by Duhem).

“This method came to be known as induction, a word derived from the Latin translation of Aristotle's term epagoge, meaning reasons for belief that fall short of proof, but now used to mean reasoning from particular observations to general statements or theories. It did not have quite that meaning for Bacon but rather an appeal to facts instead of authority or intuition for knowledge about empirical phenomena.”

“The great influence of induction, and also its basic difficulty, have been strikingly expressed by Russell and Whitehead. Russell wrote: 'Induction is an independent logical principle, incapable of being inferred from other logical principles, and without this principle science would be impossible.' Whitehead has similarly written: ‘The theory of induction is the despair of philosophy- and yet all our activities are based on it’ “

It was William Bartley took  up Popper's ideas about justification in his theory of  'pancritical rationalism' with a considerable program to unpack the consequences of “non-justificationism”. He held that criticism has to be completely dissociated from justification, and that almost all traditional and modern philosophies have erred by fusing these two ideas in a search for justification of criticism, thus making some form of authoritarianism inescapable.

Popper was persuaded by these arguments to restate his position about rationality. In Realism and the Aim of Science he wrote:

“Bartley observes that my approach has usually been mistaken for some form of justificationism, though in fact it is totally different from it. For even though I offer a negative solution to the classical problem of justification, resembling in this respect the sceptics and irrationalists, at the same time I dethrone the classical problem and replace it by a new central problem which allows of a solution that is neither sceptical not irrationalist. For my proposed solution to the new problem is compatible with the view that our knowledge -our conjectural knowledge-may grow, and that it may do so by the use of reason: of critical argument.” 

It is helpful to make a distinction between justification in the strong sense demanded by justificationists and the formation of a critical preference for one or other among rival theories. It is possible to justify a preference, with the understanding the new evidence or arguments could alter the position, without any problem of regress or an arbitrary stopping point.


This is a very short chapter noting the different responses of conventionalists, instrumentalists and pragmatists to the problems of induction and justification.

“Instrumentalist and conventionalist ideas, spreading beyond physics, have even reached economics, largely through the writings of Milton Friedman. Taking truth to be somehow divisible, he has argued that the truth or falsity of a theory's assumptions do not affect its scientific worth. That depends only on its usefulness in yielding accurate predictions-predictions that could conceivably be wrong but which have not yet been proved to be wrong. Its usefulness, of course, could extend to policy or other applications of a theory because these would depend upon its predictive power. To this extent Friedman has been judged an instrumentalist. But he is, at the same time, a conventionalist in making preference between rival theories that appear to give equally good predictions depend upon comparative simplicity.”

The American school of pragmatists includes William James, John Dewey and the great Charles Peirce. Simkin pointed out that Peirce can be seen as a forerunner of Popper, describing scientific method as matter of critical discussion and experimental testing, without hope of certainty. He used the term 'fallibilism'. He also became unhappy with the popular interpretation of “pragmatism” and decided that his own philosophy should be called “pragmaticism”.

From Wikipedia: Pragmaticism is a term used by Charles Sanders Peirce for his pragmatic philosophy starting in 1905, in order to distance himself and it from pragmatism, the original name, which had been used in a manner he did not approve of in the "literary journals". Peirce in 1905 announced his coinage "pragmaticism", saying that it was "ugly enough to be safe from kidnappers". Peirce's pragmatism, that is, pragmaticism, differed in Peirce's view from other pragmatisms by its commitments to the spirit of strict logic, the immutability of truth, the reality of infinity, and the difference between (1) actively willing to control thought, to doubt, to weigh reasons, and (2) willing not to exert the will, willing to believe. In his view his pragmatism is, strictly speaking, not itself a whole philosophy, but instead a general method for the clarification of ideas.


"From the time of Laplace onward, various attempts have been made to show that the probable truth of an inductive inference follows from the mathematical theory of probability."
Bertrand Russell 

“Inductivists were perturbed by Hume's devastating critique but they could see no better way of justifying empirical theory, always for them a major but wrong-headed concern. They tried, therefore, to buttress the inductive method in one way or another. The great mathematician, Laplace, had two ideas which seemed helpful for this purpose. One was a bold leap from Newton's mechanics to the assertion, not altogether original, of universal causation, and the other was an appeal to the theory of probability, which he was doing so much to develop.”

J.S. Mill's incorporated the principle of the uniformity of nature in his very influential System of Logic (1843) but he was forced to resort to a circular argument to justify the principle on the basis of experience.

J.M. Keynes, who made his name as an economist, earlier made a serious effort to give induction a probabilistic basis in A Treatise on Probability. He proposed to apply probability apply to propositions, as well as to events, so that an inductively justifiable conclusion would have a high probability which might even approach unity if it became practically certain.

However the problem persisted and Russell wrote that 'Induction…is not quite the universal principle that we need to justify scientific inference. But we certainly need some universal proposition or propositions. . . And whatever these principles of inference may be, they certainly cannot be logically deduced from facts of experience'.

When Popper began his own study of the problem, Reichenbach's theory was being discussed among the logical positivists. His theory was based on the proposition that, for a large number of occurrences of two phenomena, the relative frequency of their joint occurrence becomes constant, and this relative frequency is the probability of their joint occurrence. He also proposed that probability, so defined, should take the place of truth in logic.

Simkin wrote that his treatment of induction is a “very incomplete sketch of attempts to make something of the idea of inductive probability. No apology is made for that because the attempts were wasted intellectual effort. For, just as Hume had exploded the idea of induction by enumeration, so Popper has exploded that of inductive or ampliative probability.”

It is possible that the notion of objective probability has been given up in the form that Carnap and others were seeking and nowadays there is a vogue of Bayesian subjective probability theory that has become the latest effort to save justificationism along empirical and probabilistic lines.


"To use one's eyes as far as they can be used, and then to guess the way forward - the art of science is no more than that; but it is quite enough, nevertheless, because the point is to guess right, that is to say, so that the result may be tested, and confirmed by experience."
Knut Wicksell

"We know that science cannot grow out of empiricism alone that in the constructions of science we need to use free invention which only a posteriori can be confronted with experience as to its usefulness. This fact could elude previous generations, to whom theoretical creation seemed to grow inductively out of empiricisnz without the creative influence of a free construction of concepts. The more primitive the status of science is the more readily can the scientists live under the illusion that he is a pure empiricist. In the nineteenth century many still believed that Newton's fundamental rule 'hypotheses non fingo' should underlie all healthy natural science."
Albert Einstein

In this chapter Simkin pointed out that Popper's rejection of all forms of induction did not lead him into scepticism about our ability to reach towards objective truth through critical reasoning. Simkin recalled the four-stage problem-solving schema which Popper used to demonstrate that all forms of life have some ability to improve understanding of their environments, to explore it and to learn through trial and error. Humans are greatly helped in this process by the development of language and writing.

Popper used to enjoy telling his students that there is no such thing as Scientific Method, reminding them that he was therefore the Professor of a nonexistent subject (in his chair of Logic and Scientific Method). By “method” he meant a formula or algorithm or a process that would guarantee success. But there is no such thing.

The process of scientific discovery is the process of solving problems by trial and error, with no guarantee of success. The logic of scientific discovery (better called the logic of scientific investigation which is the appropriate translation of Der Logik der Forschung) is the logic of testing. And the key logical problem here is the relation between singular statements of fact and the universal statement of a theory.

Popper addressed the question: what can we say about a general theory on the basis of true singular statements? The aim of the positivists was to verify universal laws by reference to observational statements, but that does not work (due to the black swan effect). Resort to probability does not work either.

Popper’s response was to play up the significance of the well-known principle that a single true statement (regarding a black swan) refutes the law that all swans are white.

He did not fall into the trap of “naïve falsificationism” that is, the claim that a law can be decisively refuted/falsified by observations. The point is that adverse tests render a theory problematic, so more work is required to check whether the refutation can be repeated in other laboratories or whether more work is required on the theory.


"But as for certain truth, no man has known it
Nor will he know it; neither of the gods
Nor yet of all the things of which I speak."

Popper’s rejection of induction and other forms of  justification raises some questions from people who think that the objectivity of science depends on some kind of assured access to the Truth and the capacity to transcend normal human frailties and emotional commitments in apprehending it.

How does Popper's methodological analysis fit in with the idea that science and scientific knowledge are objective?

Does not its stress on imagination for conjectures, on judgement in regard to refutations, and on the continuous argument of critical discussion point rather towards subjectivism?

Can it be properly described as rational if it denies that empirical or any other type of knowledge is ever certain or even probabilistic?

In this short chapter Simkin explains that the objectivity of science, such as it is, does not depend on special qualities of scientists as much as on the existence of public debate and various institutional vehicles to conduct the debate.

Recall Popper’s views on objective “world 3” knowledge.

As for the truth, Tarski rehabilitated Popper’s confidence in the correspondence theory of truth after he wrote Logik der Forschung without mention of truth at all.

“For Popper, the objective character of science is different from what it is often taken to be by those seeking certainty or justification, and it has a dual aspect. The first aspect is that scientific theories, although springing from individual minds, acquire an independent and objective existence once they have been published and so become open to critical examination and testing by other scientists; i.e. they become part of what he has called an objective World 3. The second aspect is that the testing of a theory has be repeatable by other scientists before statements which report tests can be accepted as true; testing, that is, acquires an objective character by being public or intersubjective. Scientific knowledge thus belongs to World 3 and is objective because, for a metaphysical realist like Popper, World 3 is itself part of objective reality. A distinction has to be made between knowledge and knowing, between thought and thinking, the one being objective and the other subjective. The objectivity of science depends, not on the impartiality or subjective processes of individual scientists, but on the social and public character of science as a human activity-on its traditions and institutions for the dissemination and critical discussion of knowledge, theories and tests. The rationality of science, similarly, is not the same thing as the rationality of individual scientists, which may sometimes be questionable. It depends on the tradition of free, open and critical discussion of scientific problems, together with a general readiness to accept criticism and to learn from mistakes.”


“The difficulties arose from semantical and logical paradoxes which seemed to involve self-contradiction, emptiness or redundancy. They were cleared by Tarski's demonstration that, in order to speak of correspondence to fact, we need a 'metalanguage' in which we can speak of both statements and the facts to which these refer. Thus the statement 'she cooked the dinner' is true if, and only if, she did in fact cook the dinner. Here English is a metalanguage for the whole sentence and also an object language for 'she cooked the dinner'; if we put this statement into French that would become the object language and nothing would be altered thereby. Tarski's theory of truth, it may be noted, has nothing to do with ideas of true belief; it is quite objective.”


"The only connection or relation of objects which can lead us beyond the immediate impressions of our memory and senses, is that of cause and effect."
David Hume

Popper gave an account of causal laws in the Logic of Scientfic Discovery (Ch. III), focussed on the link between a set of initial conditions (the cause) and a deductive consequence (the effect). This became known as the  Popper-Hempel model of  explanation. Popper also proposed a methodological rule which corresponded to the principle of causality; never to abandon the search for causal explanation.

“Exact causal laws have been taken to be the hallmark of an advanced science, such as classical mechanics which, indeed, is their origin. Russell, however, pointed out that simple causal laws belong mostly to an early stage of a science because, as it becomes well developed into a comprehensive theoretical system, the initial conditions for explaining any effect become much too complex for any simple statement about cause and effect. Hayek, indeed, would limit causal explanation to relations between only a few phenomena, and so sees no place for it in analysing the complex problems of social science.”

Popper’s views changed over time although I am not aware of any clear account of the stages in the move towards explanation by propensities. Simkin wrote:

“He has, indeed, moved more and more to the view that strict causal laws are limiting, and rather rare, cases of scientific explanations which have the general form of probability statements. An exact causal law, asserts that predictions made from it, together with all relevant initial conditions, hold with certainty; i.e. with probabilities equal to 1. For probabilistic laws the stated probabilities of predictions are obviously less than 1.”


"I now regard the analysis of causal explanation in section 12 of L.d.F. (and therefore also the remarks in the Poverty and other places) as superseded by an analysis based on my propensity interpretation of probability."
Karl Popper

Simkin describes how Popper came to probability theory after Heisenberg announced the  famous uncertainty principle, giving impetus to the subjectivist view that something does not exist unless it is observed. Schrodinger showed the absurdity of the subjectivist view with his analogy of the cat in the box which may be alive or dead.

“Schrodinger developed a wave mechanics and showed that it gave the same results as Heisenberg's particle theory. Born unified the two theories by showing that the amplitude in Schrôdinger's wave equation could be used to give the probability of finding the particle within any given region of space. He thus gave a statistical interpretation of quantum theory but held, like others, that the probabilities were subjective, a kind of inverse measure of human ignorance, arising from inevitable limitations on our knowledge of quantum phenomena.”

Popper was dissatisfied with this subjective interpretation of probability and he turned to the more objective interpretation of Richard von Mises' theory which took probabilities to be relative frequencies of occurrence that became stable over a sufficiently large number of identical trials.

Popper developed his frequency theory over a long period, hoping to provide an alternative to subjective probabilities in quantum theory.

When Born introduced his statistical interpretation of quantum mechanics he made important use of the idea of singular probabilities, although interpreting them as a measure of subjective ignorance. Popper later realised that Born's singular probabilities were meant to be actual and so could not be treated as formally singular. This difficulty for the relative frequency theory of objective probabilities was, of course, also recognized to some extent by others, but with the result that quantum theorists would inconsistently swing between objective and subjective interpretations of probability in dealing with their problems. In 1956 Popper reached a new objective interpretation of the probability calculus which covered singular probabilities. He proposed, in place of the relative frequency theory, what he called a 'propensity' theory.

A probability or propensity distribution is the distribution of  relative frequencies for all the events to which a specific situation could ever give rise over an infinite number of repetitions. It is to be distinguished from a statistical distribution, which relates to actual relative frequencies recorded for a finite sequence of repetitions.

The propensity distribution is a real relational property of the whole objective situation, not of any of its elements and still less of the events which it generates. Nevertheless this distribution of virtual relative frequencies cannot itself be observed, so that it has to be conjectured; (stated as a hypothesis). The hypothesis, however, can be tested against one or more corresponding statistical distributions of actual relative frequencies, and may well be revised in the light of such testing.

It is important to realize that this propensity theory makes all scientific laws probabilistic, including exact causal laws as the special case of probabilities being unity. He sees an analogy here between his probability propensities and Newtonian forces, and so between propensity distributions and physical fields of force (e.g. gravitational, magnetic or electrical fields).

“Watkins points out that the propensity theory of probability refutes Hume's claim that there is no middle ground between sheer chance and complete determinism. If a dice is loaded so as to make 6 turning up more probable, then it is not a matter of sheer chance that 6 turns up more frequently nor is it bound to turn up every time. He also points out that there can be propensities superimposed on propensities, perhaps in a hierarchy as in the case of a pinboard whose pins are altered in accordance with the falls of a loaded dice so that its propensity to give a certain relative frequency distribution for dropped balls is continually altered in accordance with the propensity of the loaded dice to generate relative frequencies of dice outcomes. This, of course, is a very artificial example; a better one could be the tendency of weather variations to affect the tendency of an area's sheep to grow fleeces of various weights.”

Recently Popper saw that his propensity theory, first devised to meet difficulties in interpreting quantum mechanics, had implications going well beyond it and, indeed, far beyond physics. They reach to any field of knowledge which uses probability laws, including their special case of causal laws, and so to all aspects of the physical, biological and social sciences. All of them can work with an objective theory of probability, and all can give up the idea that the universe, or any part of it, is completely determined.


Simkin explained that no theory, scientific or otherwise, can be completely free of presuppositions, some of  which, being beyond empirical testing, are metaphysical. Newton's laws were supposed to operate within a Euclidean space of infinite extent, over a time of infinite duration, in a completely deterministic framework. Dante's Commedia combined Christian theology with Ptolemy's cosmology, and the American Declaration of Independence assumed the doctrine of natural rights.

“Metaphysical ideas, in one way or another, permeate all our thought because we absorb them, often unconsciously, from our society's traditions and culture, as well as from our education and general intellectual development. Positivists, from the time of Hume, have deplored this permeation of our thinking by metaphysics, and have sought to correct it. Logical Positivists have taken the view that any metaphysical idea is 'meaningless' be- cause it is not reducible to elementary statements of experience.”

It was very awkward for the positivists to discover, as Popper has pointed out, that the laws of science (universal generalizations) would be meaningless by that criterion.

Popper at first granted metaphysics a role in leading the way towards properly testable scientific theories, later, (during the 1950s, perhaps earlier) he started to think seriously about “metaphysical research programs”. This development was no doubt assisted by Joe Agassi’s  interest in the topic. Agassi was Popper’s research assistant for much of the 1950s, through the time that Popper wrote hundreds of pages of manuscript for a companion volume to The Logic of Scientific Discovery (1959).

The second volume was tentatively titled The Postscript to The Logic of Scientific Discovery: After Twentyfive years. It began as footnotes to the LSD but they grew to a point were a second book was required. People may recall how the two volumes of The Open Society and its Enemies grew out of  the section on Essentialism in The Poverty of Historicism. The twenty five years referred to the period from 1934 (the date Popper liked to ascribe to Logic der Forschung) to 1959. In the event 25 years became almost 50 because the three volumes of The Postscript eventually appeared in 1982 and 1983.

All of Popper’s students and associates had access to the manuscript of the Postscript and Bryan Magee described the Metaphysical Epilogue as possibly Popper’s finest piece of writing (without giving any hint of the contents). [Lakatos appropriated the theory and made a splash with his theory of  Scientific Research Programs. These are practically identical to Kuhn’s paradigms, indicating the ambition of Lakatos to affect a Hegelian synthesis of Popper and Kuhn and hence to manifest the Lakatosian Spirit of the Age. The tormented progress of the Lakatos and Kuhn project caused a great deal of wasted effort on several continents, especially in the methodology of economics.]

But I digress. 

Simkin explains that Popper expanded the scope of metaphysical ideas from leading the way or providing ideas for testable theories into a more pervasive role in shaping and selecting our problems, the way we formulate them and the way we evaluate our tentative solutions.

“Metaphysical research programmes have thus been indispensable to scientific development. They often began as sweeping generalizations about the universe or life, deriving from intuitive ideas most of which are no longer accepted. But these speculative theories gave, for a time, unity to thought about various aspects of reality, suggested problems for investigation and gave purpose to investigations which became more scientific as the theories were changed to permit empirical testing. Popper has confessed to changing his mind about one important aspect of the difference between science and metaphysics. He now looks upon a metaphysical theory as similar to a scientific one in regard to its implicit claim to be considered, tentatively, as true. It is, to be sure, vaguer and lack of testability is 'its greatest vice'. But so long as a metaphysical theory can be rationally criticized it should be evaluated by appraising its claim to truth.”

Simkin completed the chapter with more explanation of the role of critical discussion of these programs, how this process worked in physics and how Lakatos and Kuhn produced rival “program” theories which purported to advance and correct the Popperian program, using invalid arguments against Popper’s views on testing and the rational appraisal of rival theories.

“The difference between Popper and Kuhn is thus epistemological and, from the standpoint of the growth of knowledge, which Popper takes to be the basic aim of science, his epistemological stance is superior to Kuhn's. Popper has taken the central issue between Kuhn and himself to have been stated in the title of Kuhn's paper, 'Logic of Discovery or Psychology of Research?', the paper which he contributed to The Philosophy of Karl Popper. In it Kuhn argued that the explanation of scientific progress 'must, in the final analysis, be psychological or sociological'. Popper's rejoinder was that science is part of World 3 and so can be investigated only logically. Hence 'any good psychology of research will have to depend on, and be guided by, the logic of discovery'.”

“Kuhn's ideas had a vogue among economists and other social scientists soon after their publication. But his relativism led to dissatisfaction, and there was a strong shift in the 1970s towards the adoption of Lakatos' ideas, especially by those who incline towards ideology or scepticism. Blaug had begun as a supporter of Popper's ideas but, becoming dissatisfied with his ahistorical stance, is now a leading and influential advocate of Lakatos' ideas among economists.”

This was published before the vogue of Lakatosian theory among economists had run out of steam and was clearly a degenerating program.


"Popperian epistemology marks a major achievemint: the first and only unfied theoiy of knowledge. One single coherent process of knowledge, a problem- solving process, is seen as stretching from the earliest inklings of life to the latest advances in science and technology."
Günter Wãchterhãuser

This chapter could have been placed after the first chapter on Growth of Knowledge because EE represents the extension of Popper’s views on scientific knowledge to encompass all kinds of learning.

“It was pointed out in §1 that Popper's main interests have been in epistemology and cosmology, and that both are part of metaphysics. It seems appropriate to consider these interests further because his earlier views about them have undergone considerable changes - not so much revisions as bold and illuminating developments. In this section I consider his invention of what Donald Campbell has called evolutionary epistemology.”

Bartley and Radnitzky edited a brilliant collection of papers on Evolutionary Epistemology  and Rationality, including two of their best papers on the topic by Popper and Campbell. That collection is summarized here and another collection on evolutionary epistemology is reviewed here.

Simkin noted that there is a passing reference to the Darwinian process of survival of the fittest theories in Logik der Forschung but it was not until 1961 that he began to develop evolutionary epistemology and surprised his seminar with the unveiling of well developed ideas that emerged in the Herbert Spencer Lecture at Oxford.

Key works are  Objective Knowledge: an Evolutionary Approach (1972), some passaged in his intellectual autobiography in The Philosophy of Karl Popper (1974), The Self and Its Brain (1977) where he wrote on 'The Biological Approach to Human Knowledge and Intelligence'  and on 'The Biological Function of Conscious and Intelligent Activity. Then  the first Darwin Lecture in Cambridge, 'Natural Selection and the Emergence of Mind', reproduced as Chapter VI of Evolutionary Epistemology (1987, edited by Gerald Radnitzsky and W.W. Bartley). His latest contribution is 'Towards an Evolutionary Theory of Knowledge', published as the second part of A World of Propensities (1990).

“The result has been a unification and generalization of his philosophy. This is indicated by the change in his view about the main task of epistemology; in The Logic of Scientflc Discovery it had been stated as 'the problem of the growth of knowledge', but in The Philosophy of Karl Popper it is stated as 'to understand it as continuous with animal knowledge; and to understand also its discontinuity - if any - from animal knowledge'. Emphasis was thus shifted from the top end of the evolutionary scale, scientific knowledge, to the whole of it, and knowledge widened to include cognitive structures of men and animals. It has dealt a further blow to the naive belief of the British empiricists and their followers that human sense experience is the foundation for justifiable knowledge. The basis of this theory of evolutionary epistemology is a very wide concept of knowledge as adaptation to a partly unknown environment, and hence an uncertain adaptation.”

Simkin explains many of the implications and ramifications of this perspective, and notes the remarkable contribution of  Wãchterhäuser who applied Popper’s ideas to produce a novel theory to account for the evolution of the eye, starting with the sensitivity and survival advantage of bacteria that were sensitive to light.


This chapter provides a clear and concise summary of  Popper’s major arguments against determinism.

Popper has defined scientific determinism as ‘the doctrine that the state of any closed physical system at any given future instant of time can be predicted, even from within the system, with any specified degree of precision, by deducing the prediction from theories, in conjunction with initial conditions whose precision can always be calculated…if the prediction task is given'.

Newton’s achievement moved many great thinkers from Laplace and Kant to Einstein to accept the truth of  absolute scientific determinism but Simkin explained that the inference from the truth of a theory to the validity of scientific determinism is invalid because, although a true theory would describe certain properties of the world, not every property of a true theory will necessarily be a property of the world.

“A deterministic theory, even if successful within its limits, does not guarantee a completely deterministic world. The reason is that the inference could be valid only from a complete system of physics that would allow for the prediction of all kinds of physical events. We cannot have such a system because our theories reflect our own fallibility. They are necessarily simplifications, and hence approximations, not only because of our mental limitations, but because complex theories may well be untestable, even if they are true, and whatever a theory does not explain remains obscure. Scientific progress has meant replacing defective theories by better ones and, from the standpoint of the replacement, the previous theories appear, at best, as cruder approximations. It is conceivable that someday we would come to an end of this process of making ever better approximations - that we might reach a complete and true theory of the world - but if so we could not possibly know that we had found it.”

A second argument draws on Einstein's special theory of relativity according to which, there is, for every observer or local inertial system, a past region of spatio-temporal points from which physical influences can affect the system, and a future region of all such points as may be affected by past and present physical influences. But from no place in the future region can a physical influence be exerted on any place in the past and recent region. This means that the future becomes open to us in the sense that we cannot fully predict it. The reason is that any event in the future region may be affected by an influence which is in the distribution of its own past region but also in our future region, and so unknown to us because no influence from the future can reach us in the present.

Using another analogy, we know that our future will be influenced by our children but we can’t predict precisely what that influence will be.

A third argument against determinism is that 'we cannot scientifically predict results which we shall obtain in the course of the growth of our own knowledge'.

Determinism is self-defeating in a way because it unintentionally entails the non-existence of arguments. Simkin reports that Arthur Compton, the American physicist, had similar misgivings about determinism in relation to morality. What scope, he asked, can there be for morals and moral efforts if  the atoms of our bodies follow physical laws as immutable as the motions of the planets? He welcomed the new indeterminism of quantum mechanics as liberating him from this dilemma, and felt able to assert that 'it is no longer justifiable to use physical law as evidence against human freedom'.

In a Compton Memorial Lecture Popper introduced the idea of 'plastic controls' to designate intermediate positions between complete determinism and pure chance or complete freedom. (Hume and others had asserted that there was no such middle ground.) To a considerable extent we are free, but our freedom is always limited by physical, biological or social conditions; these are our controls but they are plastic because we can affect them by our actions so that there is a kind of feedback.

This point was not spelled out by Popper or Simkin but the concept of plastic control is especially relevant in the human sciences when we explore the role of conventions (man-made laws) in regulating and directing human actions. These man-made laws only work when people know about them (not necessarily consciously) and when we deliberately observe them (when we realise that there is discretion, like speed limits on the road).

Simkin noted that the idea of plastic controls is particularly important in biological evolution as an organism can be regarded, 'a quasi-hierarchical system of plastic controls’. In addition, Popper insisted that world 3 has a (plastic) control over much of our action and if this is so, then the essential unpredictability of world 3 enters world 1 along this path.


"Quantum mechanics is very impressive. But an inner voice tells me that it is not yet the real thing."
Albert Einstein

Popper views on determinism, subjectivism and the subjectivist interpretations of probabilities evolved in the course of his  critique of quantum theory, in defence of indeterminism (and human freedom), realism (both in the sense of the reality of the external world and the existence of  the world of ideas) and objectivism (especially the frequency interpretation of probability). The latest key work in this project is Quantum Theory and the Schism in Physics, published in 1982, though “in progress” from the 1930s.

Simkin wrote:

“Inevitably much of the criticism deals with highly technical considerations so that, as a layman writing for lay people, I shall attempt only a general indication of its drift. He speaks of a crisis in understanding or interpreting quantum theory. It has arisen, he holds, for two main reasons. One is an intrusion of subjectivism into physics which led to a denial or questioning of the objective reality of physical phenomena. The other is an idea that quantum theory is the 'end of the road' in physical theory-the truth beyond which we cannot penetrate.”

I will attempt to summarise Simkin’s summary, which I think is very good.  Elsewhere I have listed some of the main ideas in the three volumes of The Postscript.

Simkin wrote this book especially for economists and so it is interesting to see the role of von Neumann in supporting the “end of the road” theory of quantum theory. This is because he played a similar role in supporting the program of General Equilibrium Theory in economics.

“The end-of-the-road view came out in a celebrated argument between Einstein and Bohr whether quantum mechanics was complete. Einstein held that it could not be, but Bohr was thought to have won the debate with the opposite view, taking Heisenberg's uncertainty principles as revealing an impassable limit to further basic progress in quantum mechanics. Von Neumann seemed to have clinched the argument with a mathematical proof that there was nothing more basic in physics than the reality presented by quantum mechanics. His proof asserted that there could be no 'hidden variables' in quantum mechanics but, as Popper points out, in the same year, 1932, that this proof was published two new particles, the neutron and the positron, were discovered
and, a few years later, a number of mesons. Some variables thus emerged from hiding, and they were then explained by a powerful mathematical formalism which much enlarged quantum mechanics beyond what had been a theory of the motion of a negatively charged electron in the positive field of the nucleus-the theory which was to have been the end of the road in physics.”

Popper's argued the problems of quantum mechanics are essentially statistical ones, and quantum theorists have misinterpreted its results because they have not sufficiently
appreciated that statistical problems require statistical answers, and because they have held erroneous subjectivist ideas about probability.

“The statistical character of quantum mechanics has been clear from its beginnings with Planck's radiation formula and Einstein's hypothesis about protons up to Born's disposal of Bohr's supposed problem of atomic stability by a statistical interpretation of Schrodinger's wave function. Subjectivism intruded through a mistaken belief that the probabilistic character of quantum mechanics arises from limitations on knowledge of quantum phenomena.”

That was manifest in Heisenberg’s response to the problem of measuring position and momentum the same time (as though the value did not exist if it could not be measured). Yet Simkin notes that great physical theorists, such as Pauli, von Neumann and Born  wavered between objective and subjective interpretations of quantum mechanics and many if not most physicists have turned away from philosophical discussion  to follow the instrumentalist view that it is enough to understand, develop and apply, the mathematical formalism of quantum mechanics. In other words they have given up the attempt to understand the physical meaning of this formalism and with it any chance of  gaining insight into its difficulties and shortcomings.

The parallel in economics is the scholars who are devoted to mathematical formalism and do not look out the window to find whether they are gaining insight into the way the way that real markets work, especially in their relations to other social institutions.

“As was shown in § 5, instrumentalism is an old doctrine, and one which attempts to evade falsification by raising no claim to truth, and for Popper, it is 'uncritical, irrational, and objectionable'. He expresses admiration for Schrodinger who fathered the formalism of quantum theory yet tried hard, in 'the grand style' of physics, to understand the physical world, Einstein had a dream of reaching a unified field theory which would supersede the dualism of matter and field and which would explain particles as resulting from the properties of this unified field. It was not realized, and Popper has a similar dream of overcoming the multitude of field theories associated with a multitude of fundamental particles by developing a unified field theory based on his propensity interpretation. No physical theory can do no more than explain a physical system in terms of its propensities. By adopting the admittedly metaphysical view that propensities are real and described by field equations we would reach a monism of propensities that could overcome the dualism of matter and field yet retain the practical advantages which this dualism may have for treating physical problems. For propensities are both potentalities and potentialities to realize something, and whatever may be realized must again be a set of potentialities to realize something else. This dream of Popper's is metaphysical, not because propensities are unreal, but because it is based on the metaphysical idea of indeterminism. This does not mean that it is useless; for it gives a metaphysical research programme such as those which have proved valuable in directing the growth of scientific knowledge.”

Part II Applications to Social Science

IndexFull Index