He uses the somewhat controversial term 'metaphysical' because these programmes embody general views about the structure of the world which cannot be tested by experiments or empirical evidence. He uses the term 'programme' because these ideas tend to cluster and support each other in various ways; also they provide historical continuity despite changes in the status of testable theories. For example the 'Austrian' tradition in economics contains a number of features (the subjective theory of value, methodological individualism, the recognition of the time factor and the uncertainty of knowledge) that have persisted from 1870 despite many important differences between workers in that tradition on other issues. The development of the program by Ludwig von Mises also involved the postulations of certain a priori principles, and the status of these assumptions in relation to the testable components of economic theory continue to cause concern. Perhaps this discussion (which has been known to become spirited) may be advanced if the Misean a prioris are conceived as the elements of a metaphysical research program, in which case the demand for empirical testability of these principles would be relaxed, without prejudice to the scientific (testable) status of models or theories developed subsequently.

Popper's work can be seen as a defence of a set of epistemological, metaphysical and methodological principles and a critical demolition of rivals. One can discern a 'Popperian programme' and a 'Rival programme' as follows:

Fallibilism or non-justificationism Dogmatism or justificationism

Non-essentialism Essentialism

(definitions are not fundamental) (conceptual analysis)

Realism Idealism

Objective knowledge Knowledge is belief

Non-determinism Determinism

Non-reductionism (emergence) Reductionism

Situational analysis and Holism

methodological individualism

Virtually any of the doctrines in the 'Popper programme' can be used as an organising principle to explain the systematic nature of Popper's work. Watkins used indeterminism in the Shilp volume and one could just as well nominate the critical approach, fallibilism or realism. Popper's 'situational logic' affords a suitable starting point to explain his ideas in the social sciences.

Situational logic postulates that particular events and the patterns of social processes can be explained as the product of countless individual decisions and their unintended consequences. This theory allows for a certain amount of choice by individual actors and it also allows for contingencies (such as natural calamaties) that are beyond the realm of human choice or control. Popper endorses methodological individualism against the claims of collectivists and holists who regard social 'wholes' as the fundamental units of analysis although this does not entail any claims about individuals being more 'real' than, or having any form of priority over, groups or external situations. Some critics of methodological individualism such as Lukes have suggested that this method does not take account of the situation of the individual actor but the very term 'situational logic' draws attention to the context of the actor.

Contexts usually offer alternative courses of action for the actor and this element of choice (which is disputed by historical determinists) clearly requires a metaphysics of indeterminism, a theme which Popper has pursued in both physics and history.

The type of explanation afforded by situational logic is conjectural and does not claim to provide ultimate explanations, unlike the various forms of essentialism which claim to penetrate to the hidden essence of the phenomena. Essentialism occurs in two linked forms, the first being a metaphysical theory about the possibilities for the understanding of things, the second being a methodological theory which results in an obsession with definitions and the meanings of terms which attempt to capture 'essences'.

In situational analysis, various elements are selected from the totality of the situation, guided by the problem under investigation, by theoretical considerations or by a point of view. This process of abstraction and the methodological individualism associated with it support the doctrine of 'piecemeal social engineering' which is Popper's contribution to the debate on methods of reform and social technology. In view of the unhappy connotations of 'social engineering' as noted by Hayek, it must be emphasised that Popper's method is agnostic about the type or extent of intervention that may be attempted in any situation.

Kuhn rejected the inductivist theory that science grows in a steady fashion as observations accumulate. He and his followers also made much of his claim, directed against Popper, that actual or attempted falsification of theories did not play the major or decisive role in the advance of physics. Instead, scientists stood by their theories in the face of refutations and progress consisted in periodic revolutions whereby a ruling paradigm or world view was replaced by another that was incompatible with it. This process of replacement may not be rapid and indeed the older generation of scientists may never accept the new paradigms which triumphs literally over their dead bodies. [Footnote a comment by Sammuelson, cited by Hutchison, regarding the Keynesian revolution "funeral by funeral, science advanced"]. In the period between revolutions there occurs what Kuhn labeled 'normal science', conducted by 'normal scientists' who work on relatively minor 'puzzles' without making any attempt to criticise the wider framework of ideas (the paradigm) in which the puzzles are located.

However it is misleading to represent Kuhn's theory as a critique of Popper's, as Kuhn himself pointed out.

Thus Kuhn's ideas do not provide a critique of Popperian falsificationism and indeed it is not clear that they provide any kind of methodology at all. Perhaps some kind of sociological explanation is required to explain why his book was widely received as a major contribution to the philosophy of science, rather than a contribution to history, or as Blaug suggests, to the sociology of science.

For whatever reasons, Kuhn's views carried sufficient weight to convince many people that Popper's ideas were inadequate and this prompted Lakatos to attempt to "retrieve" whatever could be saved from Popper's program.

Lakatos proposed a methodology of scientific research programmes in which the unit of analysis is not a single theory but a cluster of theories. The programme is not decided by its probability or even its degree of corroboration, but rather by its performance over a period of time. Progress is indicated by 'progressive problem shifts' whereby the content of the program increases with the corroboration of predictions. In contrast, 'degenerative problem shifts' are signaled by the proliferation of ad hoc hypotheses.

The programme has two parts; one is the 'hard core' which is treated as irrefutable by "the methodological decision of its protagonists". This part of the programme provides its distinctive character. The second part of the programme is described as the "protective belt" of auxiliary hypotheses which bear the brunt of tests and criticisms of the programme. These hypotheses are used to absorb or deflect criticism from the hard core.

This theory can be depicted as an attempt to improve on Popper and Kuhn by getting the best of both worlds. From Popper he took the theory of programmes, from Kuhn he took the idea that the most important parts of the programme should not be subjected to criticism.

His first major paper on programmes appeared in 1968 and by the time he died in 1974 it was clear that his ideas were converging with those of Kuhn despite his best efforts at demarcation. In response to one of Lakatos's papers "History of science and its rational reconstruction" Kuhn concluded

However from a more critical point of view the theories of Lakatos and Kuhn share a common weakness in that they tend to deflect criticism from those elements of the 'paradigm' or the 'hard core' which provide the essential framework of discourse. Apparently these frameworks may be attacked from the vantage point of another program but this raises the question of criteria that are to be used for critical appraisal of the merits of rival frameworks. In addition, it appears that these frameworks need to be treated as 'wholes' despite the mixture of metaphysics, ethics, ideology, and methodology that are likely to constitute a real life tradition of research.

An alternative approach to that offered by Lakatos and Kuhn calls for the relentlessly critical and piecemeal appraisal of all elements of paradigms, hard cores and frameworks. Kuhn and Lakatos fail to provide assistance in this enterprise, though they may have some merit in identifying a problem, namely the existence of 'paradigms' and 'hard cores' that are not subjected to criticism. It is precisely here that Popper's theory of MRPs is helpful.

Popper's theory of metaphysical research programs can be used to identify and criticise ideas in 'rival programmes' which are often regarded as the basic categories of thought, ultimate presuppositions or framework assumptions. These theories, such as determinism, justificationism, subjectivism (knowledge as belief) and the method of conceptual analysis (essentialism) are shared by most philosophical schools and for this reason they are seldom in dispute. They operate as invisible boundaries that dictate the type of problems that people choose to work on, the way that they are formulated and the type of solutions that are deemed acceptable. These boundaries render certain theories and methods either wrong, or stupid, or ideologically unsound, or in extreme cases, simply unthinkable. They provide the intellectual equivalent of an ecological niche where some species of ideas are stunted or killed outright while others survive and flourish. They are not amenable to experimental test or falsification, so they have persisted unchallenged by the rise of empiricism in the philosophy of science and by the triumphs of science itself. Far from being challenged by science, some of these ideas are located at the core of modern physics and they derive strength from their association with popular interpretations of quantum physics.

The influence of metaphysical ideas is in no way reduced by the positivists' aversion to metaphysics in toto. The philosophy of science from the time of Hume has been intensely hostile to metaphysics and in this century alone , the logical positivists have wanted to brand metaphysics literally meaningless, while the pragmatists and instrumentalists have hardly been more hospitable. The positivist program aimed to drive metaphysics out of science and philosophy, but people who try to deny a role to metaphysical ideas simply render them invisible, hence immune to criticism.

Metaphysical ideas such as determinism and reductionism have proved almost impossible to subject to effective criticism because discussion of them has mostly been pursued by conceptual analysis under the influence of the errors that Popper labeled 'essentialism'. The resulting proliferation of unhelpful verbiage has done much to prejudice people against metaphysics. It is often assumed that metaphysical theories are beyond the limits of rational criticism because they provide the framework of discourse. A widespread view (held by Popper in his youth) claims that the domain of rational criticism is limited to areas where empirical tests can be used.

Popper has addressed this situation in two ways. First, the theory of metaphysical research programs has made metaphysical theories visible, in the same way that an improved microscope brings objects into sharp focus where previously they were either invisible or indistinct. Second, Popper has shown that metaphysical theories can be critically discussed, even though they are not empirically testable, and it is possible to criticise and compare rival theories. He has provided criticisms of (and alternatives to) the theories of reductionism, determinism and subjectivism in the context of live scientific problems. A prime example of this is Quantum Theory and the Schism in Physics, which explores the impact of the old metaphysics on the interpretation of theories in quantum physics. Unfortunately we do not know of any comparable piece of criticism in the social sciences.

Not surprisingly, Popper's account of MRPs in Quantum Theory and the Schism in Physics has not yet engaged the attention of many social scientists. Similarly Kuhn and Lakatos developed their ideas in relation to the history of physics and they made no serious attempt to apply their ideas in the social sciences. Other people have done this and there is now a huge literature on the implications of the "post-positivist" philosophy of science in economics, psychology, sociology, history and even in literature. Two features stand out in all of this. First is the failure to take up some important ideas from Popper and to rely on Lakatos as though he represented the latest developments in Popperian thinking. Second is the lack of interaction between the philosophical articles and reports of empirical or policy-oriented work in progress.

In economics, Hutchison has some illuminating comments on revolutions and the progress of knowledge. He drew attention to the methodological revolution brought about by Ricardo's Principles which removed the historical dimension of the subject which was so important for Adam Smith. He also quoted a comment from Menger on the Methodenstreit where Menger insisted that the issue was not, as often reported, between the inductive method of the German historical school and the deductive method of the Austrians, or between the empirical and the rationalist approach.

It should be apparent from the above discussion that this is precisely the kind of disagreement which may be illuminated by the theory of metaphysical research programmes.

Blaug and Caldwell have also attempted to bridge the gap between philosophy and substantive issues though the results are not encouraging. They tend to operate at the level of historical reconstruction of theoretical progress and it is hard to find any concrete event or issue systematically explored in a way that could contribute to a debate on methodology. Their critique of Popper fits the standard pattern that is criticised in previous extracts [in the proposed collection] and they have not yet taken account of the theory of metaphysical research programmes.

Return to full list of papers