1. How Did Popper's Social View of Science Emerge?

Popper's introduction of a social aspect into his theory of science cannot be explained without taking into consideration those essays that Popper wrote before Logik der Forschung (Reference Popper1994), since it was a response to difficulties that arose in these essays. Jarvie does not do this. Rather, he points to three problems that he finds in The Logic of Scientific Discovery (Reference Popper1962b) that, he says, caused Popper to make his methodological, that is, his sociological, turn in the philosophy of science. One of these—the introduction of methodological rules to meet the objection that any theory could be protected from refutations with small changes—was in fact crucial for his introduction of methodological rules. This introduction took place as he moved from his first view, which was an attempt to demarcate scientific sentences with the tools of logic alone—to his second attempt—his methodological theory of how progress in science is made. The first is found in Die beiden Grundprobleme der Ekenntnistheorie (Reference Popper1979) without Chapter 5 and the latter in Die Logik der Forschung (Reference Popper1994). Jarvie takes no notice of this first view, and therefore cannot explain the switch. As regards the other two problems, one—the objection that, since science should give positive information, it is wrong-headed to characterize it negatively—was solved with methodological rules, but after he had introduced them for other reasons, and the third—falsifications can no more be verified than inductions—played no role in the turn at all, since before the turn he thought that singular statements were veridical and falsifications were, therefore, secure. The fact that observations statements are not veridical as he presumed they were in Die beiden Grundprobleme without Chapter 5 did play a role in the change, since he introduced a methodological rule to account for this fact. This history has been related elsewhere and may be passed over here (Wettersten Reference Wettersten1985c, Reference Wettersten and Eschbach1988, Reference Wettersten1992a, Reference Wettersten2005)

2. Should We Read Popper's Philosophy of Science in Logik der Forschung as a ‘Protoconstitution’ of Science?

Jarvie presents Popper's early view as containing, if not consisting of, a ‘protoconstitution’ of science (2001, 51ff.). This ‘protoconstitution contains rules that scientists (ideally) practice. These rules demarcate science: to the degree that would-be scientists do not follow them, they are not doing science. This does not mean that rules cannot be reformed, criticized, and improved. But this project goes beyond Popper's inquiries.

Jarvie offers his own list of Popperian rules, but one should note that it is relatively easy to translate into rules aspects of Popper's theory that he did not present as rules. One can do that for any philosophy of science. This is not wrong, since rules may express how scientists should act, if some explanation of scientific progress is true. But we should be careful not to translate Popper's theory into a mere rule book for science. This approach distorts it by tearing it out of its theoretical context. Popper's theory does not merely say what scientists do when they do research well, but also explains why this activity leads to knowledge. Jarvie wants to evaluate these rules by looking at them socially, in real institutional contexts, and this is surely a good thing to do, but they also need to be evaluated philosophically. Jarvie would apparently agree, but strips too much from the deeper philosophical context in order to present the rules—the content of the constitution of the ‘republic of science’. (The term ‘republic of science’ comes from Polanyi.) Jarvie notices the difficulty, but responds by merely emphasizing the need for a social dimension (2001, 83). This is not enough: the problem of how to evaluate such rules calls for an integration of the philosophical dimension into the social studies of science.Footnote ¹

3. Popper's Development of His Theory of Science in The Poverty of Historicism and The Open Society and Its Enemies

Jarvie suggests that in The Poverty of Historicism (Reference Popper1960) and The Open Society and Its Enemies (Reference Popper1962a) Popper further developed the skeletal, ‘protoconstitutional’ social view of science found in Logik der Forschung (Reference Popper1994). In The Open Society and Its Enemies (Reference Popper1962a) Popper did not, Jarvie perceptively observes, merely apply the social view of science found in Logik der Forschung (Reference Popper1994) to problems of social science and social theory. He had to further develop his theory because he had previously never noticed the institutional context in which scientific research takes place. The views Popper expressed in The Poverty of Historicism (Reference Popper1960) and The Open Society and Its Enemies (Reference Popper1962a) can even, Jarvie suggests, be taken as an implicit criticism of the earlier view as unrealistic: the earlier view emphasizes the need for methodological (social) rules, but it does not recognize the need for institutions, the care that these institutions require, their imperfections, and the dangers that the open society, even science, could die. It is primarily through this historical thesis that Jarvie offers a new perspective on the growth of Popper's philosophy.

Jarvie says that in The Poverty of Historicism (Reference Popper1960) Popper extended his view of science proper in two ways. “First, in his ideas about ‘the social or public character’ of science. … and second, in the connection he makes between his individualistic and competitive model and the possibility of progress” (2001, 14). Progress and objectivity are now portrayed as products of institutions that allow individuals to put forth ideas and subject them to critical appraisal. The nonsocial elements of Popper's view—logic, truth and falsity, and transcendence—enable Popper to develop a social view of science without explaining the objectivity of science away, as the so-called strong program in sociology does (2001, 119).

4. How Good Is an Updated Version of the Framework Developed in The Open Society and Its Enemies for Developing Social Scientific Explanations?

In his advocacy of the use of Popper's theory to explain social events, Jarvie concentrates on two oppositions that Popper developed in The Open Society and Its Enemies (Reference Popper1962a), that is, that between closed and open societies, on the one hand, and that between abstract and concrete societies, on the other. Jarvie hopes to overcome criticism of these oppositions as Popper presented them by treating them as ideal types. He justifies this strategy by appealing to a methodological rule proposed by Popper: one should always interpret a view to be criticized in the strongest possible way (2001, 149). This seems quite reasonable and generous. But it is a dangerous policy. Just which interpretation is the strongest depends on the point of view of the interpreter, which may be quite at odds with that of the author. Popper was often read as providing a theory of meaning. This is not what he said, but from the point of view of many, Popper's theory was more interesting and stronger when it was interpreted that way. This was quite wrong and Popper and his followers protested vociferously that they really meant what they said, that Popper offered no theory of meaning and no theory of justification.

Jarvie uses this seemingly generous strategy of interpreting Popper in the strongest way to protect Popper's view from some effective criticism, but in doing so, he seriously diminishes its explanatory power. Both as a matter of promoting a clear discussion and as a matter of fairness to Popper, it is better to note that Popper's theory is inadequate and to construct new ones. We may then thank Popper for instigating us to do better and to see more clearly what is gained or lost by the change.

Popper's opposition between closed and open societies has been criticized due to Popper's assumption that all individuals have a longing to return to a tribal state. By treating Popper's opposition as an ideal type, Jarvie frees it of this assumption. But Popper's theory that all individuals long to return to a tribal state was crucial for his analysis of the problems of the difficulty of maintaining open ones. If this theory goes by the board, a completely new theory of the closed society and the dangers its attractions pose is needed (Wettersten, Reference Wettersten, Miller and Jarvieforthcoming).

Popper sought to explain how a civilized society differed from a closed society. One aspect of his solution to this problem was his characterization of the open society as an abstract society. He used the example of a government that organized such factors as train schedules, but that did not intervene in the operation of real social groups. The building and operation of such groups should be left to individuals. They are then free to accept them or to leave them or to build new ones, as they are not when a society has only one coordinating theory of how to live. The view of an abstract as opposed to a concrete society should solve the problem of explaining what a government should and should not do: it should set the abstract conditions within which individuals can develop concrete, real social relations independent of the influence of the State.

At first blush this is a very attractive view and points to conditions that any open society must meet, above all, the freedom of individuals to choose their associates and to build groups that satisfy their individual needs.Footnote ² But it will nevertheless not do. It will not do, because, even though all modern societies need abstract elements in order to enable strangers to interact with each other in business, education, and so forth—as Gellner (Reference Gellner1994) has pointed out—no society can be built on merely abstract elements. At first blush, Agassi's (Reference Agassi1999) book on liberalism in Israel may appear to be merely a study of problems peculiar to Israel as a Jewish state. How can Israel be both Jewish and liberal? But it is not. It is an example of the central political problem of today. Americans struggle to answer the question of how much unity they need and when the quest for unity begins to sacrifice the open society as the lively debate about Michael J. Sandel's call for a ‘public philosophy’ indicates (Sandel Reference Sandel1996; Allen and Regan Reference Allen and Regan1998). The French have a secularist consensus that Americans hardly understand, as the reaction to the French banning of head scarves in public schools shows, but that is just as fundamental to them as the American Constitution is to Americans. Europeans struggle with the problem of how to maintain their view that it is the responsibility of the state to see to it that no suffering is too great and that all individuals are integrated into society in the face of an assault brought on by globalization neoliberal economics. Muslim societies face the problem of how much control and influence Islam may have in a modern state. Rwanda hopes desperately to build up nationalist sentiment in order to overcome the bitter opposition of Tutsis and Hutus and Iraq needs some form of nationalism to provide for peaceful coexistence of Shiites, Sunnis, and Kurds. It is naive to push these problems aside as Popper does with his opposition between abstract and concrete. We need to face the far more difficult problem, which Popper's theory unsuccessfully seeks to avoid, of how an open society can be constructed, maintained, and improved even when we need some coordinating consensus about values, the extent of law, and the problems the state should solve. We cannot return to Durkheim's exclusive emphasis on a unifying consensus that makes society function well. Rather we need to improve the political debate. And here Popper's fallibilisitic theory of rationality is quite useful, by preserving rationality in the face of continuing disagreement. But in itself this is also not a solution, but an element of one.

Jarvie sees that Popper's theory of an abstract society cannot solve the problem of explaining how the demands of individuals for something more than an abstract society can be met while keeping a society open. But he centers his criticism of Popper's theory on the distinction Popper makes between nature and convention and the use Popper makes of it (2001, 154ff.). Here he calls for improvement but does not supply it.

The distinction Popper uses between nature and convention is inadequate, Jarvie says, to account for many of the features of political life. We do not regulate our political affairs merely on the basis of convention, but also on the basis of what we take to be human nature. We cannot merely say, then, that society is conventional and nature is not, as, I might note, Agassi has so forcefully argued (Agassi Reference Agassi1977). But Jarvie does not offer any view of the nature-convention distinction or what we may use in its stead. Rather, he slides from nature-convention opposition to fact-decision opposition without tackling the problem of how to find an alternative for the first. He says that, whatever we make of the nature-convention distinction, Popper is right to indicate that humans are responsible for the decisions they make.

Jarvie also follows Popper in associating the denial of the nature-convention distinction with dogmatism and acceptance with a critical approach. Quite intuitively this may seem right, because those who claim that their social programs are God's will may very well be dogmatic and often are and those who deem programs mere conventions may be open to change and often are. But one can also be open and critical about what God's will may really be—as many religious people are—and one can, as Jarvie recognizes (2001, 187), be dogmatic about conventions—as many are when they stubbornly say that some policy or convention cannot be changed even though they are quite clear that it is conventional. They may say, for example, that it is part of their culture and cultures must be preserved.

In the philosophy of science things tend to go the other way. Those thinkers such as Duhem, Quine, Polanyi, and Kuhn, who endorse conventionalist views or views close to them, tend also to approve some form of dogmatism. Theoretical frameworks should not be changed, but not because they are true but because change of convention is bad. And this argument for conservatism in science based on conventionalism can also be used to support conservatism in political or social matters: reform is too difficult to engage in because we often overlook values that existing arrangements possess and we cannot anticipate the unintended consequences of our actions. This is, indeed, a core of traditional conservative political theory.

Jarvie's exposition of Popper's theory leaves the problem of explaining the relation between nature and convention unsolved. This is important for both the social sciences and politics and thus shows serious limits to Popper's view. We cannot merely say, as Popper does, that we have to take responsibility for our decisions. We also need to say how this responsibility may be exercised. And this is not merely a matter of technology. That Popper seems to reduce these problems to technological ones is one reason why his view has been found by many to be quite inadequate. We need something more to explain the relation between convention and nature, or as Agassi has argued, we must find our way to understanding those aspects of humanity that include aspects of both without being either one or the other (Agassi Reference Agassi1977; Wettersten Reference Wettersten1982, Reference Wettersten1985b). We need to understand how we may improve theories of the intertwining of nature and convention and how these theories may help to understand how consensus in political frameworks that protect openness may be achieved.

5. Can We Use Popper's Framework to Improve Our Understanding of Nationalism?

In his discussion of Popper's treatment of nationalism as a modern version of tribalism, Jarvie points out that Popper's theory of treating social arrangements as conventions does not offer a way of preventing tribalism when it is embodied in nationalism (2001, 182–188). He calls for a better theory of nationalism, which will enable us to better combat the tendencies to move toward a closed society. Such a theory, he says, requires a distinction between nationalism and tribalism. Gellner has shown that, although nationalism sometimes is tribalism, it is not always tribalism and may even serve to block tribalism. This interesting theme and Gellner's important contributions to it are not pursued further in this book. Jarvie has, however, furthered this discussion elsewhere (Hall and Jarvie Reference Hall and Jarvie1996).

Jarvie's nod in the direction of Gellner once more shows the inadequacy of the oppositions offered by Popper that Jarvie nevertheless recommends so highly. Neither Popper's opposition between closed and open nor the opposition between abstract and concrete are adequate for the task of combating movements toward closed societies. Popper's idea of the closed society is so intimately bound up with his theory of primitive tribal societies that it loses its power when interpreted merely as an ideal type. We need a far better theory of closed societies than this to make progress. And Popper's distinction between abstract and concrete cannot take into account the need for broad social consensus on aims and values. There are relations that individuals value that are neither quite face to face, because they involve strangers, and that are not quite abstract, because they involve feelings of unity. These relations exist between individuals sharing nationality, religion, or ideology. However we evaluate them, we need to take them into account. Popper's theory cannot do that adequately.

6. Can Popper's Model of Science Serve as a Model for Society?

Jarvie points out that science, as Popper conceives it, requires criticism and openness to change and thus institutions that encourage them (2001, 188ff.). Following Gellner, he notes that these characteristics, as well as the interest in truth, pose a danger not only to nationalism but also to traditions. In all societies individuals feel some need to protect traditions. This fact has to be taken into account in any theory of the open society. Individuals value established institutions and are not willing to put all of them at the disposal of a critical process.

Jarvie's point is well-taken, but his view on this point confirms the observation that Popper's distinction between abstract and concrete societies is not a sufficient basis for a theory of the open society. Abstract rules can always be subject to criticism and, if that is all we needed, we could also use the critical model of science for a model of the open society. But all societies—even those we call open—have substantive and not merely abstract rules and traditions whose influence has to be taken account of in any realistic theory of open societies.

7. Can We Apply Popper's Standards of an Open Society to Science?

Jarvie asks whether Popper's theories of justice and piecemeal engineering can be applied to science. Popper's theory of justice, according to Jarvie, says that justice sets five requirements:

1. 1. An equal distribution of the burdens of citizenship, i.e., those limitations of freedom that are necessary in social life;

2. 2. Equal treatment of the citizens before the law;

3. 3. Provided the laws themselves neither favor nor disfavor any individual, group, or class;

4. 4. Impartiality of the courts of justice;

5. 5. An equal share of the advantages that membership of the state may offer to the citizen.

Jarvie argues that this theory of justice does not describe science as it is nor could it describe science; science cannot be a model for society, as Popper thought (2001, 161ff.). These requirements are so high they can hardly be applied to any society. Requirements 2, 3, and 4 are certainly ideals of democratic states. But surely Requirements 1 and 5 are utopian. How they may be approximated may be a problem for any society.

Jarvie asks whether Popper's theory of piecemeal engineering could apply to scientific institutions. He sees here a further opening for his institutional criticism of Popper's institutional view of science: science must also be subject to process of correction and reform. On this point he is surely right. Criticism and reform can include efforts to bring science closer in line with standards of justice, openness, and decency that we expect of open societies.

We can easily agree that science as it is practiced today is often a poor model for an open society. As Polanyi and Kuhn have argued so forcefully, it is often authoritarian and elitist. But that does not mean that we cannot take the qualities of science that favor openness and seek to implement these in both science and society. These qualities include respect for truth and those who disagree with one's own views. The recognition that ideas have to be judged on their content rather than on the authority who proposes them and that the public forum should be open to anyone who has ideas. This is all, I think, that Popper had in mind, when he viewed science as a model of an open society. At another point in his argument Jarvie seems to agree. Science can come closer to meeting the standards of justice, as Jarvie suggests, if it is viewed as part of society at large. We may then work toward the admittedly unattainable goals of giving everyone access to it who has any ability or interest and that of allowing anyone a hearing who is seeking to say something. As Agassi has argued, the more inclusive and open science and technology are, the better they are.

8. Does Jarvie Offer a New Program for Future Research in the Social Studies of Science?

Jarvie concludes his book by looking backwards to Popper's view of science in Logik der Forschung (Reference Popper1994) from the point of view he finds in The Poverty of Historicism (Reference Popper1960) and The Open Society (Reference Popper1962a). He proposes using the latter view as a basis for criticism of the former. Popper's later more sophisticated view reveals problems of institutions, and rules and the relation of science to culture and more, as Jarvie points out. But these were just those problems that Popper shied away from. Due to his suspicion of any sociology of science Popper rejected moves in the direction of a social view of science as often as he supported them. He saw a threat from this direction to his realism. All too often sociologists of science explain away the rationality of science by explaining why scientists took the views they did as a result of social interests rather than of interest in truth. Jarvie rightly speaks about the difficulties others had in understanding this aspect of his view, but also notes that Popper had the same difficulty and did not know what to do about it (2001, 175).

Jarvie's thesis, that the social aspect of Popper's philosophy has been, with a few exceptions, neglected, is important and correctly describes a large portion of the literature on Popper. But Jarvie does nothing to explain the nature of these exceptions or how far they have come. Following Popper, he complains about the bad criticism while ignoring the good.Footnote ³ This leads to a misleading picture of the reception of Popper's philosophy. It leaves the false impression of a wasteland in which nothing has occurred in the direction Jarvie sees as progressive.

The task of critically studying the rules of scientific research has been underway for some time. Agassi led the way in applying Popper's theory of the logic of science to the practice of science. One of his earliest suggestions was that Popper's rule that scientists should always favor the theory with the highest degree of testability could not be upheld, because scientific theories also should have high explanatory power and high explanatory power is not, as Popper argued in Logik der Forschung (Reference Popper1994) the same as high degree of testability. This suggestion was the result of Agassi's examination of how research programs were used in the history of science. This was followed shortly thereafter by his application of Popper's philosophy of science to the history of science in his Towards an Historiography of Science (Reference Agassi1963). He also tackled the problem of the appropriateness of the second rule that was crucial in Popper's sociological turn, that is, that all basic statements should be testable, and thus repeatable, in his discussion of Boyle's rule. Agassi suggested that here too, independent testability would do. He has conducted further sociological studies of refereeing, studied and named the so-called Salieri effect—established thinkers often block the innovations of less established ones when the contributions of the latter threaten the social position of the former. For some time he has worked on the worthwhile project of finding the way to more openness and better criticism in scientific societies. I have elsewhere described his approach to the sociology of science and contrasted it with alternative approaches.Footnote ⁴ Agassi and others who have followed Popper's lead have long since treated science as an institution with good and bad aspects, sought to study them, and offered proposals for improvement. All of this is in accord with Jarvie's interpretation of Popper's social theory of science as developed in The Poverty of Historicism (Reference Popper1960) and The Open Society and Its Enemies (Reference Popper1962a). One of the most important aspects of this development, Jarvie says, was Popper's recognition that the institutions of science should not be idealized. They need to be studied sociologically, critically appraised, and actively preserved and improved by social reform. And all of these activities have to be carried out in the context of a normative philosophy of science.Footnote ⁵ Jarvie is right, of course, that it would be desirable to give the tasks of appraising this work and of bringing it forward a more prominent place of the agenda of contemporary philosophy of science, or, if that is too much to hope for, a more prominent place on the agenda of those who seek to build on Popper's revolutionary philosophy.