IT IS CLEAR to all readers of Popper’s work that there is some sort of natural affinity between the account he gives of the rationality of science and his commitment to political liberalism. The object of this essay is to explore the nature of that affinity. The claims I make about it are initially very uncontentious and hardly go beyond Popper’s own words; I end, however, by making the more contentious claim that Popper’s account of scientific rationality is itself in a broad sense political and that what sustains his commitment to some awkward epistemological views is his liberalism. That is, it is not so much that his philosophy of science supports his liberalism as that it expresses it. This is not a claim that I imagine Popper himself would accept; indeed, I imagine that he would be extremely hostile to it. Nonetheless, I should perhaps say at this point that it is not a claim made in any very hostile or critical spirit. Defenses of liberalism, like defenses of science, are almost doomed to waver somewhat between the thought that it is the process that justifies the result—a political decision or an accepted theory—and the thought that it is the result that justifies the process.
Since one of the things I wish to explore here is what kind of liberalism it is that Popper espouses, I shall follow his recommendation that we should avoid starting with elaborate definitions of our subject matter (Popper 1962, 2:17). So I start by observing only that liberalism must, whatever else, place a high value on liberty and equality. It is true that Ronald Dworkin has recently tried to persuade us that liberalism is concerned only with equality; but even he concedes that there will be liberal and nonliberal conceptions of equality and that the liberal conception of equality will be characterized by its concern for individual autonomy (Dworkin 1978, 127ff.). So I think we shall come to no harm if we start from the thought that liberals are concerned to achieve as much liberty and equality as may be.
Now, Popper remarks in his autobiography that he long ago came to the conclusion that although equality might be a good thing, it was excessively costly in terms of liberty (Popper 1976, 36). But here, it is clear, the sort of equality he has in mind is something like equality of wealth or income. In The Open Society and Its Enemies, Plato is attacked for his inegalitarian views not, of course, because he defends inequality of wealth or income, for in these terms Plato’s guardians are worse off than the rest of the population of the republic, but because he defends a kind of political elitism that is at odds with the sort of equality to which Popper is committed (Popper 1962, 1:94ff.). That is, there must be at least an equality of basic rights; we may not wish to try to secure that the most energetic and the least energetic end up with the same amount of wealth or the same annual income, but we shall almost certainly want to start by giving everyone the same rights to acquire wealth or income, and we shall certainly want to give everyone the same political rights and immunities (256–57).
In accordance with another of Popper’s own claims, we can agree that most liberals will see that these are values that we cannot pursue as “absolutes.” If, say, freedom was understood as the absence of authoritative control over our behavior, there is a strong case for saying that the creation of “absolute” liberty by abolishing all such control would be self-destructive by virtue of permitting the strong to tyrannize the weak without restraint. Similarly, if we were to start with the notion of equality as absolute equality of wealth or income, we should soon find that any attempt to create that equality would be destructive of other sorts of equality—in particular, the need to police whatever system of achieving equality we dreamed up would destroy any sort of equality of power (Popper 1976, 36–37). The idea that we can have absolute equality in the sense of equality in everything and in all ways is seemingly absurd, as is the idea that we can have absolute liberty; creating some equalities threatens others, just as leaving some liberties alone threatens the survival of others.
In this view, then, liberals will try to secure as much liberty and equality as possible; sometimes equality will strengthen liberty, and sometimes it will be in competition with it, just as sometimes liberty will reinforce equality and sometimes not. The kind of political equality that is enshrined in democratic procedures will, under favorable conditions, assist in the preservation of liberty by restraining rulers from trying to control every aspect of their subjects’ lives, but under unfavorable conditions, it may simply replace the tyranny of one man by the tyranny of the majority. There cannot be any conceptual demonstration that if only we pick the right conception of liberty and equality, we can secure that there is no competition between them and that we can have all we want of both—properly understood (cf. Dworkin 1978, 123–26). Or, more guardedly, any attempted demonstration of this compatibility is no use; we shall not know until after the event whether a given political choice embodied the right conception.
Given this minimal characterization of liberalism, it is easy to show how many of Popper’s concerns have been central to the concerns of mainstream liberalism. Consider The Poverty of Historicism. The object of that work is to show that social prophecies of the kind embodied in Marx’s social and political theory are no part of a rational program of scientific research. The idea that it is the task of the social sciences to uncover the laws of history is attacked with a variety of weapons, but chiefly with the claim that the “laws” on offer—say, Marx’s “law of increasing immiseration”—have not been laws at all, but trends (Popper 1961, 115–16). They have not been properly framed universal hypothetical statements, but extrapolations of conjunctions of singular statements. The defender of Marx might argue that this is by no means a fatal objection. If a trend could be shown to be properly grounded in true laws and the appropriate initial conditions, it ought in principle to be possible to show that the trend would continue—in some appropriate fashion—for an indefinite period. There are, perhaps, two slightly different ways of explaining why this will not do. One might be to construe laws as Mill did, as asserting that whenever A, then a, unless something interferes; human affairs, of course, amount to a continuous attempt to interfere with what would have happened but for the interference, so there will be very little predictive power to be had from laws of this sort in social matters. Of course, we can sometimes predict how people will interfere, and make allowances for it, but we cannot do so at anything beyond short range. A man setting a booby trap engages in ad hoc predictions about how people will try to interfere with the workings of his device; but he cannot hope to predict how long it will take for them to get wise to everything he does—all he knows is that since they will be trying very hard, he himself had better keep innovating. A second way of putting the point, perhaps truer to Popper’s own formulation, is to say that a law of the form (x)(Fx → Gx) may hold so long as people do not know of it; once they know of it, or perhaps even once they know of something else on which this law’s holding depends, the law becomes false (vi–vii). Again, the man setting the booby trap illustrates the point; the generalizations about the way people set out defusing a device on which he has hitherto depended turn out not to hold once their knowledge increases. The little semiformal demonstration of the impossibility of determinism that prefaces later editions of The Poverty of Historicism makes the central point: we cannot predict what we are going to discover in future, but we do know that what we shall do depends on those discoveries (v–vi).
It is in tracing the bearing of these arguments upon traditional conceptions of liberalism that we begin to see what sort of liberal Popper is. For in the liberal tradition, we may discern two distinctive approaches, one most clearly espoused by Mill, one most clearly espoused by Kant; and these would yield rather different interpretations of why the arguments of The Poverty of Historicism matter so much. Suppose there were some people, social scientific experts, who did possess the sort of far-ranging knowledge of society’s operations that we currently do not suppose they could actually have. What authority would this give them? On the difference between Kant’s reply to this question and Mill’s reply to the same question hang the two different liberalisms at stake here.
Kant’s answer is unequivocal; such knowledge would not give its possessors authority over us. Their position vis-à-vis us is rather like that of a “civilized” country proposing to colonize some “primitive” people. However enlightened we may be, we have no right to colonize other countries (Reiss 1970, 21). However much our experts may know, they have no right to our obedience unless we give them that right. In Kant’s formulation of the point, he appeals to the idea of a social contract, not as a historical fact, since he knows as well as his successors that no such event took place, but as an Idea of Reason—a methodological device to remind us that we can obey with a good conscience only those authorities to whose rule we could have consented in the appropriate circumstances (79). Popper’s hostility to “the myth of the origin” is such that it is hard to guess whether he would have much sympathy for what Kant was trying to do; at least, he would certainly want to rewrite Kant’s appeal to a hypothetical contract in noncontractual terms. But there is much in Popper to suggest that he would at any rate be sympathetic to the thought that each individual possesses a sort of moral inviolability that limits anyone else’s claim to authority over him.
Now, Mill, who was perhaps a more consistent liberal than a utilitarian, can accommodate something like Kant’s restrictions on the grounds of authority only by appealing to liberal ideals rather than individual rights, and at some cost to the supposed utilitarian foundations of his liberalism. That is, Mill’s claim was that a sufficient disparity in knowledge between ruler and ruled did amount to a license for despotism (J. S. Mill 1914, 73). Colonization was morally quite acceptable so long as its aim and effect were the improvement of the subject population. He was at least tempted by the appeal to expertise that ran through all of Comte’s work. To get back to something like liberal principles, Mill resorted to history; once a society was “civilized,” its people were improvable by argument; once people were improvable by argument, the difference between more and less informed ceased to matter. All claims to knowledge are fallible—a view plainly congenial to Popper’s criticalist account of science—and those who claim to know more than the rest of us are to be trusted only as far as they are willing to subject themselves to criticism from wherever it may come (80–81). Whether Mill’s appeal to fallibility is consistent with his generally inductivist approach to the growth of science is a moot point, but one we can leave on one side here.
More crucially, Mill had to appeal to the ideals of liberalism to defuse the claims of the knowledgeable. Liberty is one of the liberal’s supreme values: he therefore minds very much about the way in which the less knowledgeable obey the more knowledgeable. They must be led with their eyes open; they must give the best assent they can: unforcedly, freely, and on the basis of such information and argument as they can be given. To this extent, Mill might even have held that Popper’s appeal to “piecemeal social engineering” was dangerously misleading as an account of the politics of the liberal state. For it is characteristic of the inert, nonhuman material out of which bridges and automobiles are built that it has no view of its own, no values of its own, and could neither give nor withhold its assent to anything proposed for it. But what makes us human is precisely that we do have our own values, our own views, and a taste for self-government rather than heteronomy (J. S. Mill 1914, 114ff.).
Although I think that Popper’s liberalism is much more nearly Kantian than Millian, it cannot be said that the texts are absolutely conclusive on their very faces. For Popper does not develop an elaborate account of rights in the way that Hayek, for instance, does; even worse, from my point of view, Popper’s clearest view is that the best defense of the political decencies is a kind of negative utilitarianism. That is, the constitutional problem is not Kant’s problem—how to reconcile the freedom of each with the collective authority of all—but a problem more familiar in James Mill’s Essay on Government. Since those who hold in their hands the power to do the good things we desire from government also hold in their hands the power to do evil, how can we design institutions to minimize the evil (James Mill 1957, 50)? To the extent that this implies any very determinate position in political morality, it is that of a general humanitarianism, and a reminder that the miseries of the least significant member of society are miseries nonetheless and therefore to be taken account of. At first sight, it is the negative case that is doing all the work: like Isaiah Berlin, Popper is frightened of theorists who believe that if only the right leader and the right doctrine coincided, everything would go perfectly (Popper 1962, 1:120). Like Berlin, he thinks there are no indisputable criteria for leadership, no special knowledge or special qualities of character to which they can lay claim. In this sense, liberalism is the product of skepticism even more than of a theory of inalienable rights, or of a commitment to the ideals of liberty. Nonetheless, I think the beginnings of the positive case for liberalism are also to be found here—a positive case that is built up along with Popper’s description of the practice of scientific research.
But one feature of the skeptical case also suggests the Kantian affiliation of Popper’s politics. Popper insists on the logical distinction between facts and decisions—his version of Hume’s famous distinction between is and ought (Popper 1962, 1:60–61). This makes him more nearly Kantian than Millian on the connection between knowledge and authority. Mill seems to envisage superior minds knowing more than the rest of us not only about economics and sociology but also about what constitutes true elevation of character and the like. He is at least insecure about whether there can or cannot be moral experts, and this has given his critics a field day, with the more energetic claiming that his liberalism was the defense of an authoritarian secular enlightenment (Cowling 1962). Kant, however, had no doubt that moral ideals were something that each person had to accept for himself. Technical advice could be proffered ad lib, and if we were sensible, we would take the best advice we could get. But it was tyranny of the worst kind to force others to adopt our conception of their good or to force upon them our moral ideals.
It is central to Kant’s thought that the state enforces external rules, which can be obeyed without further moral consideration, and that the authority of the state is limited to enforcing these. It is why property rights matter so much to Kant; through property rights and their cognates, we can deal with each other coercively but at arm’s length (Reiss 1970, 135). I promise you my bicycle for your stamp collection; once you have handed over the stamps, I must hand over the bike; if I do not, you may properly force me to. What you and the state may not do is demand that I hand over the bike in any particular state of mind. It is my external performance only that is subject to control. This marks a concern to distinguish what is public, nonmoral, subject to law, and external from what is private, internal, and moral. There is no such concern in utilitarianism; one of the oddities of Mill’s Liberty, indeed, is that there is no mention of the distinction between public and private.
If this is right, the Kantian flavor of Popper’s liberalism is not accidental. Further evidence for this comes in the conjunction of his uncertain attitude to democracy and his negative utilitarianism. Popper’s defense of democracy is odd in that it is not based on what one might call the positive moral attractions of the democratic formula. Against Plato, with his contempt for the ordinary man, and his mystical faith in the qualities of the guardians, Popper certainly defends democracy, in the sense that he defends the rights of the common man to a say in the government of his country, and attacks the guardians’ claims to any such moral and intellectual infallibility as their authority is supposed to rest on. Moreover, says Popper, if the common man does not have the power to throw out rulers he wishes to throw out, he will be mistreated; all experience shows the truth of Lord Acton’s dictum about the corrupting effect of power (Popper 1962, 1:136–37).
The justice of this as an attack on Plato is, of course, open to dispute. If one makes the first concessionary step of agreeing that Plato’s intellectual universe makes sense, then the claim that experience shows that power corrupts will not seem quite such a knockdown argument; Plato’s reply is that, of course, power has always corrupted in the past because philosophers have not been kings and kings have not been philosophers. To my mind, this response really shows something else—that Popper was right to launch a wholesale attack on Plato; the aim cannot be to show the internal incoherence of Plato’s case, but to show that this whole approach to politics is misguided from the ground up. Then, the defense of democracy can be made in the casual way Pericles makes it in The Peloponnesian War (Popper 1962, 1:77). In essence, it is the old claim that the wearer of the shoe knows whether it pinches; not everyone can aspire to formulate policy, but what everyone can do is say whether the policies dreamed up by others are painful in their consequences.
There is more to it than that, of course. For one thing, what Popper and Pericles are eager to defend, and what Plato hates, is variety, vivacity, experiments in living. What Plato denounces as mere ignorant and childish running about after one fancy after another, Popper and Pericles praise as an attractive and enjoyable feature of the life of democratic Athens. As a defense of democracy in the narrow sense of the rule that decisions should be taken by some sort of majority vote, this would all be irrelevant, of course; but Popper is not concerned to defend that rule, since he knows as well as the next man that it can produce ugly results and paradoxical ones. The defense is more nearly a defense of toleration, what you might call a democratic view of cultural and moral issues, to the effect that the ordinary man has a right to live as he pleases within the limits of the law, and so long as he neither disturbs the peace nor prevents others exercising a similar freedom (Popper 1962, 1:186–87).
If this is plausible, it reinforces the point that this is a Kantian liberalism, since Kant’s defense of the individual’s rights is couched in very much the same terms. It offers a limited defense of democracy in suggesting that individuals have a right to a say in the affairs of their country, and, more importantly, that they are not likely to preserve their other rights unless they have some such say.
Primarily, however, it is a defense of individual rights, and only secondarily a defense of a decision-making formula—which is one reason why Popper is more a constitutionalist than a democrat, just as Kant was. The thought that this is the right lineage of Popper’s views is not unduly disturbed by observing that Popper and Pericles side with Mill in defending experiments in living (J. S. Mill 1914, 115). For Kant was every bit as concerned as Mill to defend such experiments. Kant’s conception of the hidden point of history, it may be recalled, is that the goal of history is the development of all human faculties to their highest point. Of course, there is no empirical proof that this is the point of history; it is another of the Ideas of Reason that we employ to make sense of human experiences. We might be anxious about the difference between Mill’s defense of variety for its own sake and Kant’s notion of a historical task; but, of course, it is axiomatic in Kant’s system that we have a right to harmless enjoyments—we would not have assented to a government that could curb them—and it is implicit in his slightly awkward relation to his scholastic predecessors that he believes that the universe can contain as much good as possible only if it is as varied as possible too.
But it is at this point that Kant’s liberalism and Popper’s view of science begin to touch. Kant’s defense of an evolutionary but nonutopian picture of human history partly rests on the claim that out of the crooked timber of humanity no straight thing is made; but mostly, it rests on the view that nature sets us a series of problems, the solutions to which set us new problems, and so indefinitely on. Kant’s view that “progress” is the goal of history is a sort of transcendental hypothesis, in something of the same way that Popper’s conviction that increasing verisimilitude is the goal of science is a transcendental hypothesis. And one implication of this is that each individual who shares in the growth of knowledge or the development of human capacities may contribute only a little to the development of the whole species, but can, nonetheless, find a meaning to his life in so doing (Reiss 1970, 42–43). One ought not to be unduly surprised that two thinkers who adamantly reject the “bucket” theory of the mind and insist on the active role of the intelligence in interpreting experience should share more than epistemological allegiances.
If we place experiment, social progress, and fallibility at the center of our politics, we have good reason to support a strenuous constitutionalism, which guarantees each man the maximum liberty consistent with the liberties of others. Democracy—in some sense other than conferring absolute power upon a majority—is one expression of this sort of constitutionalism, since anything other than equal political rights seems to flout the view that everybody has the same claim as everybody else not to be under anyone else’s control without good reason. There is, of course, a more obviously utilitarian case to be made for democracy too, namely, that unless people can voice their views, even a well-meaning government may be ineffective for lack of information about what people want. But we must recall Popper’s view that the problem of government is not to secure that good men have the information to do good, but to secure that the nonrulers can throw the rascals out. Democracy, conceived as a device for the circulation of elites, is an answer both to the question of how to select competent rulers—by getting rid of the less competent ones peacefully—and to the question of how to retain power to throw the rascals out (Popper 1962, 1:124–25). The reason why I mention the consistency of this defense of democracy with the “elite theories” of writers like Vilfredo Pareto and Gaetano Mosca is to further draw out the affinity between Popper’s conception of democracy and his conception of science. In science, the essential is that no hypothesis can be protected by an appeal to authority. It does not follow that every scientist is equally good at testing old hypotheses or equally imaginative in suggesting new ones—let alone at carrying out tests of new ones. What does follow is that the merits of a hypothesis are a matter of what happens to it under test, not a matter of who suggests it. In a democracy, not all ideas are equally good, and many policies are quite silly; nonetheless, we ought not to restrict the right to propose policies, and we must always remember that the merits of a policy are independent of the social or academic status of the proposer of it.
The way Popper’s defensive view of democracy and constitution building place him in the Kantian tradition emerges even more clearly when we turn to his negative utilitarianism. His defense of this ethic is not elaborate or sophisticated—he has nothing much to say to the observation that a moral theory devoted to misery minimization would prescribe an unexpected and painless death for the whole of the human race. But this, too, is significant. Mill, whose On Liberty claims in passing that governments do best when they stick to harm prevention, is aware that his official theory of government holds that governments ought to do whatever maximizes the general welfare; he sees that he needs to claim that liberty is an essential element in happiness, and that misery is more readily identified and more easily dealt with, if he is to square negative utilitarianism in practice with positive utilitarianism in theory (J. S. Mill 1914, 132–33). Hence the complexities in his case that have irritated and excited critics for a hundred and thirty years.
Another view of the matter achieves the same results less awkwardly. Kant held that we simply had no business making other people happy. Governments might hinder hindrances to the search for the good life; but that was all (Reiss 1970, 134). Of course, all this becomes very much less simple when we probe it, and the line between hindering hindrances and positively promoting happiness in a paternalist fashion is harder to draw than it looks. All the same, the point of the doctrine is clear enough. Whereas Mill starts from the view that governments may legitimately do anything that promotes the general welfare, and then erodes that view in order to find room for individual liberty, Kant credits us with a right of self-defense and allows governments to do only those things that help us to help ourselves without attacking others.
I do not wish to exaggerate the definiteness of Popper’s case here, nor the clarity of its lineage. Popper is plainly at one with Mill in thinking that one reason for pursuing a negative utilitarianism is shortage of information. We are much more readily able to discover what makes people unhappy than what makes them happy, and on the whole, what makes people unhappy is more uniform than what makes them happy; it is, also, quite often easier to cure. (This is, of course, very far from being universally true, and it is simple enough to think of exceptions.) Painters, poets, and pianists all suffer from cold and hunger, but a piano is no use to a painter, nor an easel to a poet. The state may properly engage in the provision of a range of basic welfare services, but is ill advised to do much more. It is not a major concern of Popper’s to draw sharp lines around the proper tasks of the welfare state, and it would be wrong to try to extract too much from his argument (Parekh 1982, 146–53). All the same, it is perhaps worth noticing that the slightly blurred argument here, and the uncertainty about quite what tradition the argument belongs to, is characteristic of writers who have made much more of it than Popper. Hayek, for instance, sometimes employs utilitarian arguments against welfare-state utilitarianism—or welfare-state “liberalism,” in the American sense of the term; and sometimes, he employs arguments of a more strictly rights-based kind (Hayek 1944). In one view, the ambitious state fails in its purposes because its information is inadequate and its techniques clumsy; on the other, positive steps to maximize social welfare involve illegitimate coercion in forcing members of society to part with goods and services they do not choose to part with, and for purposes other than self-defense.
I conclude, then, that Popper’s liberalism is more nearly in a Kantian than a Millian mold, that it is a genuine liberalism in defending constitutionalism above a populist form of democracy, and that even where it is eclectic in its sources, it is classical liberalism in its concern for individual inviolability (Popper 1962, 1:99–104). I turn finally to the task of showing that the connection between Popper’s defense of liberalism and his view of science is partly a matter of instrumental and sociological arguments, but in part is simply a matter of pointing out the liberal virtues of the ideal scientific community. The instrumental case is relatively straightforward. Science seems to entail an open society, and an open society is indispensable to science. That is, if we manage to establish the scientific enterprise in any solid form, its achievements are sure to have an impact on society at large. We cannot believe one thing at work and another at home; we cannot study biology in the lab and heal ourselves with witchcraft at home. The degree to which conceptual incoherence is cheerfully tolerated by many people may be much greater than the out-and-out rationalist would like to believe, and the contents of our minds are not entirely transparent to us—whence, of course, Popper’s own insistence on the need to institutionalize the pressure to think clearly and articulately. All the same, there will be a steady pressure to square the beliefs of society at large with the findings of science. This, however, is just what makes an open society: beliefs are not locked solid, rendered immune from investigation and challenge. If we live in a society with a flourishing scientific community, we shall find ourselves in a society with a pressure toward openness (2:217–22).
This, again, is a view that Popper seems to share with both Kant and Mill, and it is, of course, a contestable view; we are familiar with colleagues who seem to subscribe to one cosmology in the laboratory on weekdays and to another in church on Sundays, and no doubt there were physicists who were committed Nazis and no worse physicists for it. My own view is that a closed totalitarian society would stagnate scientifically after two generations and would never do better than replicate work done elsewhere—but that is a hunch rather than a testable sociological hypothesis. What one can surely say is that no closed society can allow its scientists to chase hunches wherever they may lead; if the license to hunt is implicit in science, closed societies cannot practice science. Conversely, it is at least arguable that science needs an open society as background. The willingness to challenge established theories, the ability to go out on a limb without undue anxiety, the combination of eager questioning and patience to wait for answers, all seem to require the social training provided in a liberal society. The scientific community is one full of people who have passionately held convictions but who need to resist the temptation to close ranks or shut up dissenters; the better the training in self-restraint provided elsewhere, the better the science. Again, the argument is contestable. Writers like Kuhn (1962), who emphasize the closed and authoritarian nature of scientific communities, cannot be lightly dismissed. The violence of Popper’s reaction to Kuhn suggests that he cannot bear to entertain the thought that liberalism and science may be systematic enemies rather than allies (Popper 1970, 51–59).
Kuhn, of course, represents himself as a modified disciple of Popper, but the point is still the same. Even if Kuhn’s claim is only that scientists need to organize themselves in a more hierarchical and authoritarian fashion than Popper supposes if they are to carry out the eminently Popperian task of rigorously testing their conjectures about nature’s laws, science would cease to be the natural ally of liberalism. Kuhn emphasizes that among scientists, ideas are not accepted regardless of who puts them forward; only some people are authorized to change the current orthodoxy, and then in orthodoxy-conserving ways. If that is inescapable, we must accept that the habits of free speech and open discussion that characterize liberalism will be irrelevant to science. Conversely, if great stress is laid on the difference between what scientific claims mean to professional scientists and what they may fail to mean to laymen, the idea that there will be a sort of leakage from science to society at large that forces laymen to adopt new ideas and new ways of thinking will also be false.
In the argument between Popper and Kuhn, one could pursue a conciliatory course by pointing out that Popper stresses what Kuhn never denies—that nature tests our hypotheses, whoever devises them—and that Kuhn stresses what Popper never denies—that scientists need tenacity as well as a willingness to change their minds, that some attempts to protect hypotheses against premature falsification are justified, and that a scientific community needs some taken-for-granted ways of deciding where to direct its efforts (Kuhn 1962, 231ff.). Still, the story I wish to advance is at odds with this conciliatory intention. It starts with a familiar epistemological point. Many of Popper’s critics have complained that once Popper’s account of the process of conjecture and refutation is deprived of any residual empiricist or positivist elements, Popper can no longer tell us what hypotheses are tested against. Popper insists on a correspondence theory of truth and also insists that there can be no statements that are simply known to be true by virtue of corresponding to the facts. The earliest version of Popper’s account of science seemed to suggest that the asymmetry between justification and falsification was important because we could know beyond doubt that some singular proposition was true, and use this as bedrock in testing general claims. This view is now said not to be Popper’s own. All statements are revisable; all statements are interpretive; and general statements may be saved by numerous different strategies, some of which are quite justified.
The claimed virtues of science now become rather harder to elucidate because the idea that scientific progress amounts to increasing verisimilitude is now more than ever an Idea of Reason in the Kantian sense, and the way is open for a much more “sociological” account of science. That is, once it is accepted that we have a great deal of choice about what statements to accept or reject as part of the corpus of science, the pressure is on us to give an account of what amounts to good and bad choosing. The “scientificity” of scientific doctrines becomes a matter of the process by which they are chosen rather than a simple matter of their factual truth. The obvious instrumental view that the goodness of the process of choice is simply a matter of the effectiveness with which the process eliminates false beliefs and replaces them with truer ones cannot be sustained in any simple way once all ties with empiricism have been cut. At best, the claim that what the progress of science achieves is theories that are increasingly true to the facts becomes a metaphysical hypothesis, a heuristic maxim: “Act as if it were true that theories mirror facts.” At worst, Popper finds himself in the company of critical theorists such as Jürgen Habermas, whose notion of truth is parasitic on what people would agree to say under conditions of freedom to say what they liked, equal ability to contribute to the consensus on what to say, and some sort of pressure toward reaching such a consensus (Habermas 1970, 360 ff.). It is going too far to suggest that Popper has arrived at the point of saying that the discoveries of science just are what the ideally liberal scientific community agrees to say they are, but it is not an exaggeration to suggest that he is under the same pressures in that direction as writers like Habermas have been.
I will not try to show what the consequences of this might be for Popper’s defense of objectivity, though I cannot forbear remarking that the sort of tension I have pointed to seems to account for the critical anxiety that Popper’s appeal to “World 3” induces. I shall end by emphasizing what I take it to show. It is not just that Popper espouses a rather Kantian liberalism; and it is not just that he thinks that the scientific frame of mind nourishes, and is nourished by, an open, liberal society. It is that the same Kantian picture of what it is like to try to make sense of the world that permeates his philosophy of science means that the bedrock reason for approving of science as an activity is that it is intellectual inquiry carried on by liberal means. To put it another way, it is the politics of scientific inquiry that makes it admirable, and the defense of political liberalism is not so much supported by an appeal to science as it is simply another part of a seamless web. This conclusion remains an uncomfortable one to square with Popper’s hostility to what he thinks of as subjectivism, and I do not seek to disguise that. I will, however, say that it is not uncongenial to Kant’s defense of the Enlightenment, with its cry of “sapere aude,” and not uncongenial to Popper’s own defense of the Enlightenment and his hostility to the irrationalism of Wittgenstein’s later philosophy, with its defense of forms of life and its claims for the inescapability of current and local habits of belief (Reiss 1970, 54).
I am grateful to Bryan Magee, John Watkins, and seminars at Oxford and Reading for their reactions to earlier versions of this essay. In two respects I feel that I have not been able to do justice to them. Watkins thinks Popper mistaken in relinquishing the “empiricist” elements in falsificationist methodology, and thus that I follow too readily down a road that leads nowhere. My only response is that I am impressed by the similarities between Popper’s problems and those of Kant and think the comparison illuminating. I am inclined to share Watkins’s doubts about Popper’s antiempiricism—but that would have been a different essay. My question here was not so much whether that antiempiricism was right, but whether the connections I suggest between Popper’s philosophy of science and his political liberalism would hold up if (or even if) his antiempiricism were right. Christie Davies (after the Reading seminar) suggested that I should have been sharper about the distinction between a liberalism founded on humanitarian concern and one founded on rights. He instances debates over abortion as occasions where one kind of liberalism is worried by the conflict between the rights of the mother and those of the unborn child, whereas the other worries only whether humanity to unwilling mothers carries unacceptable costs in other directions. I agree, of course, that these different forms of political argument have characterized debate in Britain and the United States. But Popper is not a natural-rights theorist nor simply a rule-of-thumb utilitarian, so too sharp an insistence on the distinction would have meant pressing his writings for answers to questions they do not raise. I am inclined to say that like Kant, Popper defends legal, constitutionally guaranteed rights not because we just do have (natural) rights, but because they are the legal expression of something like Kant’s principle of the inviolability of the individual. I must, however, admit that I am not sure that I could infer from what Popper has written on particular issues what he might say on those about which he has not written.
Cowling, M. 1962. Mill and Liberalism. Cambridge: Cambridge University Press.
Dworkin, R. 1978. “Liberalism.” In Public and Private Morality, edited by Stuart Hampshire. Cambridge: Cambridge University Press.
Habermas, J. 1970. “Towards a Theory of Communicative Competence.” Inquiry 13:360–75.
Hayek, F. A. von. 1944. The Road to Serfdom. London: Routledge and Sons.
Kuhn. T. S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press.
Mill, James. 1975. Essay on Government. Indianapolis: Bobbs-Merrill.
Mill, J. S. 1914. Utilitarianism, Liberty, Representative Government. London: Dent.
Parekh, B. C. 1982. Contemporary Political Thinkers. Oxford: Martin Robertson.
Popper, K. R. 1961. The Poverty of Historicism. 2nd ed. London: Routledge and Kegan Paul.
———. 1962. The Open Society and Its Enemies. 4th ed. 2 vols. London: Routledge and Kegan Paul.
———. 1970. “Normal Science and Its Dangers.” In Criticism and the Growth of Knowledge, edited by I. Lakatos and A. Musgrave. Cambridge: Cambridge University Press.
———. 1976. Unended Quest. LaSalle, Ill.: Open Court.
Reiss, H., ed. 1970. Kant’s Political Writings. Cambridge: Cambridge University Press.