Karl Popper, critical rationalist.

"IT'S NOT what people don't know that gets them into trouble," an American humorist once said, "it's what people know that ain't so."

Perhaps the most difficult problem in the history of philosophy has been: How do we know that what we know is so? Sir Karl Popper's startling solution to that problem (we don't) made him, at the time of his death last year at 92, the most influential recent philosopher in the world -- perhaps the most influential since Kant. His work was mainly in two cognate branches of philosophy: epistemology (the theory of knowledge) and the philosophy of science. Despite its theoretical character, Popper's work profoundly affected how we think not only about science and philosophy, but also about politics, art, and religion. Indeed, so powerful has been his influence (except among professional philosophers) that most of us don't even realize how much our thinking owes to him. No assessment of Popper's ideas can begin other than with his early, purely philosophical work. However, it was his landmark books The Poverty of Historicism and The Open Society and Its Enemies that brought his ideas to wider public attention. This year's fiftieth anniversary of the publication of The Open Society and Its Enemies provides the ideal occasion for an assessment of those ideas.

In the traditional view, knowledge was sharply distinguished from belief. Individuals might hold opinions with varying degrees of belief. They might even hold some opinions with absolute (subjective) certainty. But unless their truth could be rationally justified or proved, those opinions, regardless of the (subjective) certainty with which they were held, constituted not knowledge, just belief. At least since the Renaissance, it came increasingly to be accepted that scientific knowledge was unique in being rationally justified and validated. And especially after the astonishing success of Newton's theory of gravity, not only in explaining the observed movements of the heavenly bodies but in leading to the discovery of the previously unknown planet Neptune, science acquired an exalted status -- the exemplar of confirmed, positive, and certain knowledge. It became the role of philosophy not to discover truth but to judge which scientific beliefs could be rationally justified.

Ultimately only two methods of rational justification seemed tenable: 1) logic (deduction), and 2) observation and experience (induction). But even these two methods were found to be severely limited. Deductive reasoning, by itself, can provide us no knowledge of the real world. All it can provide us are tautologies -- the implications of a given set of definitions and premises, for example those of Euclidean geometry. But there is no way to determine a priori whether the implications of any set of definitions and premises correspond to the complexity of observable reality.

The principle of induction seemed more promising. According to this principle, researchers could, by repeated observation, establish (confirm) a universal law (e.g., the sun rises in the east every morning). But by the middle of the eighteenth century David Hume had shown that induction is logically inadequate to its task. No number of observations, Hume pointed out, can confirm a universal law, for no number of observations can exclude the possibility that a contrary future observation may invalidate it.

After Hume, philosophy was faced with a glaring contradiction between the incapacity of induction to justify the knowledge claims of science and the indisputable success of science in explaining an ever-expanding range of observable phenomena. Kant tried to resolve the problem by arguing that scientists do not discover universal laws embedded in observable reality, but create them a priori and then impose them, as it were, on reality. But the Kantian view of science collapsed early in this century when the Michelson - Morely experiment supported Einstein's theory of relativity and disproved Newton's theory, showing a priori claims of validity to be as insecure as inductive claims.

The philosophical need to reconcile the unique epistemological position of science with Hume's disproof of induction and with Kant's failure to circumvent Hume's disproof led to the rise in the early twentieth century of logical positivism. Positivists distinguished between two kinds of factual statements: scientific statements, which at least in principle could be confirmed by some observation, and metaphysical statements which could not be confirmed, even in principle, by any observation. The difference between scientific statements and metaphysical statements, the positivists argued, is not that scientific statements are true and metaphysical statements are not, but that scientific statements have meaning and metaphysical statements have none.

Here is where Karl Popper enters the story, for in The Logic of Scientific Discovery he rejected the positivist criterion of demarcation between meaningful (scientific, verifiable) statements and meaningless (metaphysical, unverifiable) statements. His critique was twofold. First, he denied that potential for verification is what distinguishes science from non-science. On the contrary, Hume had shown that, as a matter of logic, a universal law of science could never be verified. The positivists' own criterion thus rendered all universal laws of science meaningless. Popper turned the positivist doctrine on its head by arguing that what characterizes a universal law of science is that it forbids certain events. Should those events occur nonetheless, the law is refuted.

It is noteworthy that Popper's conception of science stemmed largely from his early interest in psychology and socialism in post - World War I Vienna, whose intellectual and political life was dominated by the Freudians and the Marxists. Both groups claimed a scientific understanding of individual human conduct and of society. And both insisted that their doctrines were confirmed by observations derived from clinical experience, from the annals of history, or from common everyday experience. But Popper found that both groups had, in fact, insulated their theories from empirical criticism. The Freudians boasted that psychoanalysis explained any conceivable behavior, so that all clinical experience constituted evidence in its favor. Popper's response was that if no conceivable observation could contradict psychoanalysis, the proper inference is not that psychoanalysis is confirmed, but that its empirical (and, hence, scientific) content is nil.

Marxists avoided criticism in a different way. Acknowledging that Marx's predictions had been contradicted by events, the Marxists nevertheless dogmatically defended their theory by advancing special ad hoc arguments to explain away those contradictions. Popper concluded that Freudians and Marxists both regarded their theories not as scientific hypotheses susceptible of criticism and possible refutation, but as dogmas immune to criticism.

After Popper advanced his claim that the characteristic of scientific theories is their vulnerability to refutation by observable events, some logical positivists proposed reformulating their criterion of meaning in terms of the refutability of statements rather than in terms of their verifiability.

But Popper's second criticism of positivism blocked that line of retreat, for he expressly denied that statements are meaningful only if they could be refuted by some observation. Popper remarked that the positivists' own criterion of meaning could be neither verified nor refuted by any observation and therefore, under their criterion, had no meaning. Indeed, how would one determine whether a statement could be verified or refuted without first comprehending its meaning? The criterion of refutability therefore could not distinguish between meaningful and meaningless statements, only between scientific and non-scientific (or metaphysical) statements. And trying to account for the specialness of science by declaring non-scientific statements to be meaningless, Popper realized, would inevitably lead to a futile preoccupation with definitions and semantic issues.

Popper's critique of logical positivism not only rescued philosophy from the positivists' self-destructive preoccupation with meaning, it revolutionized our view of science. Accepting Hume's critique of induction, Popper readily acknowledged that there is no method (scientific or otherwise) of justifying our theories about the real world. What sets science apart from other forms of intellectual inquiry is that science subjects its theories (i.e., its tentative claims to truth) to observational tests that could disprove them. When competing theories have contradictory implications about the world, scientists prefer that theory which is most consistent with observable reality. Because it aims at theories whose implications correspond with observational reality, science progressively weeds out inferior theories. And even though we can never be certain that the theories we now maintain are true, they are closer to the truth than past theories were.

Before Popper, rationalists had dismissed unjustified beliefs as nothing more than superstition. Popper, however, argued that it is criticism, not justification, that is the hallmark of rationality. Reason, for Popper, is a critical faculty, and the characteristic of all rational discourse is the exercise of that faculty. And while metaphysical theories, unlike scientific ones, cannot be empirically tested, they can still be discussed critically.

IF SUSCEPTIBILITY to criticism is the characteristic of rationality, then arguments based on meaning, or arguments by definition, are inherently suspect. At best futile, such arguments degenerate, at worst, into a tactic for avoiding criticism. How we define a word describing a real object or phenomenon is a matter of convenience. Reality can't be altered by how we choose to define words, and definitions can't be right or wrong -- only more or less useful in the light of a particular problem we are addressing.

Thanks largely to Popper, the once pervasive belief that definitions really matter has ebbed considerably. Popper's distinction between substantive arguments and semantic ones is now widely understood and accepted. Indeed, exposing an argument as being merely semantic now often suffices to put it to rest. But that is not to say that the old view about definitions does not often still hold sway. A famous example of such misplaced emphasis on definitions was the reaction to the remark made by Justice Potter Stewart in his concurring opinion to the Supreme Court's decision in Jebollis v. Ohio (1964). Justice Stewart wrote that while he could not define hard-core pornography, "I know it when I see it and the motion picture under review in this case is not that." Ever since, that frank statement has been derided by law professors and other sophisticates (usually those objecting to any restriction on obscenity) as an example of muddleheadedness.

Notice how, by suggesting that unless obscenity can be defined it doesn't exist, a response focusing on the lack of a definition for obscenity arbitrarily stifles discussion about the real issues posed by legal restrictions on obscenity. The problem is not how to define obscenity, but whether the state has a legitimate interest in banning obscene materials or restricting their distribution. Considerations might include the possible harmful effects of such materials on viewers, on those involved in their creation or distribution, or on the community at large. Alternatively, there might be moral grounds that would give the state an interest in disallowing the creation and distribution of obscene materials. Having surveyed the possible reasons for banning or restricting obscenity, one could then consider whether individuals have legal rights to read and view what they choose that could take precedence over the possible social harms or moral offenses perpetrated by the creators and distributors of obscene materials. Instead of engaging in a critical discussion of the issues, those deriding Justice Stewart sought to solve the problem by defining it out of existence.

APPLYING his conception of the growth of knowledge through criticism, Popper achieved two fundamental insights in political philosophy. The first, which he worked out shortly after completing his basic work on the philosophy of science in the early 1930s, was his discovery that all alleged historical laws of development, from Hegel and Marx to Spengler and Toynbee, suffer from an inherent logical fallacy. The fallacy becomes unmistakable as soon as an obvious fact is recognized: that the future condition of society, which theories of historical development claim to predict, depends on the future state of scientific knowledge. Thus, any law governing the future state of society entails a corresponding law governing the future state of knowledge. But if the future state of knowledge is determined by a known law, then it must already be known, which is self-contradictory. If future knowledge were already known, it would be not future, but present, knowledge. And if future knowledge must, as a matter of simple logic, be unpredictable, so must the future state of society. All purported laws of historical development are thus logically untenable.

Popper developed this idea in his book The Poverty of Historicism, which grew out of a lecture that he gave in 1936 to F. A. Hayek's seminar at the London School of Economics. Within a year, Popper, anticipating Hitler's imminent takeover of Austria, left for New Zealand to take his first academic post, at Canterbury University College. Resolved to publish his work showing the common historicist tendencies of Communism and fascism, Popper found himself, in developing his logical refutation of historicism, tracing the roots of historicist and totalitarian ideas all the way back to Plato. Realizing that an adequate historical treatment of the development of these ideas from Plato to Hegel and Marx would require a separate and much longer work, he simultaneously began writing his best-known book, The Open Society and Its Enemies.

The Open Society and Its Enemies addresses the age-old conflict between those who believe that they are possessed of a truth or of knowledge that is beyond criticism and discussion -- and who therefore seek to impose their conception of the truth and their will on the rest of society -- and those who believe that the way to arrive at truth is through open debate and discussion. The latter believe in an open society in which no individual or group claiming a monopoly over truth can enforce its beliefs on others, while the former are the enemies of the open society.

Working under the burden of a desperately heavy teaching load and the active hostility of his university administration to his research and writing, Popper nevertheless persevered. Even after finishing his two books, he faced further discouragement. In 1943, the philosophical journal Mind rejected the manuscript of The Poverty of Historicism for publication, and the early readers of the manuscript of The Open Society and Its Enemies were distinctly unenthusiastic, in part because of its attack on Plato and in part because of its attack on Marxism. In desperation Popper sent the manuscripts to his Viennese friend Ernst Gombrich, the art historian. Gombrich, now in London, had attended Popper's 1936 lecture in Hayek's seminar, and he enlisted Hayek's help in securing publication for the manuscripts. Hayek himself accepted The Poverty of Historicism for publication in the journal Economica, of which he was the editor, and he found a publisher for The Open Society and Its Enemies. A year later, Hayek arranged for Popper to be offered a post at the London School of Economics.

Both born in Vienna around the turn of the century, raised in the educated Viennese upper middle class, drawn to left-wing politics and the study of psychology as undergraduates at the University of Vienna after World War I, Popper and Hayek, who had never met then, thus became close friends in middle age. They found they had arrived at virtually the same conclusions about rationality, scientific method, freedom, and political philosophy.

Both Popper and Hayek argued that all totalitarian systems were irrational in a very basic sense. Hayek emphasized the inability of a central authority to obtain the information needed to allocate rationally the resources it commanded. Popper's point was that by decreeing its own omniscience and suppressing all criticism of the official creed and of official decisions, the ruling elite would be suppressing the very activity, critical discussion, that is essential for knowledge to grow. These two aspects of the irrationality of totalitarianism virtually dovetail into one. The inability of a ruling elite to allocate resources rationally must cause obvious errors that evoke criticism as a matter of course. Such criticism, if accepted, might improve conditions. Instead, the ruling elite suppresses criticism, which makes learning, and any improvement in conditions, impossible.

Popper's clear-eyed recognition of the horror of totalitarianism of whatever stripe earned him the reputation of a conservative despite the socialist and even Marxist leanings of his youth. Those leanings were still detectable in The Open Society and Its Enemies, in which, notwithstanding his attack on Marxism and his opposition to any system of comprehensive state control over the economy, he advocated "piecemeal social engineering," commenting favorably on the Swedish Social Democracy and the American New Deal. But gradually, presumably under Hayek's influence, Popper softened his attitude toward laissez-faire capitalism and narrowed his interpretation of the sort of state action that might legitimately be undertaken in the name of piecemeal social engineering.

EVEN SO, in his hostility to all forms of nationalism, his embrace of the values of critical rationalism and science, his rejection of authority, Popper never became a conservative in the mold of, say, Russell Kirk. Nevertheless, there were few more sensitive than Popper himself to the limits of reason. Reason operates primarily negatively, by criticism and refutation. And just as reason can offer no certainty in science, it offers none in morality, justice, or politics. Popper rejected, in particular, the notion that reason can identify basic incontestable principles -- principles either of nature or of morality from which valid scientific or moral conclusions may be safely derived.

All scientific and moral theories are in the nature of conjectures that can be criticized. Scientific theories are subject to criticism based on correspondence with observable facts. In morals no comparable empirical method is available, so any improvement in our moral understanding that may have occurred has been of a far more tenuous nature than that of the growth of our scientific knowledge. The naive rationalist hope of deducing all moral values from self-evident basic principles is illusory. Our moral values are, in the end, derived from a tradition that can never be justified in purely rational terms. The role of reason is to help us understand our moral tradition and perhaps, in the process, to improve it, not to replace it with a new moral system derived from self-evident principles. This fundamentally conservative implication of Popper's epistemology may turn out to be no less important for the philosophy of morals than it has been for the philosophy of science. Should it turn out to be so, most conservatives -- and, for that matter, most liberals -- would have yet another reason to be thankful for the life and work of Karl Popper.