A Guided Walkthrough of Popper’s Core Philosophical Works: The Logic of Scientific Discovery, The Poverty of Historicism, and Conjectures and Refutations
Section One: The Problem of Induction and the Demarcation of Science
We begin at the opening of Popper’s project, where he establishes the central problem that his philosophy of science is designed to solve: the problem of induction. Popper locates this issue within the tradition of Hume, who had shown that induction—the inference of universal laws from particular observations—cannot be logically justified. Popper presents the situation clearly: no number of black ravens can logically prove that all ravens are black, because a single counterexample destroys the universal claim. This is the logical gap between finite observation and infinite generalization.
Popper then introduces his first major move: the demarcation problem. He argues that the traditional view of science as “empirical” because it uses inductive methods is fundamentally mistaken. The real problem is not to justify induction, but to find a criterion that distinguishes genuine science from pseudo-science. Popper points to Marxism and psychoanalysis as examples of theories that, despite their explanatory power, fail this test. He writes that these systems “appeared to be able to explain practically everything that happened within the fields to which they referred.” This very ability to explain everything, Popper argues, is a sign of weakness, not strength.
What matters here is Popper’s reversal of the traditional relationship between observation and theory. Instead of beginning with observations and building up to general theories through induction, Popper insists that science begins with problems and bold conjectures. The scientist proposes a hypothesis, and then attempts to refute it through severe testing. A theory is scientific if it makes predictions that could potentially be falsified by observation. This is the criterion of falsifiability: a statement or system of statements is scientific only if it is logically possible to show it false by some potential observation. Popper’s key claim is that “the criterion of the scientific status of a theory is its falsifiability, or refutability, or testability.”
Section Two: Falsification as the Method of Science
Having established falsifiability as the demarcation criterion, Popper now develops the logical structure of falsification in detail. He presents the argument using the logical form of modus tollens: if a universal statement “All swans are white” implies a prediction “This swan will be white,” and the observation shows a black swan, then the universal statement is logically refuted. This is deductive, not inductive reasoning. Popper emphasizes that “the logic of falsification is deductive logic” and that it requires no appeal to induction whatsoever.
Popper then addresses a crucial objection: no single observation can ever definitively falsify a theory because we can always save the theory by blaming faulty instruments, poor experimental conditions, or auxiliary assumptions. Popper accepts this point but argues that it does not undermine falsification as a methodology. He introduces the distinction between dogmatic falsificationism and methodological falsificationism. Dogmatic falsificationism demands absolute, decisive falsification based on indubitable empirical statements—this is impossible. Methodological falsificationism, however, recognizes that we must adopt methodological rules for deciding when a theory is falsified, accepting certain “basic statements” as sufficiently tested for the purposes of scientific criticism.
The content of this section also includes Popper’s treatment of corroboration. When a theory survives severe testing, it is not “verified” or “proved true”—it is merely “corroborated” for the time being. Corroboration is a report of past performance, not a guarantee of future success. A theory with high corroboration has been tested severely and found not to be false, but it remains permanently fallible. Popper writes that “corroboration (or degree of corroboration) is a comparative concept” that measures how well a theory has stood up to rigorous testing. The more severe the tests a theory has passed, the higher its degree of corroboration.
Section Three: The Problem of the Empirical Basis
Popper now turns to a foundational issue: what are the empirical statements that serve as the basis for testing theories? He calls this the problem of the empirical basis. Popper rejects the view that science rests on pure sense-data or uninterpreted perceptions. Instead, he argues that all empirical statements are interpretations of experience, laden with theory. There are no “given” facts; we always approach observation with expectations, assumptions, and theoretical frameworks. As Popper puts it, “observations are always interpretations of the facts observed.”
The resolution Popper offers is that the empirical basis consists of basic statements—singular existential statements that describe observable events. These statements are not indubitable or final; they are accepted by convention, temporarily, as the result of intersubjective testing. A basic statement might be “There is a black swan in this pond at time t.” Such a statement can be checked by multiple observers, and if they agree under specified conditions, it is accepted as part of the empirical basis. But Popper insists that this acceptance is a decision or a convention, not a logical deduction from pure experience. The empirical basis is therefore “at the bottom” of scientific knowledge, but it is not “at the bottom” in the sense of being absolutely certain.
This leads Popper to his famous metaphor of science as a structure built on piles driven into a swamp. The piles—the basic statements—are accepted provisionally, not because they rest on solid ground, but because they are “sufficiently firm” for the purposes of scientific testing. Popper clarifies that “the empirical basis of objective science has thus nothing ‘absolute’ about it.” Science does not rest on a foundation of indubitable sense-data; rather, it is a system of conjectures that we test against observational reports that are themselves fallible. The crucial point is that this fallibility does not lead to skepticism or relativism, because we can still engage in rational criticism and error elimination.
Section Four: Degrees of Testability and the Logic of Preference
Popper now develops a method for comparing theories based on their testability. He introduces the concept of empirical content: the greater the class of potential falsifiers a theory has, the more it says about the world, and the more testable it is. A theory that forbids more—that says “This cannot happen under these conditions”—is a better theory than one that forbids little. Popper writes, “the more a theory forbids, the more it says.” This is his criterion of progress in science: we prefer theories with higher empirical content because they are more informative and more open to refutation.
Popper then shows how this criterion applies to comparing theories. Given two theories, we prefer the one that is more falsifiable—the one that takes greater risks. This preference is not arbitrary; it follows from the aim of science, which is to get closer to the truth. A theory that is more testable is more exposed to criticism and has more opportunities to be shown false. If it survives those tests, it is more highly corroborated. Popper emphasizes the logical point that “every good scientific theory is a prohibition: it forbids certain things to happen.” The more a theory forbids, the better it is as a scientific hypothesis, because it makes more specific and thus more vulnerable predictions.
Popper also introduces the distinction between absolute and relative degrees of falsifiability. A theory is absolutely falsifiable if there exists some potential falsifier for it. But scientific progress involves relative comparisons: we say that theory T1 is more falsifiable than theory T2 if every potential falsifier of T2 is also a potential falsifier of T1, but not vice versa. This ordering allows Popper to argue that universality and precision both increase falsifiability. A universal law like “All planets move in ellipses” is more falsifiable than “Some planets move in ellipses” because the universal statement forbids more. Similarly, “All planets move in ellipses with eccentricity e” is more falsifiable than “All planets move in ellipses” because it is more precise.
Section Five: The Criterion of Progress — Falsifiability as a Normative Standard
Popper now consolidates his argument by showing that falsifiability is not merely a descriptive criterion for distinguishing science from non-science, but also a normative standard for evaluating scientific theories. He argues that a theory that is unfalsifiable is not just non-scientific—it is dogmatic and obscurantist. The psychological appeal of such theories lies in their ability to explain everything, but this is precisely their intellectual vice. Popper contrasts this with genuinely scientific theories, which are “risky” and “bold.” A bold theory is one that makes predictions that are improbable given what we already know, and thus runs the risk of being contradicted by observation.
Popper then introduces the concept of simplicity as a methodological preference. He argues that simpler theories are preferable because they are more falsifiable: a simpler theory has fewer adjustable parameters and thus forbids more than a complex theory that can be adjusted to fit any data. Popper explicitly challenges the traditional view that simplicity is a matter of aesthetic preference or convenience. Instead, he defines simplicity in terms of falsifiability: “the simplest theory is the one that has the greatest degree of falsifiability.” This is a purely logical criterion, not a metaphysical one. A theory with fewer free parameters is logically stronger because it says more about the world, and therefore it is easier to test and potentially refute.
Popper closes this section by connecting his philosophy of science to the broader aims of knowledge. He states that science does not aim at certainty or final truth, but at truth-likeness or verisimilitude—the approximation of truth. We never know that a theory is true, but we can know that it is closer to the truth than its predecessors if it has survived tests that falsified earlier theories. Popper writes that “we must not be afraid of taking risks; we must face the possibility of being mistaken.” The progress of science consists in the elimination of error through bold conjecture and severe testing, not in the accumulation of certainties. This sets the stage for his later work on objective knowledge and evolutionary epistemology.
Section Six: Probability and the Propensity Interpretation
We are now in Popper’s sustained treatment of probability, where he confronts a concept that had become central to modern physics and statistics, and he works to disentangle it from the verificationist and inductivist assumptions that had surrounded it.
The section opens with Popper’s recognition that probability statements pose a special problem for his falsificationist framework. Unlike universal laws, which forbid certain events, probability statements assert something about frequencies or propensities that cannot be strictly falsified by any finite number of observations. If a coin lands heads ten times in a row, this does not definitively refute the hypothesis that it is fair—unlikely sequences do occur. Popper writes that “probability statements are not falsifiable in the same sense as universal statements; they are not strictly falsifiable at all.” This seems to place them outside the domain of empirical science as Popper has defined it. He therefore must either exclude them or find a way to reconcile them with his criterion of demarcation.
Popper’s solution is to argue that probability statements are not treated as universal laws but as statistical hypotheses that can be tested by their predictions about relative frequencies in long sequences. A probability statement like “the probability of heads is 1/2” is interpreted as a prediction that, in an indefinitely long sequence of tosses, the relative frequency of heads will approach 1/2. While no finite run of outcomes can decisively falsify this prediction, Popper argues that we can adopt conventions about when we regard a sequence as falsifying the hypothesis—for example, when the observed frequency deviates too far from the predicted value given a specified level of statistical significance. He emphasizes that this involves a methodological decision, not a logical deduction, and that such decisions are always revisable in light of further evidence.
But Popper goes further than this instrumentalist approach to probability. He develops his own propensity interpretation of probability, which treats probability as an objective property of the experimental setup or generating conditions, not merely as a subjective degree of belief or a frequency in an actual finite series. For Popper, the probability of heads is a physical propensity of the coin-tossing arrangement to produce heads with a certain frequency in the long run. This propensity is real, though it is measured by the hypothetical infinite sequence of outcomes. He argues that this interpretation avoids the problems of the frequency theory, which cannot meaningfully assign probabilities to single events, and the subjectivist theory, which reduces probability to ignorance. The propensity interpretation allows probability to function as a genuine scientific hypothesis about the world, testable through statistical methods.
This matters within the work because it shows Popper extending his falsificationist logic to the probabilistic sciences—quantum mechanics, genetics, statistical mechanics—without abandoning his core commitment to testability. By treating probability statements as testable by statistical conventions and by grounding them in objective propensities, Popper preserves the demarcation between science and non-science even for domains where strict falsification is impossible. He also positions himself against the dominant subjectivist and frequency-based interpretations of his time, offering a realist alternative that aligns with his broader metaphysical commitments to an objective, mind-independent reality.
Section Seven: Corroboration and the Logic of Preference
Moving from the formal logic of falsification to the practical problem of theory choice in science, Popper introduces the concept of corroboration as the central epistemic notion for evaluating theories. This section addresses a question that naturally arises from his rejection of induction: if we can never confirm a theory or assign it a probability, on what basis should we prefer one theory over another?
Popper begins by distinguishing between truth and corroboration. Truth is an objective, absolute property of a statement—a theory is either true or false, whether we know it or not. Corroboration, by contrast, is a temporal, relative measure of how well a theory has stood up to severe testing so far. A theory is highly corroborated when it has survived attempts to falsify it, especially when those attempts involved tests that would likely have refuted it had it been false. Popper writes that “corroboration is not a measure of probability; it is a measure of the degree to which a theory has been tested and has survived tests.” The degree of corroboration of a theory is therefore a function of its testability (how much it forbids) and the severity of the tests it has passed.
The logic of preference that follows from this is straightforward: we prefer theories that are more falsifiable (i.e., that have greater empirical content) and that have withstood more severe tests. Popper argues that this is a rational preference because more falsifiable theories are more informative—they tell us more about the world by forbidding more possible states of affairs. A theory that forbids nothing explains nothing. If a highly falsifiable theory survives attempts to falsify it, it has earned its corroboration in a way that a less falsifiable theory cannot. Popper emphasizes that this preference is not based on any probabilistic calculation of the theory’s likely truth, but on the rationality of conjecture and criticism: we hold onto theories that have proven their mettle, while remaining aware that they may be overturned tomorrow.
Popper also clarifies that corroboration is not a substitute for truth. A theory can be highly corroborated and still false. The history of science is full of examples—Newtonian mechanics was highly corroborated for centuries before Einstein showed its limits. Corroboration merely indicates that a theory has withstood our best efforts to find its flaws so far. It is a guide to action (we rely on highly corroborated theories in practical contexts) and to further research (we attempt to extend and apply them), but it does not license inductive inference about their future success. Popper insists that “the only logical basis for preferring one theory to another is that it has been better tested and has survived those tests”—not that it is more probable to be true.
This matters because it provides the positive dimension of Popper’s scientific method. Falsification alone might seem purely negative—we only refute, never justify. Corroboration gives scientists a rational basis for choosing among competing theories without falling back into inductivism. It also connects Popper’s logic of scientific discovery to the actual practice of science, where theories are constantly compared, tested, and provisionally accepted or rejected based on their track record. Corroboration is the bridge between Popper’s austere deductive logic and the messy, progressive reality of scientific inquiry.
Section Eight: Truth, Verisimilitude, and the Aim of Science
Having established the falsificationist logic of testing and the concept of corroboration, Popper now turns to the ultimate aim of science: truth. This section marks a decisive move from the logical and methodological concerns of earlier chapters to the metaphysical and epistemological foundations of his entire project.
Popper begins by affirming a correspondence theory of truth, following Tarski’s semantic conception: a statement is true if it corresponds to the facts, if the state of affairs it describes actually obtains. He rejects coherence theories (truth as internal consistency) and pragmatic theories (truth as what works) as inadequate, arguing that they collapse the distinction between what we believe and what is the case. “Truth is objective,” Popper writes, “it is not relative to our knowledge or our interests.” This commitment to objective truth is essential to his entire philosophy of science because it makes falsification meaningful: a theory is false if it does not correspond to the facts, and we can sometimes—though not always—know that it is false through empirical testing.
But Popper immediately recognizes a problem: if we can never know that a theory is true (because induction is invalid), and if we can only know that some theories are false, what is the point of science? His answer is that science aims not at certainty or final truth, but at truth-likeness or verisimilitude—the approximation of truth. Even if we never reach the truth, we can get closer to it. A theory with greater verisimilitude is one that contains more true consequences and fewer false consequences than its predecessor, even if it is still not fully true. Popper gives the example of Kepler’s laws versus Newton’s theory: Newton’s theory is closer to the truth than Kepler’s because it corrects Kepler’s errors while preserving his successes, even though Newton’s theory itself was later superseded by Einstein.
Popper then attempts to define verisimilitude formally: the truth-content of a theory is the set of all true consequences derivable from it; the falsity-content is the set of all false consequences. A theory A has greater verisimilitude than theory B if A’s truth-content is greater than B’s and A’s falsity-content is less than B’s (or one is greater while the other is not smaller). This definition, however, encountered well-known technical problems that Popper later acknowledged and that generated significant debate in the philosophy of science. Despite these difficulties, the intuitive idea remained central to his thought: scientific progress means moving toward theories that are more informative and more accurate, even if absolute truth remains a regulative ideal.
Popper also links verisimilitude to his earlier discussion of testability and corroboration. A theory that is more falsifiable (has more empirical content) can, if it survives testing, achieve higher corroboration. But corroboration is not the same as verisimilitude—it is merely a measure of how well a theory has performed so far. Popper insists that “we must not be afraid of taking risks; we must face the possibility of being mistaken.” The progress of science consists in the elimination of error through bold conjecture and severe testing, not in the accumulation of certainties. This sets the stage for his later work on objective knowledge and evolutionary epistemology, where the growth of knowledge is seen as a process of variation and selective retention analogous to biological evolution.
This section matters because it resolves a tension that had run through the entire work: how can a fallibilist and anti-inductivist account of science make sense of scientific progress? By introducing verisimilitude, Popper provides a goal for science—truth—that is both objective and unattainable in its final form, yet approachable through the method of conjecture and refutation. The concept also allows him to explain why later theories are better than earlier ones even when both are false, a phenomenon that is evident throughout the history of science but difficult to account for within verificationist or instrumentalist frameworks. Truth-likeness is Popper’s answer to the question of what science is ultimately about.
Section Nine: The Growth of Knowledge and the Role of Tradition
We are now in the later portions of Popper’s collected works, where he turns from the formal logic of science to the historical and sociological conditions that make scientific progress possible. This section examines how knowledge actually grows within a tradition and why tradition is not opposed to rational criticism but is its necessary precondition.
Popper begins by rejecting the widespread Enlightenment view that tradition is simply an obstacle to knowledge. He argues that all rational inquiry must start from some inherited body of beliefs, assumptions, and problems. “We cannot start from nothing,” Popper writes, “we must start from a tradition — a set of accepted theories, problems, and methods that constitute the starting point for criticism and revision.” The idea that one could begin entirely afresh, free from all preconceptions, is a myth that Popper attributes to the Cartesian tradition of radical doubt and the Baconian ideal of the mind as a blank slate.
The key move Popper makes is to distinguish between the uncritical acceptance of tradition and the critical use of tradition. A tradition provides the background knowledge against which a problem can be formulated and a conjecture proposed. Without such a background, there would be no way to identify what is problematic, no way to frame an interesting hypothesis, and no way to conduct meaningful tests. The scientist who rejects all tradition is not liberated but paralyzed, because he has no starting point for criticism.
This leads Popper to a deeper point about the nature of rational criticism itself. Criticism, he insists, is always immanent criticism — it proceeds by examining a theory in terms of its own internal consistency, its ability to solve the problems it was designed to solve, and its capacity to withstand testing against the background of accepted knowledge. There is no such thing as a purely external or absolute standpoint from which to judge theories. The growth of knowledge is therefore not a process of replacing tradition wholesale but of modifying it piecemeal through successive conjectures and refutations.
Popper also uses this discussion to clarify his relationship to historicism and the sociology of knowledge. He rejects the view associated with Hegel and Marx that all knowledge is determined by its historical and social context, because this leads to relativism and the denial of objective truth. But he equally rejects the view that knowledge exists in a timeless Platonic realm, untouched by history and practice. Instead, he proposes that knowledge grows through a process that is both historically situated and rationally accountable. The situational logic of the scientist — the specific problem-situation she inherits, the available theories and evidence, the institutional structures of scientific practice — shapes the direction of inquiry without determining its outcomes.
Clarifying this point is essential for understanding Popper’s overall project. He is not offering a purely logical or formal account of scientific method, divorced from real human practice. Nor is he offering a sociological reduction that would deny the objectivity of scientific results. Rather, he is describing the conditions under which objective knowledge can emerge from historically specific processes of trial and error. The tradition of criticism—what he elsewhere calls the “critical tradition” that began with the pre-Socratics—is itself a historical achievement, one that must be maintained and protected if science is to continue to grow.
Section Ten: The Evolution of Knowledge and the Darwinian Analogy
This section extends the previous discussion by drawing a systematic analogy between biological evolution and the growth of scientific knowledge. Popper argues that both processes follow the same fundamental pattern: variation and selective elimination. In biology, genetic mutations provide the variations and natural selection provides the elimination; in science, conjectures provide the variations and refutation provides the elimination.
Popper is explicit about the limits of this analogy. He does not claim that scientific theories are literally organisms, or that the process of selection in science operates in exactly the same way as natural selection. The analogy is structural and heuristic. What matters is the rejection of instructionist or Lamarckian models of knowledge acquisition. Just as organisms do not acquire adaptations by directly recording environmental pressures, so scientists do not acquire knowledge by passively receiving sense data or by induction from observations. In both cases, the mechanism is one of blind variation followed by selective retention.
“The growth of our knowledge is the result of a process closely resembling what Darwin called ‘natural selection’; that is, the natural selection of hypotheses: our knowledge consists, at every moment, of those hypotheses which have shown their (comparative) fitness by surviving so far in their struggle for existence; a competitive struggle which eliminates those hypotheses which are unfit.”
This passage makes clear that Popper sees the analogy as more than a metaphor. It is a way of understanding the fundamental unity of all adaptive processes, whether biological, behavioral, or cognitive. The same logic that explains the evolution of the eye also explains the evolution of the theory of optics. Both are the product of trial and error, of the elimination of unsuccessful variants by the environment—though in the case of science, the “environment” includes the logical and empirical constraints imposed by the problem-situation.
Popper draws several important consequences from this analogy. First, it implies that there is no fundamental difference between the growth of knowledge in animals and in humans. Animal knowledge, such as the knowledge that enables a bird to navigate or a beaver to build a dam, is also produced by conjecture and refutation, though the conjectures are embodied in behavior and the refutations in the consequences of that behavior. Human scientific knowledge is distinguished only by its explicit, linguistic formulation, which makes it subject to conscious criticism and testing.
Second, the analogy undermines any attempt to ground knowledge in certain foundations. Just as no organism is perfectly adapted to its environment, no theory is perfectly true. The evolutionary process does not produce final, unchangeable forms; it produces forms that are temporarily viable in a changing environment. Similarly, science does not produce final, indubitable truths; it produces theories that have survived testing so far and that remain open to future revision.
Third, the analogy highlights the importance of pluralism in science. In biology, genetic diversity is essential for adaptation to changing conditions; in science, theoretical diversity is essential for the detection of error and the discovery of new problems. A scientific community that suppresses alternative hypotheses risks stagnation, just as a population that loses genetic variation risks extinction. Popper’s defense of methodological pluralism is thus not merely a political or ethical position; it follows from the logic of evolutionary epistemology itself.
Clarifying this section is important because it shows that Popper’s epistemology is not narrowly logical or methodological but is embedded in a broader naturalistic worldview. The growth of knowledge is not a special exception to the laws of nature but an extension of them. This does not reduce knowledge to biology—the logic of scientific testing introduces norms and standards that have no analogue in natural selection—but it does show that the same basic pattern of variation and selection operates at every level of adaptive complexity.
Section Eleven: The Problem of the Empirical Basis Revisited
Popper returns here to a problem that was central to the earlier sections of the work but that has taken on new significance in light of the evolutionary account of knowledge. The empirical basis of science—the set of basic statements that serve as potential falsifiers—cannot be grounded in sense experience in the way that empiricists have traditionally supposed. But if the empirical basis is not given by perception, how is it determined?
Popper’s answer is that the empirical basis is itself the product of a decision or convention. The basic statements that are accepted as potential falsifiers are not forced upon us by the data of experience; they are chosen by the scientific community in light of practical and methodological considerations. This choice is not arbitrary—it is constrained by the need for intersubjective testability and by the background knowledge of the time—but it is not dictated by experience alone.
The process by which basic statements are accepted is described by Popper as a kind of “trial” in which the scientist proposes a statement and the community tests it against other accepted statements and against the results of observation. If the statement is accepted, it becomes part of the “empirical basis” for the time being, but it remains open to revision if new evidence or new theoretical developments call it into question. This means that the empirical basis is not a fixed foundation for knowledge but a shifting and provisional set of agreements.
“The empirical basis of objective science has thus nothing ‘absolute’ about it. Science does not rest upon rock-bottom. The bold structure of its theories rises, as it were, above a swamp. It is like a building erected on piles. The piles are driven down from above into the swamp, but not down to any natural or ‘given’ base; and when we cease our attempts to drive our piles into a deeper layer, it is not because we have reached firm ground. We simply stop when we are satisfied that they are firm enough to carry the structure, at least for the time being.”
This passage is one of Popper’s most vivid images of the scientific enterprise. It makes clear that the empirical basis is not a foundation in the traditional sense—a set of indubitable, given facts that guarantee the truth of theories. Instead, it is a provisional stopping point, a set of statements that the scientific community agrees to accept for the purposes of testing, but that can always be called into question if new problems arise.
Popper uses this discussion to address a common objection to his falsificationism: that if basic statements are themselves fallible and conventional, then the whole edifice of science is built on sand. If no statement is immune to revision, it might seem that there is no way to ever falsify a theory definitively, since one could always reject the basic statement instead of the theory. This is the problem of “the immunity of theories” —the worry that falsification is never forced upon us by the evidence but is always the result of a choice.
Popper’s response is to accept this consequence fully. He acknowledges that falsification is never final or conclusive. A theory can always be saved by rejecting the basic statement that purports to falsify it, and sometimes this is the scientifically rational thing to do. The crucial point is that such decisions are themselves subject to criticism and testing. If the rejection of a basic statement leads to contradictions with other accepted statements, or if it blocks the growth of knowledge by making the theory unfalsifiable, then the decision can be criticized and revised. There is no ultimate guarantee, no Archimedean point from which to settle all disputes—only the ongoing process of conjecture, testing, and criticism.
Clarifying this dialectical relationship between theory and empirical basis is essential for understanding Popper’s fallibilism. The provisionality of the empirical basis is not a weakness of science but a strength, because it allows science to be self-correcting. Nothing is fixed forever, and nothing is immune to criticism—not even the empirical “facts” that are supposed to test our theories. This is what Popper means by “critical rationalism” : the application of the principle of criticism all the way down, to the very foundations of knowledge itself.
Section Twelve: The Problem of the Objectivity of Science
We are now in the later stages of Popper’s epistemology, where he addresses the social and intersubjective dimension of scientific knowledge. This section examines how objectivity is possible despite the fallibility of all individual observation statements.
Popper opens by confronting a fundamental challenge: if every empirical basis statement is itself fallible and theory-laden, how can science claim objectivity? The answer does not lie in the individual scientist’s mind or methods, but in the social character of scientific criticism. Objectivity, Popper argues, is not a property of individual scientific judgments but of the critical process that subjects all claims to public scrutiny. He writes:
“The objectivity of science is not a matter of the individual scientist’s impartiality or objectivity, but a matter of the social or public character of scientific method; the individual scientist’s impartiality is, as far as it exists, not the source but rather the result of this socially or institutionally organized objectivity of science.”
The key move here is to relocate objectivity from the individual to the community. No single scientist can be trusted to be objective, because every individual is subject to bias, prejudice, and personal interest. But the institution of science—with its journals, conferences, peer review, and public debate—creates a mechanism whereby individual biases can be exposed and corrected. It is not that scientists are more virtuous than other people; it is that the social structure of science forces them to be critical of their own and others’ work.
Popper draws a comparison with legal procedure. A judge is not expected to be infallible or perfectly impartial in some abstract sense; rather, the legal system provides procedures—cross-examination, appeals, public hearings—that allow errors and biases to be identified and corrected. Similarly, the method of science is a set of institutionalized procedures for the critical testing of claims. The objectivity of science rests on the public character of the testing process: theories are published, experiments are described in sufficient detail to be replicated, and results are open to challenge by anyone.
This social account of objectivity has profound implications for Popper’s conception of scientific rationality. Rationality is not a private faculty of individual minds but a public method of criticism. A rational belief is not one that is held for good reasons in the privacy of one’s own consciousness; rather, it is one that has survived sustained public criticism. The rationality of science is therefore inseparable from its social organization, from the institutions that ensure “the free competition of ideas” and the “public character of scientific debate.”
Clarifying this point is essential because it explains why Popper rejects both the idea that science is purely subjective and the idea that it rests on some indubitable foundation. Science is objective in the only sense available to fallibilists: it is open to correction by anyone, and its claims are tested against intersubjectively available evidence. This is not a guarantee of truth but a guarantee of accountability—and for Popper, that is all the objectivity we need or can have.
Section Thirteen: The Growth of Knowledge Through Conjectures and Refutations
This section represents Popper’s mature synthesis of his philosophy of science into a model of how knowledge grows at all levels, from the simplest perceptual learning to the most advanced theoretical physics. The title phrase—”conjectures and refutations”—captures the entire movement of Popper’s thought.
Popper presents his model as a universal schema for the growth of knowledge: P1 → TT → EE → P2. Here, P1 is the initial problem, TT is a tentative theory (or conjecture) proposed as a solution, EE is the process of error-elimination (testing, criticism, refutation), and P2 is the new problem that emerges after the testing process. The schema is explicitly designed to parallel the Darwinian model of evolution by natural selection: variation (conjectures) followed by selection (error-elimination).
The crucial insight is that knowledge does not grow by accumulating facts but by solving problems. Every advance in science begins with a problem—an inconsistency between theory and observation, a conflict between two theories, or a practical difficulty. The problem is then addressed by a bold conjecture, a new theory that goes beyond the available evidence and proposes a solution. This conjecture is then subjected to the most severe testing possible; if it survives, it is provisionally accepted, but it immediately generates new problems. Popper writes:
“The growth of knowledge proceeds from old problems to new problems, by means of conjectures and refutations.”
Popper emphasizes that the schema applies not only to scientific theories but also to pre-scientific and even animal knowledge. A mouse that learns to avoid a cat is engaged in the same basic pattern: it has a problem (survival), it tries a tentative theory (the cat is dangerous), and it eliminates errors (by not approaching the cat). The difference between scientific and pre-scientific knowledge is that science makes the process explicit, critical, and self-conscious. Science is “criticism in the full light of day”—it deliberately invents bold conjectures and then deliberately tries to refute them.
This model directly challenges the empiricist view that knowledge begins with observations. For Popper, observations are always theory-laden; they are already interpretations in light of prior expectations. The starting point of inquiry is not the accumulation of neutral data but the identification of a problem. Problems arise when our expectations are disappointed, when our theories fail. The growth of knowledge is therefore a process of learning from our mistakes—of trying again, but better.
Clarifying the significance of this schema reveals its radical implications. If knowledge grows through conjectures and refutations, then there is no such thing as a definitive or final theory. Every solution to a problem generates new problems—indeed, deeper and more interesting problems. The aim of science is not to reach a final truth but to make progress by solving ever more profound problems. This is why Popper can say that “the advance of science is not towards a final state but towards ever new and deeper problems.”
Section Fourteen: The Autonomy of the Third World
In this section, Popper develops one of his most distinctive metaphysical ideas: the theory of three worlds or three realms of reality. This theory emerges from his reflection on the status of scientific knowledge as something that transcends individual knowers.
Popper distinguishes World 1 (the world of physical objects and states), World 2 (the world of mental states, consciousness, and subjective experiences), and World 3 (the world of objective contents of thought, including theories, problems, arguments, and works of art). The crucial claim is that World 3 is autonomous: although it is a product of human minds, it has its own objective properties and causal powers that are independent of any individual’s mental states.
The argument proceeds through an example. Consider the sequence of natural numbers. Humans invented the natural numbers; they are products of human thought. But once invented, the sequence has objective properties that no one created and that we discover rather than invent. The fact that there are infinitely many prime numbers, for example, was not created by anyone—it is an objective property of the natural number system that was discovered by Euclid. Similarly, a mathematical proof has objective logical properties—validity, soundness, implications—that hold regardless of what any particular mathematician thinks. Popper states:
“Thus the third world is autonomous: it is, in a certain sense, a world of abstract entities; it is a world of problems, theories, and arguments; and it is a world that is produced by human beings, yet it is largely autonomous and has its own laws of growth.”
The autonomy of World 3 is most evident in the phenomenon of unintended consequences. When a scientist proposes a theory, she cannot foresee all its logical consequences. These consequences are objectively there in the theory, waiting to be discovered, but they do not depend on anyone’s awareness of them. A theory may have implications that contradict other accepted theories, or it may open up new problems that no one anticipated. These features belong to the theory itself as an objective entity, not to the mental state of its creator.
Popper uses the three-worlds framework to explain how objective knowledge is possible. Scientific knowledge is not reducible to the subjective beliefs of scientists (World 2) or to the physical documents that record it (World 1). A book is a physical object (World 1), and reading it produces mental states (World 2), but the content of the book—the theory it expresses—belongs to World 3. This content can be understood, criticized, and improved by many different people across generations. It is this objectivity of content that makes cumulative scientific progress possible.
Clarifying this three-worlds ontology is important for understanding Popper’s anti-subjectivism. He is not a Platonist who believes that World 3 entities exist eternally independent of human activity. They are human products, but they transcend their creators. They have what Popper calls a “growing autonomy” : once created, they generate their own problems and their own standards of evaluation. The theory of natural selection, for example, once formulated, has objective implications that scientists must respect regardless of their personal preferences. In this sense, World 3 is both our creation and our master—we make it, but then it makes us.
Section Fifteen: The Mind-Body Problem and the Interaction of Worlds
This section extends the three-worlds framework to address the traditional mind-body problem. Popper uses his ontology to argue for interactionism—the view that mental states and physical states causally interact—while avoiding the difficulties of both materialism and idealism.
Popper begins by criticizing both crude materialism (which reduces mental states to physical states) and idealism (which reduces physical states to mental states). The problem with materialism is that it cannot account for the causal efficacy of consciousness and rational thought. If all mental states are merely epiphenomena of physical processes, then our beliefs and arguments are causally irrelevant—but this contradicts the very practice of science, which assumes that rational argument can change minds. The problem with idealism is that it cannot account for the reality of the physical world that science investigates.
The three-worlds framework offers a way out. Popper proposes that all three worlds are real and that they interact causally. World 1 (physical) and World 2 (mental) interact: physical events cause mental events (e.g., a blow to the head causes pain), and mental events cause physical events (e.g., a decision to move one’s arm causes the arm to move). World 2 and World 3 also interact: mental acts of thinking engage with World 3 objects (theories, arguments), and World 3 objects cause mental states (understanding a proof produces intellectual satisfaction). And through World 2, World 1 and World 3 interact: physical books (World 1) encode World 3 content that, through mental acts (World 2), can be understood and criticized. Popper writes:
“The three worlds are so related that the first two can interact, and the last two can interact. This means that the second, the world of subjective experiences, can interact with each of the other two worlds. The first world and the third world cannot interact, save through the mediation of the second.”
The crucial point for the philosophy of science is that the interaction between World 2 and World 3 explains how rational criticism is possible. When we criticize a theory, we are not merely expressing a subjective feeling of dissatisfaction; we are engaging with the objective logical content of the theory. The theory has objective properties (consistency, explanatory power, testability) that can be evaluated by anyone who understands it. Our subjective mental acts of understanding and evaluation are what mediate between the physical world (where experiments are performed) and the world of ideas (where theories live).
Popper uses this framework to argue for the irreducibility of human freedom and rationality. If World 3 did not exist, or if we had no access to it, then our mental lives would be locked inside our skulls, determined entirely by physical causes. But because we can engage with objective ideas—because we can consider arguments, weigh evidence, and choose among theories—we are not simply automatons. The existence of World 3 provides the condition for objective truth and rational choice, and thus for the kind of knowledge that makes science possible.
Clarifying this interactionist position shows how Popper connects his epistemology to his metaphysics. The objectivity of scientific knowledge (World 3) depends on the ability of individual scientists (World 2) to engage with abstract ideas and to act on the basis of rational considerations. And the physical world (World 1) provides the testing ground where our theories succeed or fail. The three worlds are distinct but inseparable, and science is the enterprise that brings them into productive interaction.
Section Sixteen: The Open Society and the Critique of Historicism
This section marks Popper’s turn from the philosophy of science to political and social philosophy. He applies his critical rationalist method to theories of history and society, particularly to what he calls historicism—the doctrine that history follows inexorable laws of development that can be predicted by theory.
Popper’s target is the view, found in Plato, Hegel, Marx, and many others, that there is a “rhythm” or “pattern” or “law” governing the course of human history. Historicists claim that by discovering these laws, we can predict the future and guide political action accordingly. Popper argues that this project is not merely mistaken but dangerous, because it leads to authoritarianism and the suppression of freedom. He writes:
“Historicism is a very old idea; it is the idea that history is controlled by specific historical or evolutionary laws whose discovery would enable us to prophesy the destiny of man.”
The critique rests on several arguments. First, the course of human history is influenced by the growth of human knowledge, and we cannot predict the future growth of knowledge because to do so would require us to know now what we will only discover later. This is the “non-predictability of future knowledge” argument. Second, the attempt to discover historical laws confuses trends with laws. A trend (e.g., the growth of population) depends on specific conditions that may change; a law holds universally. Historicists mistake contingent trends for necessary laws and thus treat the future as predetermined.
Popper offers an alternative: piecemeal social engineering. Instead of attempting to redesign society according to a comprehensive blueprint (which requires predicting the future and controlling all variables), we should approach social problems as we approach scientific problems: by proposing tentative solutions, testing them, learning from our mistakes, and making incremental corrections. This approach respects human freedom because it allows for experimentation and criticism; it does not impose a single vision of the good society but allows different proposals to compete and be evaluated.
The connection to Popper’s philosophy of science is explicit. Just as science proceeds by conjectures and refutations, so should social policy. The open society is one in which policies are open to criticism, in which fallible leaders can be removed without violence, and in which institutional mechanisms exist for the peaceful correction of errors. This is the political analogue of critical rationalism: the application of reason to social life not as a source of final truths but as a method of trial and error-elimination.
Clarifying this political application reveals the unity of Popper’s thought. The same fallibilism that drives his philosophy of science—the recognition that all knowledge is conjectural and that progress comes only through criticism—drives his political philosophy. The enemy is not error but dogmatism: the refusal to admit that one might be wrong. Historicism is dangerous because it claims to have discovered the inescapable laws of history, thereby placing itself beyond criticism. The open society, by contrast, institutionalizes criticism and thus creates the conditions for both scientific and political progress.
