{
  "id": "a-priori-and-a-posteriori",
  "title": "A Priori and A Posteriori",
  "lang": "en",
  "source_lang": "en",
  "summary": "A priori and a posteriori distinguish knowledge or justification independent of experience from knowledge or justification that depends on experience.",
  "category": "philosophy",
  "subcategory": "metaphysics-epistemology",
  "type": "concept",
  "aliases": [
    "A priori knowledge and a posteriori knowledge",
    "Apriority and aposteriority"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/a-priori-and-a-posteriori/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/a-priori-and-a-posteriori/",
      "json": "https://aiwiki.page/zh/a-priori-and-a-posteriori.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The distinction between a priori and a posteriori concerns the grounds on which a belief is justified or something is known. In its standard modern sense, a priori justification does not depend on experiential evidence, whereas a posteriori justification does. Elementary arithmetic and logical inference are traditional examples of the former; knowledge of a particular city’s population or a substance’s chemical composition exemplifies the latter. The distinction is central to epistemology because it concerns the sources, scope, and limits of human knowledge.\n\nThe expressions are Latin: a priori means “from what is prior,” and a posteriori means “from what is posterior.” Their historical meanings concerned different directions of explanation or demonstration, rather than exactly the modern contrast between experiential and nonexperiential justification. Modern discussion also distinguishes apriority from two other classifications: whether a statement is true by virtue of meaning, and whether it could have been false. Much of the subject’s history concerns the relations among these three distinctions, especially in mathematics, metaphysics, and scientific reasoning."
    },
    {
      "heading": "The distinction and its proper objects",
      "text": "The a priori–a posteriori distinction is most precisely applied to ways of knowing or to sources of epistemic justification. A belief is justified a posteriori when its evidential basis includes perception, observation, or some other relevant form of experience. Knowledge that a particular tree has lost its leaves normally depends on seeing the tree, receiving a reliable report, or consulting a record. A belief is justified a priori when reflection or reasoning supplies its warrant without observational evidence for the proposition concerned. Recognizing that no object can be both entirely red and entirely green at the same time, in the same respect, is a frequently discussed candidate.\n\nCalling a proposition “a priori” is often shorthand for saying that it can be known a priori, rather than that everyone who knows it does so in that way. Someone may learn an arithmetical result from a calculator, a textbook, or a teacher without understanding its proof. Another person may establish the same result through reasoning. The proposition is the same, but the immediate epistemic routes differ. Consequently, the classification of propositions must be distinguished from the classification of particular episodes of believing or knowing them.\n\nIndependence from experience does not ordinarily mean independence from every experiential prerequisite. Learning a language, acquiring numerical concepts, or understanding spatial vocabulary may require perception and instruction. Defenders of a priori knowledge generally distinguish this enabling role of experience from its evidential role. Seeing collections of objects may help a child acquire the concept of addition without making every later arithmetic proof an empirical generalization from those collections. Critics question whether this division can always be maintained, particularly where concepts are shaped by scientific discovery.\n\nThe relevant sense of “experience” is itself disputed. Sense perception is the clearest case, but philosophers differ over whether introspection, memory, and conscious reflection should count as experiential sources. A person’s knowledge of a present headache is usually classified as a posteriori even though it does not depend on external observation. Conversely, simply calling intellectual reflection a kind of “experience” does not establish that it is empirical in the intended epistemological sense. A useful account must explain what supplies the evidence, not merely classify mental events by their phenomenological character.\n\nNeither category is identical with certainty. Observational knowledge can be exceptionally secure, while a proposed a priori proof can contain unnoticed errors. Fallibilism allows that justified beliefs, including beliefs reached through reasoning, may be mistaken. Apriority concerns the character of a warrant; it does not automatically imply infallibility, obviousness, universal agreement, or immunity from later criticism."
    },
    {
      "heading": "Ancient, medieval, and early modern antecedents",
      "text": "Ancient philosophy developed several contrasts that anticipated the later distinction without reproducing it exactly. In the Meno, Plato used a geometrical discussion to investigate learning and recollection; in the Republic, mathematical reasoning helped illustrate cognition directed beyond changing sensible particulars. These discussions concern the possibility of intellectual access to stable objects and relations, not a modern technical definition of a priori justification. Aristotle, especially in the Posterior Analytics, distinguished knowing that something is so from knowing why it is so. Scientific understanding required grasping explanatory causes and demonstrating conclusions from appropriate principles.\n\nThe medieval vocabulary of priority was closely connected with this explanatory framework. Demonstration a priori proceeded from a cause or explanatory ground to its effect, while demonstration a posteriori proceeded from an effect toward its cause. An effect might be more immediately accessible to human observation even though its cause was prior in the order of explanation. Thomas Aquinas, for example, distinguished demonstration through a cause from demonstration through effects. These distinctions should not be equated with modern apriority: an argument from a known physical cause to its effect may still depend heavily on empirical premises.\n\nSeventeenth-century rationalism gave renewed importance to knowledge obtainable through intellectual reasoning. René Descartes sought foundations resistant to skeptical doubt and treated clear and distinct cognition as central to certainty. Gottfried Wilhelm Leibniz distinguished truths of reasoning from truths of fact. In the Monadology, written in 1714, truths of reasoning are necessary and their opposites impossible, whereas truths of fact are contingent. His appeal to the principle of noncontradiction and the principle of sufficient reason linked logical, metaphysical, and epistemological questions that later thinkers would separate more sharply.\n\nEarly modern empiricism complicated any simple opposition between reason and experience. John Locke, in An Essay Concerning Human Understanding (1690), rejected innate principles while allowing intuitive and demonstrative knowledge. His rejection of innateness did not amount to the claim that every justified judgment must be supported by observation of its subject matter. David Hume, in An Enquiry concerning Human Understanding (1748), distinguished “relations of ideas,” exemplified by arithmetic and geometry, from “matters of fact,” whose contraries remain conceivable without contradiction.\n\nHume’s account of causation gave the distinction particular force. Observing a cause does not, through reason alone, disclose its effect: knowledge that fire burns or that bread nourishes depends on experience. His analysis of the problem of induction further questioned whether past regularities rationally establish future ones. These arguments helped frame the problem inherited by Kant: how can judgments possess necessity or universality if experience discloses only particular occurrences?"
    },
    {
      "heading": "Kant and the synthetic a priori",
      "text": "Immanuel Kant made the distinction systematic in the *Critique of Pure Reason*, first published in 1781 and substantially revised in 1787. He argued that although human cognition begins with experience, it does not follow that all cognition arises from experience. The mind may contribute forms and principles that make experience possible. On his account, necessity and strict universality are marks of a priori cognition: observation can show that something has happened repeatedly, but cannot by itself establish that it must happen or that no exception is possible.\n\nKant distinguished a priori cognition from pure a priori cognition. A judgment might have an a priori basis while containing a concept acquired through experience. His example that every alteration has a cause involves the concept of alteration, which depends on experience, even though the causal principle is not supposed to be established by enumerating observed cases. Pure cognition excludes such empirical admixture. This qualification matters because Kant’s use of “a priori” is not simply interchangeable with “devoid of all empirical concepts.”\n\nKant combined the epistemological distinction with the analytic–synthetic distinction. An analytic judgment explicates what is already contained in its subject concept; a synthetic judgment adds something not contained in that way. “All bodies are extended” was his standard analytic example, while “all bodies are heavy” was synthetic. His central innovation was the category of the synthetic a priori: judgments that extend knowledge but are not justified by experience. He classified basic arithmetic and geometry in this way, arguing that mathematical construction requires intuition rather than conceptual analysis alone.\n\nHis account of \\(7+5=12\\) illustrates the proposal. The concept of the sum of seven and five does not, he maintained, yield twelve merely through analysis; the addition must be carried out in intuition. Geometry similarly depends on the pure intuition of space. These claims differ both from an empirical account of mathematics and from the later logicist ambition to derive arithmetic from logic. Kant also treated fundamental principles governing possible experience, including causal principles, as synthetic a priori.\n\nWithin transcendental idealism, space and time are forms of human sensibility, while categories such as substance and causality organize objects of experience. The resulting principles apply to appearances, not to things considered independently of the conditions of human cognition. This limitation was intended to explain the success of mathematics and natural science while restricting speculative metaphysics. Later developments in geometry and physics challenged important aspects of Kant’s account, but his central question remained influential: whether objective knowledge presupposes structures that cannot themselves be justified by ordinary observation."
    },
    {
      "heading": "Apriority, analyticity, and necessity",
      "text": "Three distinctions must be kept separate. The a priori–a posteriori distinction concerns how something is justified or known. The analytic–synthetic distinction concerns, roughly, whether a statement’s truth depends on meaning or instead makes a substantive claim not settled by meaning alone. The necessary–contingent distinction concerns whether something could have been otherwise. These are respectively epistemological, semantic or conceptual, and modal classifications. Their frequent alignment in familiar examples does not establish that they coincide.\n\nA standard analytic example is “Every bachelor is unmarried,” under the usual interpretation of “bachelor.” Someone who understands the relevant expressions appears able to recognize its truth without surveying bachelors. This makes it a plausible case of a priori knowledge. But theories of analyticity differ: Kant’s subject–predicate containment account does not straightforwardly cover every logical truth or mathematical statement. Later accounts appeal to definitions, synonymy, logical form, or linguistic rules. Whether these notions adequately explain a priori justification is a further question rather than part of a universally accepted definition.\n\nNecessity introduces another issue. In contemporary treatments of modality, a necessary proposition is often characterized as true in every relevant possible world, whereas a contingent proposition is true in some possibilities and false in others. This framework does not itself explain how anyone knows which propositions are necessary. A necessary truth may be difficult to establish, discovered only after long inquiry, or beyond the cognitive resources of a particular person. Metaphysical status is not a measure of psychological accessibility.\n\nLikewise, deductive reasoning is not automatically a priori knowledge of its conclusion. From the empirical premise that every sample in a collection contains copper, together with the premise that a particular object belongs to that collection, a conclusion about the object may follow deductively. The inference can be valid while knowledge of the conclusion remains a posteriori because it depends on empirical premises. Logical validity concerns truth preservation from premises to conclusion, not the origin of the premises’ justification.\n\nApriority must also be distinguished from innateness. A belief might be innate yet false or unjustified; a capacity for reasoning might be innate without containing any explicit knowledge. Conversely, an a priori proof may require years of education before it becomes accessible. The psychological question of how a belief or capacity develops therefore differs from the normative question of what warrants a judgment. These distinctions prevent historical debates over innate ideas, conceptual truth, certainty, and necessity from being treated as a single dispute."
    },
    {
      "heading": "Logical empiricism and challenges to the distinction",
      "text": "In the early twentieth century, logical empiricism sought to reconcile respect for mathematics and logic with an empiricist account of substantive knowledge. Thinkers associated with the Vienna Circle generally rejected Kant’s synthetic a priori, although their views differed substantially. Mathematics and logic could retain a special status if their truths were understood as analytic, formal, or grounded in linguistic rules rather than as discoveries about a nonempirical realm. Empirical science, by contrast, supplied information about the world through observation and testing.\n\nRudolf Carnap developed accounts in which logical and mathematical statements were assessed within specified linguistic frameworks. In The Logical Syntax of Language (1934), formal rules played a central role in distinguishing kinds of sentences and consequences. This approach redirected attention from allegedly self-evident structures of reality to the construction and interpretation of languages. It did not simply declare every choice of framework true or false: practical considerations such as fruitfulness and convenience could help govern the adoption of a system, while statements within the system were evaluated according to its rules.\n\nWillard Van Orman Quine challenged this picture in “Two Dogmas of Empiricism” (1951). One target was a fundamental boundary between analytic truths grounded in meaning and synthetic truths grounded in fact. Attempts to explain analyticity through synonymy, definition, or semantic rules seemed to him either incomplete or circular unless the relevant notions could be independently clarified. His second target was a reductionist conception under which individual meaningful statements could each be associated with their own confirming experiences.\n\nQuine’s alternative emphasized confirmation holism: statements confront experience as interconnected bodies of theory rather than as isolated units. When prediction fails, different parts of a theoretical system can be revised. Some commitments, including elementary logic and mathematics, are more central and resistant to revision than ordinary observational beliefs, but centrality need not constitute a categorical exemption from empirical reconsideration. On this account, differences in revisability may be differences of degree rather than evidence of two fundamentally different kinds of truth.\n\nNeither Quine’s arguments nor logical empiricism ended the debate. Critics of Quine distinguish the possibility of revising a belief from the source of its present justification. A proposition might be justified a priori even if a thinker could later discover a mistake in the reasoning or alter its conceptual framework. Other philosophers defend restricted forms of analyticity without claiming that every meaningful sentence possesses a unique observational test. The continuing dispute concerns whether the proposed boundary is explanatory and principled, not merely whether scientific theories sometimes change."
    },
    {
      "heading": "Necessary a posteriori and contingent a priori truths",
      "text": "Saul Kripke sharply separated epistemological and modal classifications in the lectures published as Naming and Necessity in 1972 and as an expanded book in 1980. His account of names as rigid designators held that a name designates the same individual in every possible situation in which that individual exists. If two rigid names actually designate one and the same object, their identity is not merely accidental. Yet discovering that the names have the same referent may require observation.\n\nThe classic example is “Hesperus is Phosphorus.” The names historically associated with the evening star and the morning star both refer to Venus. Astronomical investigation establishes the identity; it is not obtained merely by analyzing the names. Nevertheless, if both names rigidly designate Venus, there is no possible situation in which Venus is distinct from itself. The identity is therefore necessary a posteriori. A situation in which different objects occupy the relevant observational roles need not be a situation in which Hesperus itself is not Phosphorus; this is the difference between an epistemic possibility and a metaphysical possibility.\n\nRelated examples concern natural kinds. The claim that water is H₂O was established through empirical investigation, not by examining ordinary speakers’ understanding of “water.” On influential essentialist accounts associated with Kripke and Hilary Putnam, chemical constitution is central to what makes a substance that kind of substance. A superficially similar liquid with a different underlying constitution would not thereby be water. Such examples depend on substantive assumptions about reference and essence; they are not neutral consequences of defining the word “necessary.”\n\nKripke also offered a contingent a priori example involving a stipulation that fixes the reference of a unit. Suppose “one meter” is introduced as the length of a particular bar at a specified time. Under the relevant stipulation, a suitably situated stipulator may know without measuring that the bar is one meter long at that time. Yet the bar could have had a different length. Once the unit has been fixed, it remains possible to describe a counterfactual situation in which that very bar is shorter or longer than one meter.\n\nThe example does not imply that the actual definition of the meter is still based on a bar; the International System of Units has defined the meter through the speed of light since 1983. Its philosophical function is to distinguish fixing a reference from asserting a synonymy or a necessary identity. Whether stipulations of this kind yield genuine propositional knowledge a priori remains disputed. Even so, the broader lesson has been influential: apriority cannot simply be identified with necessity, and empirical discovery can reveal facts that are not metaphysically contingent."
    },
    {
      "heading": "Accounts of a priori justification",
      "text": "A theory of a priori justification must explain why reasoning supplies knowledge rather than merely producing conviction. Traditional rationalist accounts appeal to intellectual insight: a thinker grasps a relation among concepts, properties, or propositions and thereby appreciates why a judgment must hold. The apparent immediacy of recognizing a contradiction or an elementary entailment provides a model. The difficulty is to explain what such grasping amounts to and how genuine insight can be distinguished from a deeply entrenched but mistaken appearance of necessity.\n\nOne influential family of accounts connects justification with understanding or conceptual competence. On these views, understanding certain concepts involves dispositions to accept particular judgments or patterns of inference. Competence with conjunction, for example, includes appreciating that if \\(p \\land q\\) is true, then \\(p\\) is true. This approach connects apriority to the constitutive conditions of thought rather than to a special perceptual faculty. Its challenge is to explain why a rule’s role in thought guarantees, or at least supports, its truth-preserving character.\n\nAnother family emphasizes entitlement or the legitimacy of basic inferential practices. Any attempt to justify every rule through a further argument threatens an epistemic regress: the argument itself must use some inferential rules. A proof of the reliability of classical logic, for instance, may already presuppose logical reasoning. Some philosophers conclude that certain basic warrants need not derive from further evidence. Others distinguish harmless rule-circularity from arguments that merely assume the very proposition they are meant to establish. No single account of these distinctions is generally accepted.\n\nReliabilist approaches instead connect justification or knowledge with reliable belief-forming processes. Reasoning could yield a priori knowledge when the relevant process is reliable and its warrant does not derive from experiential evidence. This shifts attention away from the subject’s ability to articulate an intellectual insight. It also raises a classification problem: if a process owes its reliability to evolutionary history or environmental regularities, does that make its products empirically justified? Many accounts distinguish the causal explanation of a capacity from the evidence on which a particular judgment rests.\n\nThere is also disagreement about defeasibility. An apparently valid proof can be undermined by discovering an equivocation, a hidden assumption, or evidence of a cognitive malfunction. If such discoveries defeat a priori justification, apriority does not entail indefeasibility. A further question is whether experience can directly rebut an allegedly a priori proposition or only show that its purported proof was defective. These disputes connect the topic with foundationalism, coherentism, and broader theories of rational belief without forcing every defender of apriority into the same epistemological system."
    },
    {
      "heading": "Mathematics, logic, and scientific frameworks",
      "text": "Mathematics supplies the most prominent examples of a priori knowledge, but its philosophical interpretation is contested. A proof establishes a conclusion from axioms and previously accepted results through specified inferential steps. Its force does not normally depend on measuring a large sample of cases. Checking many even numbers, for example, does not by itself establish a universal conjecture about every even number. Proof and computational confirmation can interact, but they play different epistemic roles.\n\nGottlob Frege and Bertrand Russell pursued logicism, the project of grounding arithmetic in logic. If successful in the intended form, it would explain arithmetic’s a priori status without Kant’s appeal to pure intuition. The discovery of paradoxes in unrestricted set-theoretic reasoning complicated this project, and rival programs offered different accounts. Formalists emphasized formal systems and consistency; mathematical intuitionism associated mathematical truth with constructive proof. These approaches disagree about mathematical objects, acceptable reasoning, and what a proof establishes.\n\nKurt Gödel’s incompleteness theorems of 1931 showed important limitations of sufficiently strong, effectively axiomatized formal systems. Under the relevant assumptions, such systems cannot settle every arithmetical truth, and an appropriate consistency statement cannot be proved internally by the system if it is consistent. These results do not establish that mathematical knowledge is empirical. They instead limit particular ambitions for complete formalization and complicate the relationship among truth, proof, and axiomatic knowledge.\n\nGeometry makes the connection with experience especially clear. Nineteenth-century non-Euclidean geometries showed that coherent alternatives to Euclidean geometry could be developed. Physics then distinguished the mathematical study of geometrical structures from determining which structures describe physical space or spacetime. General relativity, published by Albert Einstein in 1915, uses non-Euclidean spacetime geometry in a theory tested through observation. Theorems within a geometry can be established deductively, while the physical applicability of that geometry is an empirical question.\n\nThis development motivated accounts of a relativized a priori. Hans Reichenbach, in The Theory of Relativity and A Priori Knowledge (1920), distinguished the role of principles in constituting a scientific framework from the claim that they are unrevisable for all time. Michael Friedman later developed a historically sensitive account of constitutive principles that make empirical testing possible within a scientific framework. Such principles differ from ordinary hypotheses, yet can change during major scientific transformations. This usage broadens “a priori” beyond timeless rational truths and remains controversial among philosophers who reserve the term for justification independent of experience."
    },
    {
      "heading": "Testimony, computation, and prior probabilities",
      "text": "The distinction becomes less straightforward when knowledge depends on other people or technological instruments. A reader may know that a theorem is true because a reliable mathematician states it, without possessing its proof. Under a justification-based classification, the reader’s warrant appears to depend on testimony, including evidence about what was said and whether the source is trustworthy. The theorem may nevertheless remain knowable a priori. Social epistemology therefore distinguishes an individual’s immediate warrant from the broader practices through which communities establish and transmit results.\n\nLarge proofs make this issue concrete. The four-color theorem, established by Kenneth Appel and Wolfgang Haken in 1976, used extensive computer checking. Later work developed different proofs and formal verification. Such cases raise questions about whether knowledge of a theorem is wholly a priori when no single person surveys every computational step. The mathematical consequence relation may be independent of observation, while confidence that a particular machine correctly executed a program can depend on empirical assumptions about hardware, software, and error detection.\n\nFormal proof assistants introduce a related distinction. A small trusted kernel can check that a formal derivation conforms to specified rules, greatly reducing some forms of human error. Yet the claim that a physical computer produced the correct result is not identical with the abstract claim that a sequence is a valid proof. Philosophers disagree about whether technological mediation merely helps subjects access an a priori warrant or adds an indispensable empirical component to their knowledge. The answer may vary with the subject’s role, the proof’s transparency, and the intended standard of justification.\n\nThe words “prior” and “posterior” also have a technical use in probability and statistics that must not be conflated with the philosophical distinction. In Bayesian inference, a prior probability is assigned before conditioning on some specified evidence, while a posterior probability results from conditioning on that evidence. For a hypothesis \\(H\\) and evidence \\(E\\), with \\(P(E)>0\\), Bayes’s theorem states:\n\n\\[P(H\\mid E)=\\frac{P(E\\mid H)P(H)}{P(E)}.\\]\n\nThe labels concern a relationship to a particular update, not necessarily independence from all experience.\n\nA Bayesian prior may incorporate extensive historical data, earlier experiments, or expert judgment. A posterior distribution from one investigation can become the prior for another. Conversely, a principle used to justify a choice of prior might itself be defended through considerations alleged to be a priori, such as symmetry. The mathematical vocabulary and the epistemological vocabulary thus intersect without being synonymous: “before these data” does not mean “independently of experiential evidence.”"
    },
    {
      "heading": "Applications and continuing disputes",
      "text": "In ethics, the distinction concerns whether moral knowledge can arise through reasoning alone or must depend on experience. Kant’s moral philosophy assigns a central role to a priori principles, including the categorical imperative. Other traditions emphasize moral perception, sentiments, social practices, or knowledge of human flourishing. Even an account that treats a general moral principle as a priori need not classify every practical judgment that way. Applying a principle to a particular action usually requires empirical information about intentions, circumstances, consequences, and available alternatives.\n\nIn the philosophy of mind, a major issue is whether complete physical knowledge would make truths about consciousness knowable through reasoning alone. Arguments involving conceivability, explanatory gaps, and imagined cases often depend on claims about what is or is not a priori derivable. Some forms of physicalism maintain that mental truths are a priori entailed by a sufficiently complete physical description; others allow that necessary mind–body identities might be knowable only a posteriori. The modal and epistemological distinctions therefore matter directly to the interpretation of mind–brain identity theories.\n\nConceptual analysis and the thought experiment also bring apriority into philosophical method. Imagined cases concerning knowledge, identity, or responsibility are often intended to reveal consequences of concepts without requiring actual instances. The Gettier problem, introduced by Edmund Gettier in 1963, uses short cases to challenge the sufficiency of justified true belief for knowledge. Yet a judgment about an imagined case may draw on background beliefs about causation, psychology, or social practice. Merely using a hypothetical scenario does not establish that the resulting judgment has a wholly a priori basis.\n\nExperimental philosophy investigates how people respond to such cases, including whether judgments vary with wording, presentation, or population. These studies bear on claims about the stability and universality of philosophical judgments, but their epistemological implications require further argument. Evidence that people disagree does not by itself show that no answer can be justified a priori. Conversely, widespread agreement does not establish that the shared judgment results from nonempirical insight rather than common training or background assumptions.\n\nA central unresolved question is how to characterize independence from experience without making the category either implausibly strict or uninformatively broad. Requiring no experiential contribution whatsoever threatens to exclude ordinary human reasoning, which depends on learned concepts and cognitive development. Requiring only that a conclusion is not directly observed risks admitting many ordinary empirical inferences. Current accounts therefore focus on the precise role of experience: whether it enables understanding, supplies premises, warrants an inference, defeats a belief, or explains the development of a reasoning capacity. Different answers produce different boundaries for a priori knowledge, even where philosophers agree about the examples and the reasoning involved."
    }
  ],
  "links": [
    "epistemology",
    "knowledge",
    "epistemic-justification",
    "plato",
    "aristotle",
    "thomas-aquinas",
    "rationalism",
    "rene-descartes",
    "principle-of-sufficient-reason",
    "empiricism",
    "john-locke",
    "david-hume",
    "causation",
    "problem-of-induction",
    "immanuel-kant",
    "critique-of-pure-reason",
    "analytic-synthetic-distinction",
    "transcendental-idealism",
    "metaphysics",
    "possible-worlds",
    "deductive-reasoning",
    "logical-validity",
    "logical-empiricism",
    "vienna-circle",
    "rudolf-carnap",
    "willard-van-orman-quine",
    "saul-kripke",
    "water",
    "international-system-of-units",
    "classical-logic",
    "foundationalism",
    "coherentism",
    "mathematics",
    "gottlob-frege",
    "bertrand-russell",
    "physics",
    "hans-reichenbach",
    "probability",
    "statistics",
    "ethics",
    "categorical-imperative",
    "philosophy-of-mind",
    "consciousness",
    "physicalism",
    "thought-experiment"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}