Synthetic A Priori

The problem it answers

What kinds of knowledge are humans capable of having? The traditional distinctions are two. Analytic judgments are true in virtue of the meanings of their terms (all bachelors are unmarried); they extend our understanding only by clarifying what is already contained in the concepts we have. Synthetic judgments add to our understanding by connecting concepts in ways that the concepts themselves do not require (the cat is on the mat); they extend our knowledge beyond the analysis of meaning. Cross-cutting this distinction is the distinction between knowledge that is a priori (knowable independently of experience) and knowledge that is a posteriori (knowable only through experience).

The combinations: analytic a priori (true by definition, knowable without experience; all bachelors are unmarried is the standard example); synthetic a posteriori (factual knowledge requiring experience; grass is green). The puzzle: is there synthetic a priori knowledge — knowledge that both extends our understanding and is justifiable independently of experience? This is Kant's central question in the Critique of Pure Reason (1781), and his answer (yes, and the conditions under which it is possible are the conditions Kant aims to articulate) organizes the entire Critical philosophy.

The core claim

The core claim has three parts.

Synthetic a priori knowledge exists. Mathematics provides clear cases (7 + 5 = 12 is, Kant argues, synthetic rather than analytic; the concept 12 is not contained in 7 + 5 but requires the synthetic combination of the two through the form of intuition). Pure natural science provides further cases (the principle that every event has a cause is, Kant argues, synthetic a priori). Some claims of traditional metaphysics, if they could be established, would also be synthetic a priori.

Synthetic a priori knowledge is possible because the mind contributes structuring conditions to experience. Space and time are not features of mind-independent reality; they are the forms of intuition through which the mind receives sensory data. The categories of understanding (substance, causation, etc.) are not features of mind-independent reality; they are the structures the understanding applies to any object of experience. These structures, because they are conditions of any possible experience, can be known a priori; because they extend our understanding (they tell us what any possible object must be like), the knowledge is synthetic.

The bounds of synthetic a priori knowledge are the bounds of possible experience. What can be known a priori about objects is what the mind contributes to the structure of any possible experience of objects. What lies beyond possible experience (the noumenon, the thing-in-itself, the traditional objects of metaphysics: God, the immortal soul, the world as a totality) cannot be known synthetically a priori, because the conditions the mind contributes apply only within possible experience.

History in one paragraph

The distinction between analytic and synthetic judgments is original to Kant in its developed form, though predecessors are visible (especially in Hume's distinction between relations of ideas and matters of fact). Kant introduces the systematic distinction in the Critique of Pure Reason (1781) and the Prolegomena to Any Future Metaphysics (1783) as the foundation of the Critical philosophy's central question. The nineteenth-century reception was substantial; the German philosophical tradition through Fichte, Schelling, and Hegel accepted the framework while developing it in directions that complicated the original distinctions. The major twentieth-century challenge to the synthetic a priori came from logical positivism, especially through W.V.O. Quine's Two Dogmas of Empiricism (1951), which attacked the analytic/synthetic distinction itself — if there is no clear analytic/synthetic distinction, the category of synthetic a priori has no clear content. Quine's argument was substantially influential and was widely taken as having undermined the Kantian framework. The subsequent literature has been more complicated; the Quinean argument has been challenged (Paul Boghossian's Analyticity Reconsidered, 1996); the analytic/synthetic distinction has been substantially restored in modified form; the contemporary literature on synthetic a priori knowledge is active in epistemology and philosophy of mathematics. Saul Kripke's Naming and Necessity (1980) complicated the picture further by arguing for the existence of necessary a posteriori truths (water is H2O) and contingent a priori truths (the standard meter is one meter long) that the Kantian framework had not accommodated.

The four classes of judgment

Kant's framework generates four logically possible classes; he argues that three are actually instantiated and the fourth is empty.

Analytic a priori. Universally accepted as instantiated. All bachelors are unmarried is true by the meaning of the terms and knowable without any particular empirical investigation.

Synthetic a posteriori. Universally accepted as instantiated. Snow is white extends our knowledge beyond conceptual analysis (the concept snow does not by itself contain white) and requires empirical investigation to know.

Synthetic a priori. Kant's central claim is that this class is instantiated. The instances include the principles of mathematics, the principles of pure natural science (the principles like every event has a cause that the Critique of Pure Reason's Analytic of Principles derives), and some claims of traditional metaphysics that the Critique's Transcendental Dialectic shows actually fail to deliver knowledge.

Analytic a posteriori. The remaining logical possibility, which Kant treats as empty. A judgment that is analytic (true by meaning) does not require experience to be known; the meanings are already given. The class is empty by the structure of the distinctions.

How synthetic a priori is possible

The central question of the Critique ("How are synthetic a priori judgments possible?") receives its answer through the transcendental analysis. The mind contributes the forms of intuition (space and time) and the categories of understanding to any possible experience. These contributions are a priori (they are not derived from experience but are conditions of any possible experience); they are synthetic (they extend our understanding by telling us what any possible object must be like, not merely by analyzing the meanings of concepts we already have).

The argument therefore turns on the doctrine of transcendental idealism: the world we know is the world as structured by the mind's contributions, and what we can know synthetically a priori about this world is exactly what the mind contributes. The bounds of synthetic a priori knowledge are therefore the bounds of possible experience; beyond those bounds, the apparent synthetic a priori claims of traditional metaphysics actually fail to deliver knowledge.

The Quinean challenge

W.V.O. Quine's Two Dogmas of Empiricism (1951) is the central twentieth-century challenge. Quine attacks the analytic/synthetic distinction itself: there is no clear way to draw the line between truths that hold by meaning and truths that hold by fact, because the meanings of terms are themselves shaped by empirical considerations. If the analytic/synthetic distinction collapses, the category of synthetic a priori has no clear content.

The argument has been substantially influential; for much of the late twentieth century, the synthetic a priori was widely treated as a pre-Quinean category that had been superseded. The more recent literature has been more complicated. Paul Boghossian, Frank Jackson, and others have defended versions of the analytic/synthetic distinction against the Quinean attack; the contemporary literature on a priori knowledge (Laurence BonJour's In Defense of Pure Reason, 1998) has restored substantial interest in synthetic a priori knowledge.

Common confusions

A priori is not the same as innate. A priori knowledge is knowable independently of experience for its justification; it need not be innate (present in the mind from birth). Kant is clear that the categories may require experience to be brought to consciousness while remaining a priori in the relevant sense (not requiring experience for their justification).

Synthetic a priori is not the same as necessary truth. The distinction between necessary and contingent (which way could the proposition have been) is logically independent of the distinction between a priori and a posteriori (whether experience is required for justification). Kripke's Naming and Necessity showed that the two distinctions can come apart in unexpected ways.

The category does not require Kant's specific transcendental idealism. Some contemporary defenses of synthetic a priori knowledge (especially in philosophy of mathematics) do not commit to the broader Kantian framework. The category names a kind of knowledge; what explanation of how such knowledge is possible should be given is a separate question.

Live debates

The Quine-Carnap debate and its descendants. The contemporary literature on the analytic/synthetic distinction continues to engage Quine's original challenge.

A priori knowledge in mathematics and logic. The status of mathematical and logical truths as synthetic a priori remains contested, with substantial work in both directions (the Hilbertian/formalist tradition treats them as analytic; the Gödelian/Platonic tradition often treats them as synthetic a priori).

The new contingent a priori and necessary a posteriori. Kripke's expansion of the framework has generated substantial subsequent literature on how the various distinctions interact.

Contemporary engagement

Major recent scholarly work includes Laurence BonJour's In Defense of Pure Reason (1998), Paul Boghossian's papers on analyticity, Tyler Burge's substantial work on a priori knowledge, and the relevant chapters in the Cambridge Companion to Kant (Paul Guyer, ed., 1992). The Stanford Encyclopedia of Philosophy entries on a priori knowledge and on the analytic-synthetic distinction document the contemporary state of the debate.

Further reading

• Kant — the originator of the technical framework
• Critique of Pure Reason — the central text
• Transcendental Idealism — the metaphysical framework that explains how synthetic a priori is possible
• Episteme — the broader cognitive achievement
• Hume — the predecessor whose distinction between relations of ideas and matters of fact Kant transformed
• German Idealism — the tradition

This is a Pillar concept. Satellite concepts should link here as their parent.