fallibilism

Works

Three formulations, a century apart at the ends.

• Peirce's 'The Fixation of Belief' (1877) and 'Some Consequences of Four Incapacities' (1868) introduce fallibilism as the engine of inquiry rather than its embarrassment, with the community of inquirers as the unit that converges.
• Popper's The Logic of Scientific Discovery (1934) and Conjectures and Refutations (1963) supply falsifiability as a demarcation and the institutional argument that criticism, not verification, produces reliability.
• Quine's 'Two Dogmas of Empiricism' (1951) supplies confirmation holism: beliefs face experience as a web, so no single claim is refuted in isolation — which complicates falsification without abandoning fallibility.
• The tradition's practical wing runs through Merton's norms of science and modern replication reform, where the argument moved from epistemology to institutional design.

Central ideas

Four ideas, and the third is what keeps the tradition from collapsing into doubt.

• Fallibility without scepticism: that a belief could be wrong is compatible with its being justified now, and acting on it is not a contradiction.
• Conjecture and refutation: theories are proposed boldly and tested severely; a theory that forbids nothing has said nothing.
• Self-correction as a property of communities, not persons: no individual inquirer is reliable, and reliability is an achievement of arrangements that let others find one's errors.
• Holism: because beliefs are adjusted as a web, a failed prediction underdetermines which belief should go — which is why refutation is a judgment call rather than an algorithm.

Distinctive vocabulary

Five terms, each guarded.

• Fallibilism: the view that any belief may be mistaken. Not the view that no belief is justified.
• Falsifiable: forbidding some observable outcome. Not false, and not disproven.
• Self-correcting: converging through public criticism over time. Not automatically improving.
• Conjecture: a serious proposal held open to refutation. Not a guess.
• Underdetermination: evidence compatible with rival theories. Not evidence being useless.

Debates and disagreements

The tradition's quarrels are about whether its own method survives contact with practice.

• Falsification against holism: Quine's web means a failed test can always be accommodated elsewhere, so Popper's clean refutation rarely occurs — the Duhem-Quine problem is the standing internal objection.
• Whether science actually works this way: Kuhn's paradigms and Lakatos's research programmes describe practice in which anomalies are tolerated for long stretches, which fits the history better than naive falsificationism does.
• Whether self-correction is real or a professional flattery: the replication crisis was the tradition's own norms discovering that its institutions were not implementing them — evidence both for the ideal and against its automatic operation.
• Whether fallibilism is livable: acting requires commitment, and the objection that permanent provisionality is a posture rather than a practice is answered only by showing what it changes in conduct.

Intellectual relationships

Its neighbours in this record are the epistemic lenses and the correction machinery.

• The epistemological lens documents the wider question of what knowing is; this tradition documents the specific stance of holding beliefs correctably, and the two are cited for different work.
• Historical source criticism, documented separately here, is fallibilism's method applied to testimony — provenance, chain, and interest as the checkable properties of a claim.
• The hermeneutic lens is the counterweight: understanding is historically situated in ways no amount of criticism dissolves, and the record keeps both admissible rather than letting method claim neutrality.
• The media-theoretic lens supplies the hard floor: correction requires a record, and what was never recorded cannot be corrected.

How it changes this reading

Ten placements — this tradition underwrites the record's whole kind of material.

• The claims about correcting in the open, naming the gap, and accuracy over affirmation are the tradition's norms stated as personal conduct rather than as scientific procedure.
• The claims about evidence before pattern, primary before secondary, and understanding before intervening are its sequencing rules: error is cheapest to catch early and in the shortest chain.
• The claim about complete artifacts is its precondition — a self-correcting practice needs a record intact enough to be corrected against.
• The through-lines extend it to relationship: frame-correction over self-defence is criticism applied to conditions rather than conclusions, and architecture over perception-management is Popper's institutional argument at domestic scale.

Useful comparisons

Against its neighbours.

• Against scepticism: same premise, opposite conclusion — fallibilism treats corrigibility as the reason to inquire rather than the reason to abstain.
• Against dogmatism: not a difference in confidence but in what the confidence is attached to — a fallibilist can be certain in practice while holding the belief revisable in principle.
• Against relativism: fallibilism holds that beliefs can be better or worse supported, which is exactly what a relativism denies; the two are frequently confused and are opposites.

Where the ideas meet

What the formulations share.

• Error is normal: the tradition's founding move is to make being wrong an ordinary state rather than a disgrace.
• Criticism is the engine: reliability comes from exposure to correction, not from care taken in private.
• Inquiry is collective and temporal: no single act of checking settles anything; the unit is a community over time.

Where they part

Where they part.

• On convergence: Peirce's long run converges on truth; Popper's approaches it without a guarantee; Quine's web has no such destination built in.
• On the unit of testing: a hypothesis (Popper) against a whole web (Quine) — a difference that decides whether refutation is decisive or negotiated.
• On method versus institution: whether reliability is secured by a rule inquirers follow or by arrangements that make error visible whatever they intend.

Limits

What the tradition cannot deliver.

• It cannot tell you which belief to revise: holism guarantees that a failed expectation leaves the choice open.
• It cannot supply the resources correction needs: openness is cheap to endorse and expensive to practice, and the tradition is silent about who pays.
• It cannot license the inference from corrigibility to equal doubt: everything being revisable does not make everything equally uncertain, and treating it so is the tradition's most common corruption.

Criticisms

Standing objections, at strength.

• Against Popper: falsificationism describes little actual science; scientists rightly protect promising theories from early anomalies, and Kuhn and Lakatos both showed why.
• Against Peirce's convergence: the long run is unfalsifiable as a claim about inquiry, and there is no argument that a community's beliefs must converge on anything.
• Against the tradition's rhetoric: self-correction is invoked far more often than it is demonstrated, and institutions cite the ideal as if it were an achievement.
• Against its social blindness: whose criticism gets heard is a question of standing and power, which fallibilism's classical statements do not address — the standpoint tradition documented here does.

Common misreadings

Four, and the first two are the ones this record guards.

• Fallibilism as universal solvent: 'we could all be wrong' deployed to level a real asymmetry in evidence — the move that borrows humility's language to protect a weaker claim.
• Falsifiability as a slur: labelling a claim unfalsifiable to end discussion, when the demarcation was a criterion for science and not a test of meaningfulness.
• Provisionality as evasion: holding everything open in order never to commit, which the tradition's own emphasis on bold conjecture forbids.
• Correction as performance: announcing revisability without ever revising anything, which is the norm worn as a credential.

What remains outside this idea

The tradition's own boundary.

• That inquiry converges: the long-run claim is a hope with a structure, not a result.
• That any particular belief is wrong: fallibility is general, and asserting a specific error requires specific evidence.
• That a practice is self-correcting because it says so: the property is demonstrated by corrections that actually landed, and nothing less.