Coherence Theory of Truth

The basic idea

The correspondence theorist says: truth is matching reality. The coherence theorist replies: there is no way to compare a belief to bare reality. Whatever we are aware of, we are aware of through beliefs and perceptions, which are themselves further beliefs. Inquiry never escapes the web of belief. So truth, whatever it is, must be a feature of how beliefs hang together — not an external relation to a world we cannot reach un-mediated.

The starting motivation is epistemological: how would we ever check a correspondence? We'd need to compare a belief with the bare world, but every comparison is itself another belief. Bradley pressed this in Appearance and Reality (1893): any judgment we make about correspondence is itself just one more belief, evaluable by its coherence with the rest. The metaphysical conclusion follows: if coherence is what we ultimately have, coherence is what truth amounts to.

Bradley and Blanshard's system

F.H. Bradley (1846–1924), the most influential British idealist, argued that reality is a single all-encompassing whole and that any partial truth is therefore partly false. Truth comes in degrees: the more comprehensive and coherent a judgment, the more truth it has. The fully true system would be the complete description of the Absolute — and only it would be fully true. This is "degrees of truth," one of British idealism's most distinctive doctrines.

Brand Blanshard (1892–1987), in The Nature of Thought (1939, two volumes), gave the cleanest modern statement. For Blanshard, coherence is more than logical consistency: it requires that each member of the system necessitate the others, so that no proposition could be removed without unravelling the whole. The maximally coherent system is one in which everything mutually entails everything else — a structure he calls the concrete universal.

This is a demanding standard. No actual human belief system reaches it. Blanshard accepts this: truth is an ideal limit. We approximate it by progressively making our system more coherent.

Neurath and the Vienna Circle

The other historical strand comes from a very different place. Otto Neurath (1882–1945), member of the Vienna Circle, defended a coherence theory on radically empiricist grounds. The Circle's usual line — championed by Schlick — was that protocol sentences (basic observation reports) compare directly to experience and ground all other knowledge.

Neurath argued this picture is incoherent. Protocol sentences are themselves linguistic items; we never compare a sentence to bare experience but to other sentences (other observation reports, predictions, theories). Famously: "Statements are compared with statements, not with 'experiences,' not with a world, not with anything else." The 1932 boat metaphor — "We are like sailors who must rebuild their ship on the open sea" — became coherentism's slogan.

Coherence vs the alternatives

Theory	Truth-bearer relation	Champion(s)	What "truth" demands	Where it struggles
Correspondence	Proposition ↔ fact	Aristotle; Russell; Tarski	Match with worldly fact	Specifying "fact" without circularity
Coherence	Proposition ↔ system of propositions	Bradley; Blanshard; Neurath	Maximal mutual support	Isolation: many coherent worlds possible
Pragmatist	Belief ↔ inquiry / practice	Peirce; James; Dewey	What works at end of inquiry	Conflates truth with utility
Deflationary	Sentence ↔ disquotation	Ramsey; Quine; Horwich	Just the T-schema; no metaphysics	Can't explain truth's explanatory role
Identity theory	Proposition ≡ fact	Bradley (mature); McDowell	True propositions are facts	Strange ontology
Internal realism	Proposition ↔ idealised theory	Putnam (1981–1990)	What ideal theory would say	Putnam himself abandoned it
Constructivist	Proposition ↔ socially constructed system	Goodman; some sociologists of science	Endorsement by relevant community	Threatens objectivity entirely

A worked example: the murder mystery

Imagine a detective solving a murder. She has clues: a footprint, a missing watch, an alibi. She constructs a hypothesis: the butler did it. The hypothesis explains the footprint (his shoe size matches), accounts for the missing watch (he had a debt), and is compatible with the alibi witnesses (who are revealed to be lying). Each new clue either fits or forces her to revise.

The correspondence theorist says: the hypothesis is true iff the butler really did it — iff there is a fact in the past matching the proposition. The coherence theorist says: there is no separate access to that past fact. What the detective does, all she can do, is build a maximally coherent narrative integrating all the evidence. Truth, for the coherentist, just is winning that game maximally — having the most comprehensive, mutually supporting story.

The detective example reveals the appeal: this is roughly how scientific inference, historical reconstruction, and legal reasoning actually proceed. We never compare a theory directly to the unmediated world; we compare it to the totality of evidence. The coherentist asks: why posit a separate fact-relation when coherence-with-evidence is what we always rely on?

Counterarguments

• Russell's isolation objection. For any coherent system, you can construct an equally coherent rival. A well-written novel coheres. A consistent fairy tale coheres. Coherence alone cannot prefer reality to fiction. Bradley's reply — that the relevant system must be comprehensive, including all our experience — is itself contested, since multiple comprehensive systems are imaginable.
• The plurality problem. A generalisation of isolation. Pluralist worries say there might be infinitely many maximally coherent total systems with no fact of the matter to choose between them.
• The specification problem. "Coherence" needs definition. Mere consistency is too weak. Mutual entailment is too strong. Anything in between is contested. Critics charge that "coherence" is not yet a precise enough relation to fix truth.
• Realist intuitions. Most people's pre-theoretical conviction is that the world makes our beliefs true, not the other way around. The coherence theorist owes an explanation of why this strong intuition is mistaken.
• The verification trap. If truth just is coherence, and coherence is what we have epistemic access to, then whatever we believe with maximum coherence is automatically true. This conflates truth with justification.
• Modal worries. Could a falsehood once have been part of the maximally coherent set? Could the actual maximally coherent set differ from what is in fact the case? The coherentist's answer must be no — but explaining why is tricky.

Variants

• Idealist coherence (Bradley, Bosanquet, Blanshard). Coupled with the metaphysical doctrine that reality itself is a coherent whole.
• Empiricist coherence (Neurath). Coupled with the rejection of any comparison between sentences and bare experience.
• Internal realism (Putnam, 1981–1990). Truth is what an idealised theory would say; reality is mind-dependent in a sophisticated sense. Putnam later abandoned this for a "natural realism."
• Davidsonian coherence epistemology. Donald Davidson (1983) defends coherence-based justification while keeping correspondence as the theory of truth. He famously argued that most of our beliefs must be true if they are intelligible at all.
• Probabilistic coherence (Bovens & Hartmann, 2003). Modern formal accounts measure coherence with Bayesian probability.
• Pragmatist-coherence hybrids (Rescher 1973, 1979). Coherence is the criterion; pragmatic success drives the choice between coherent rivals.

Common confusions

• Coherence theory = consistency theory. No. Mere consistency is too weak — any random consistent set qualifies. Coherence requires positive mutual support.
• It denies an external world. Most coherentists do not. They deny only that we can compare beliefs to it un-mediated. The world's independent existence may be among the propositions of the coherent system.
• It is the same as coherentism in epistemology. No. Coherentism in epistemology is about justification; the coherence theory of truth is about truth. Many epistemological coherentists reject the truth theory.
• Coherentism makes truth subjective. Not necessarily — the maximally coherent system is supposed to be unique and inter-subjectively binding.
• Tarski refuted coherence. Tarski showed how to define a truth-predicate satisfying the T-schema; this is neutral between correspondence and coherence as metaphysical theories of what truth is.
• Coherentism = postmodernism. No. The classical coherence theorists (Bradley, Blanshard) were rationalist metaphysicians. Coherentism is compatible with absolute, objective truth.