The Real
Ground truth proper — necessary and contingent — and why reality never waits for our axioms
Before asking what exists, ask what makes statements true. — a paraphrase of Michael Dummett’s programme.1 Dummett pressed the question toward anti-realism; this work takes the question and reaches a different answer.
The previous article distinguished atoms of truth — decisions, observations, conjectures, derivations. But those are what a system can hold. Beneath them lies what those records are about: the Real — ground truth proper, tier zero.
A thermometer’s reading is captured evidence (tier one). The patient’s actual thermal state is Real (tier zero). A judge’s verdict is captured evidence. Whether the accused acted as charged is Real. A proof logged in a journal is captured evidence. The mathematical structure the proof reaches toward is Real. Tier zero is never a record. It is what records and claims approximate.
In the ladder of Part III, this is tier zero. Everything above — captures at tier one, claims at tiers two through thirteen — is ranked by distance from the Real.
Rᶜ and Rⁿ — contingent and necessary
Two faces of tier zero, reached by different routes but equally Real:
Rᶜ — contingent Real. What could have been otherwise: the payment at 11:04:22, the fever, the apple’s trajectory, the patient’s lungs. Knowable in principle through experience — Kant’s a posteriori knowledge,2 the observable world. Laplace’s demon, given complete state and laws, could deduce the future in principle;3 in practice capture is always incomplete.
Rⁿ — necessary Real. What could not have been otherwise: logical structure, mathematical relations, the composition of quantities. Knowable in part independent of particular experience — Kant’s a priori,2 though Kripke showed necessary truth is not always knowable without observation (water is H₂O).4
Do not confuse Rⁿ with tier-two claims (“logically necessary relative to stated axioms”). The structure is Real; our proofs and axiom choices are captures and claims.
Necessary and contingent (in prose)
Contingent truth could have been otherwise. Almost everything a system observes or decides about is contingent.
Necessary truth could not have been otherwise — not because we stipulated it, but because structure any world must have. Confusing the two — dressing a contingent claim in necessary authority — is necessity laundering (named on the epistemic ladder).
They fail differently. Contingent truth is lost by not recording when we could have. Necessary structure is misreached by bad formalisation — but the structure itself was never ours to invent.
Logic and mathematics are discovered, not stipulated
It is tempting to treat mathematics as stipulation — axioms chosen, theorems true by convention. Carnap distinguished internal questions (inside a framework) from external ones (whether the framework fits the world).5 Formalists stop at internal truth. We do not.
Reality does not wait for our axioms. Energy, information, composition behaved as they behave before any proof was written. Our systems are best formalisations so far of Rⁿ — discovered, revisable, sometimes wrong.
Kurt Gödel showed that in formal systems rich enough for ordinary arithmetic (such as Peano Arithmetic), truths exist that no proof within the system can reach.6 Truth exceeds capture even in mathematics. Proofs inherit certainty only relative to fallible premises — tier-two claims, not tier zero.
Euclid’s fifth postulate was one choice among several; non-Euclidean geometries later described actual space. A formalisation that cannot express something true is a tool at its limit, not the last word on Rⁿ.
What makes a statement true
Dummett reframed metaphysics: ask what makes statements true, not only what exists.1 We adopt the question; we part from anti-realism.
- Regulative realism. Act as if mind-independent truth answers our statements — because without that pressure, inquiry has nothing to be wrong about.
- Fallibilism. Every formalisation, measurement, and theory is revisable.
- Openness. Hold the realism/anti-realism debate itself as open rather than letting either side foreclose inquiry.
Sharp analytic/synthetic and observation/theory divides are epistemically unsafe.7 Our provenance graph (claims link captures link Real) traces specific dependencies across levels — the stance of emergent physicalism, multi-level explanation, and emergent entities.
The one truth we cannot recover
Statements about the unrecorded past unite realist and anti-realist intuitions. If a contingent event happened and nothing captured it, it was still some way (Real) — but may be permanently beyond recovery.
Reconstruction from consequences is a best guess, not the fact.
The mandate:
- Never lose what can be known with certainty — unrecorded decisions and overwritten observations convert certainty into permanent guesswork. Lost certainty is the deepest failure this series names.
- Make remaining uncertainties smaller — and never dress a guess as substrate.
Reichenbach’s context of discovery (messy, psychological) versus context of justification (public, logical)8 parallels private cognition versus tier-one capture versus tier-two-plus claims — developed in Thinking by Writing.
Failure modes
Frozen axiom. Treating a formalisation as closed — refusing stronger axioms because the current frame “defines” truth.
Necessity laundering. Presenting contingent policy or model output as if it were necessary truth.
Lost certainty. Failing to record when capture was possible — irreversible damage.
Probabilistic truth mistaken for precise truth. Harding a reconstruction into substrate.
What this prepares
Tier zero is the Real — Rᶜ and Rⁿ — never held in a record, always larger than our formalisation. Tier one is captured evidence: how fragments entered the log — fallible, never “ground truth” in the tier-zero sense. The next article states what that substrate can and cannot claim: Captured Evidence.
Then: how agents think by writing (Thinking by Writing), the four capture modes (Four Capture Modes), and the moment conjecture becomes binding consequence.
Footnotes
-
Dummett, Truth and Other Enigmas (1978). Essays “Truth” (1959) and “What is a Theory of Meaning? (II)” (1976). ↩ ↩2
-
Kant, Critique of Pure Reason (1787 B ed.). Guyer & Wood (trans.), Cambridge University Press, 1998. Introduction IV; B1–18. ↩ ↩2
-
Laplace, Essay on Probabilities (1814). Introduction (demon passage). ↩
-
Kripke, Naming and Necessity (1980). Lectures II–III. ↩