Truth Machines
Summary
This essay traces the lineage from Turing’s computational formalism through Tarski’s semantic definition of truth to Gödel and Chaitin’s incompleteness results, arguing that together they establish Conditionalism: truth is not absolute but conditional on interpretation. Computation manipulates syntax without intrinsic meaning; Tarski showed truth is relative to interpretation; Gödel/Chaitin proved systems cannot fully capture their own semantics. The synthesis: computation IS interpretation, and truth is a measure of coherent interpreter-pattern matching within finite resources.
Key Concepts
1. Syntax Without Meaning
- Turing machines manipulate symbols as syntactic tokens with structure but no intrinsic semantics
- Can manipulate “0011” and “⊢P→Q” identically as mere patterns
- Original schism: computation explains form, not meaning
2. Tarski’s Bridge Formal definition of truth: “A statement ‘P’ is true under an interpretation I if and only if P holds in the model I.”
- Truth is not absolute; evaluated relative to interpretation
- Even “Snow is white” is only true given specific referents
- Formally:
I ⊢ P(truth conditional on interpretation)
3. Gödel and Limits of Absoluteness
- Sufficiently expressive formal systems contain true but unprovable statements
- Exposes semantic dependency: truth of propositions depends on interpretation system cannot fully capture
- Chaitin: boundary of provability = boundary of compressibility
- No system carries its own semantics; interpretation leaks out of syntax
4. Computation as Interpretation Inversion: computation IS interpretive act
- Given input P and program I: output O = I(P)
- Program I interprets P semantically
- Every evaluation is mapping from pattern to meaning under interpreter
- Change program → change meaning
Conditionalism rendered operational: “There are no absolute truths, only consistent interpreter-pattern pairs.”
5. Information, Compression, and Meaning
- Shannon: entropy measures surprise, not meaning (intentionally semantic-free)
- Chaitin: incompressible numbers resist interpretation (no compact description)
- Meaning arises when pattern admits compact interpreter that reconstructs it predictively
- Therefore: Interpretation = compression
- Truth = measure of how coherently interpreter maps pattern to outcome within finite resources
6. The Conditional Turn
Core equation: I ⊢ P
- When philosophers speak of “unconditional truth,” they assume hidden universal interpreter (divine Tarski model)
- Conditionalism removes that ghost
7. The Meta-Interpreter
- Every reasoning act presupposes interpreter evaluating another interpreter
- Gödel’s self-reference + Turing’s halting problem: system cannot perfectly interpret itself
- No ultimate vantage
- Conditionalism: recursion without closure
- Truth, like computation, is locally consistent but globally ungrounded
Unified Formula
Truth = Conditional Evaluation
Truth is not a property of statements, but of evaluations.
To compute is to interpret, and to interpret is to condition.
Every truth lives inside an interpreter. Outside interpretation: “unparsed chaos—patterns without meaning, syntax without truth.”
Relation to Axio Framework
Critical foundational post connecting:
- Conditionalism: truth as conditional on interpretation (core Axio epistemic principle)
- Chaos Reservoir: unparsed patterns lacking interpretation = chaos
- Coherence: interpretation as compression of patterns
- Physics of Agency: agents as interpreters mapping patterns to actions
- Beliefs: beliefs as interpretive models within models
- QBU: measurement as interpretation event conditioning truth of outcomes
This establishes the formal/computational foundations for Axio’s epistemology.
Philosophical Implications
Against Correspondence Theory: Truth isn’t matching reality to propositions; it’s coherent mapping within interpretive framework.
Against Foundationalism: No bottom-level absolute truths; only conditional truths within interpretation contexts.
Recursion Without Ground: Infinite regress of meta-interpreters isn’t a bug—it’s how meaning works. No privileged vantage.
Computation and Semantics Unified: Dissolves traditional divide: computation becomes semantic act; meaning becomes computational process.
Processed on 2026-02-10 as part of batch 26-50