Gödel’s Incompleteness Theorems
Takeaway
Any sufficiently strong, consistent formal system cannot prove all true statements about arithmetic (first theorem) and cannot prove its own consistency (second theorem).
The problem (before → after)
- Before: Hilbert’s program sought complete, consistent, finitely checkable foundations for all mathematics.
- After: Gödel shows inherent limits: truth outstrips provability; consistency cannot be proven from within.
Mental model first
It’s a legal code that cannot fully adjudicate all cases about numbers because you can construct a case that says, “This statement has no legal proof.” If the code is consistent, that case is true but unprovable.
Just-in-time concepts
- Arithmetization of syntax: Encode statements and proofs as numbers (Gödel numbering).
- Diagonalization: Construct self-referential sentences via fixed-point (the “this sentence is unprovable” trick).
- Ω-consistency vs consistency: Technical conditions ensuring no contradictions.
First-pass solution
Build a sentence G that asserts its own unprovability. If the system proves G, it proves a falsehood (inconsistency). If it proves ¬G, it proves that G has a proof, contradicting consistency. Hence, if consistent, the system proves neither: it’s incomplete.
Iterative refinement
- Rosser’s improvement weakens assumptions to plain consistency.
- Second theorem: Formalize “the system is consistent” and show it unprovable if the system is consistent.
- Scope: Applies to any system interpreting a fragment of arithmetic (e.g., PA).
Principles, not prescriptions
- Self-reference via arithmetic encoding exposes system limits.
- Completeness in logic (Gödel’s completeness theorem) does not imply completeness in arithmetic truth.
Common pitfalls
- Confusing semantic truth with syntactic provability.
- Overextending to everyday reasoning without care.
Connections and contrasts
- See also: [/blog/halting-problem-undecidability], [/blog/p-vs-np-problem], [/blog/cantors-diagonal-argument].
Quick checks
- Why “sufficiently strong”? — Needs to express basic arithmetic.
- Does this refute all formal systems? — No, but any that capture arithmetic face incompleteness.
- Can a stronger system prove consistency of a weaker one? — Yes (relative consistency).
Further reading
- Gödel’s original paper; Nagel & Newman, “Gödel’s Proof”