The Undecidability of the Halting Problem: Computability's Foundational Limit — E8 Intelligence Research
Year: 2026
Type: preprint
Abstract: FINDING: The halting problem is undecidable — no algorithm can determine whether an arbitrary program halts; this is the foundational limit of computable knowledge. | MATH: Formalized via diagonalization: assume HALT exists, define D(x) = loop if HALT(x,x) else halt; then D(D) yields contradiction. Equivalent to non-existence of a total computable function h: ℕ²→{0,1} deciding membership in the halting set K = {e | φ_e(e)↓}. Semidecidability: K is recursively enumerable but not recursive; complement is not r.e. | CONNECTION: The proof structure mirrors a fixed-point/retraction obstruction: no continuous (computable) map from the space of programs to {halt, loop} can be invariant under self-application — analogous to Brouwer's no-retraction theorem for the disk. The diagonal argument is a discrete analogue of the golden-ratio-like self-similarity (self-reference) but yields a *gap*, not a ratio. No direct 0.618/1.618 appears; however, the halting set's Turing degree 0′ is the first jump Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
Author: Caldin, Andrew Stewart
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7212301752
Cites:
Cited by:
Related to: