One diagonal fixed point sits behind Cantor, Russell, Gödel, Tarski, Turing and the recursion theorem (Lawvere 1969). The same engine forks by a single question — can self-reference close? On the wall side (μ) it cannot: the argument runs as a proof that no fixed point exists. On the floor side (ν) it does: the fixed point is produced — and at ⊥ it is the framework’s own.
One engine, two fates. Every argument here is the same diagonal fixed point under one question — can self-reference close on itself? On the μ wall side the answer is no, and the classical argument (Cantor, Russell, Turing, Tarski, Curry) is the proof that no fixed point exists. On the ν floor side the answer is yes: the fixed point is produced (the Quine atom, the Kleene quine, Löb, Gödel’s second, Rice). Gödel’s first theorem sits between — the diagonal sentence the engine builds. The μ wall faces plus Löb and Gödel’s second are axiom-free; the two computability floor faces (Kleene quine, Rice) carry Classical.choice from Mathlib’s recursion theory.
What is drawn vs. what is claimed. The unification is Lawvere’s (1969), restated by Yanofsky (2003) — the arguments as one scheme is prior art, cited not claimed. This figure is a placement of the framework’s ⊥ among these recognized arguments, not a new theorem; the cross-face identity stays a type boundary — the same shape, provably distinct carriers. The one floor face that is the framework’s own bottom is the Quine atom ⊥={⊥}, the seam the Bottom Family tree grows from.
← The Bottom Element