Why We Laugh: Humor as Forced Gist Integration and the Information-Theoretic Structure of Comedy

Why is anything funny? For 2,500 years—from Aristotle's superiority theory through Freud's relief theory to contemporary incongruity-resolution models—the question has resisted formal answer. This paper provides one. Using the Adaptive Compression Advantage Theory (ACAT; Murata, 2026), humor is formalized as forced gist integration: the comedian destroys the audience's verbatim prediction and simultaneously presents a deeper gist that the audience's brain must integrate in real time. The reward signal from this forced integration is experienced as laughter. This formalization explains why incongruity alone is not funny (horror films are also incongruent), why timing ('ma' in Japanese comedy) is critical (it controls the audience's verbatim prediction inflation), why comedy genius is unteachable through instruction (the compression function cannot be transmitted verbatim), and why AI cannot yet generate humor (humor requires modeling the audience's compression state, which requires a self-model that current AI lacks). Five testable predictions are provided.