This paper introduces Causonomy — a formal science of failure and causation in normative systems — developed by Pascal Etcheber. It constitutes the foundational paper for the science. Problem solving is universally practised yet has never been grounded in a formal scientific foundation. Existing approaches — root cause analysis, fault trees, FMEA, HAZOP, statistical methods, and systems thinking — rely on heuristic search, domain expertise, or probabilistic inference because failure itself has never been defined as a closed and enumerable object. Without such a definition, completeness of causal explanation cannot be guaranteed, diagnosis cannot terminate by necessity, and recurrence cannot be structurally prevented. This paper establishes that such a definition exists and derives from it a complete formal science. The object of the science is the Negative Outcome: the measurable deviation D(f) = R(f) − O(f) between a required state and an observed state at the boundary of an activity, where D(f) ≠ 0. This definition locates failure in a structural relation rather than in domain language, making it universal, precise, and observable across all normative systems. The derivation proceeds in four stages: From a primitive ontology of form (structure, quantity) and activity (existence, magnitude, time), the paper derives a closed grammar of exactly twelve deviation types. These twelve deviations, combined with six universal activity classes (Store, Move, Acquire, Release, Transform, Check), yield a finite and complete space of seventy-two Negative Outcomes — the complete surface of observable failure in any normative system. No additional failure type can exist outside this space. Causation is derived as the failure of necessary conditions supporting activity. Four support types (Process, Organisation, Tools, Data) combined with three activity failure modes (Existence, Magnitude, Timeliness) yield twelve structural failure positions. Their product with the seventy-two Negative Outcomes produces a closed operational cause space of 864 positions — every distinct way in which any observable failure can arise. Root cause is located at the governing layer, where required states are defined and authorised. Governing forms are structured by six governing functions (SFRCAV), four requirement primitives (RACE), and eight lifecycle states (LIFESPAN). Their combination yields a governing-form universe of 1,152 structurally distinct root-cause positions. A Structural Reduction Matrix (SRM) encodes the admissible mapping from governing defects to observable failures through two necessary conditions — projection (identity perturbation) and connectivity (structural reachability within the system) — determining admissibility a priori before any evidence is gathered. From this structure, four reasoning operations are derived as traversals of the same system: diagnosis (failure to admissible causes), prediction (governing defect to admissible consequences), stress testing (system structure to complete failure exposure), and preventive design (modification of governing structure to eliminate admissible failures). Each proceeds by elimination within a closed space rather than by hypothesis generation. The central results are: a closed failure space (72 Negative Outcomes), a closed operational cause space (864 positions), a closed governing-form universe (1,152 root-cause positions), a pre-computable Structural Reduction Matrix, and a deterministic reasoning system admitting formal termination conditions. The framework is falsifiable: it would be invalidated by any failure, cause, or admissible relation outside the defined spaces. Scientific status is established by four criteria: a precisely defined object, a finite domain derived from structural necessity, necessary relations governing the domain, and explicit falsifiability conditions. The framework is independent of domain content — the same structure governs failure in manufacturing, healthcare, aviation, software, finance, and any other system operating under requirements. The work transforms problem solving from a heuristic practice into a deterministic discipline. Brainstorming and open-ended causal search are replaced by classification within a known structure and elimination through evidence. The science provides not only a method for explaining failures but a foundation for designing systems in which certain failures cannot structurally occur.
pascal etcheber (Mon,) studied this question.