For critical systems design, the verification tasks play a crucial role. If abnormalities are detected, a diagnostic process must be started to find and understand the root causes before corrective actions are applied. Detection and diagnosis are notions that overlap in common speech. Detection basically means to identify something as unusual, diagnosis means to investigate its root cause. The meaning of diagnosis is also fuzzy, because diagnosis is either an activity-an investigation-or an output result-the nature or the type of a problem. This paper proposes an organizational framework for structuring diagnoses around three principles: that propositional data (including detection) are the inputs of the diagnostic system; that activities are made of methods and techniques; and that associations specialize that relationships between the two preceding categories.