Interpreters often cite Kant’s Stufenleiter of representations (A320/B376) as providing a “definition” of intuition. This misunderstands the peculiar logic of Porphyrian classification, which I clarify by reviewing its history. Porphyrian trees do offer conceptual analyses, I argue, but do not purport to provide a uniquely correct, much less exhaustive, account of the analysand. Different orderings of differentia, as well as different differentia, are often possible. Which terms and which kinds of terms appear as differentia in the analysis depends on the goals and constraints on the philosophical inquiry to which the analysis contributes. It is therefore illegitimate to extract a “definition” of intuition from Kant’s Stufenleiter. First, its analysis targets <idea of reason> (not <intuition>), and, second, it does not purport to give a uniquely correct account of that concept (much less of <intuition>), as a definition must.