The twentieth-century accounts of typology are often both historiographically problematic and conceptually imprecise. They reinforce an understanding of typology as mainly an interchangeable functional and graphic classification, and present Jean-Nicolas-Louis Durand as a key figure of the discourse, despite him dealing with buildings according to their genre and not their organisational and structural diagrams of typology. In contrast, one can posit that all theories of type are foremost epistemological and discursive arguments. Although not prescriptive in a formal sense, they are concerned with a rational synthesis of form by thinking through conceptual and diagrammatic organisation. This diagrammatic abstraction had already become instrumental to architectural theory and history in the eighteenth century, long before the modern discourse on the diagram was consolidated in the 1990s.

While the architectural diagram is regularly explained as a generic and generative description, it can equally be defined as a typological diagram specific to the architectural discipline and its production of knowledge. Clarifying the concept of type as emerging in parallel with ideas of abstraction and diagrammatic reasoning reveals a richer set of connected problems deriving from architectural practice, pedagogy and disciplinary knowledge, which permits a different framing of the historical discourse. This is explored by discussing its meaning for a distinction between typal and typological reasoning, how this arises from a problem of history and theory, and how the evolving typological discourse relates to the concepts of invention, disposition and style. Whereas historiography commonly recognises the French academics Antoine-Chrysostome Quatremère de Quincy and Durand, the often-overlooked Gottfried Semper and Julien-David Le Roy were central to a modern conception of architecture that developed ideas of typal and typological abstraction through historicist processes of cultural and diagrammatic reduction.