In the article "Forms of Expression: The Proto-Functional Potential of Diagrams in Architectural Design," Greg Lynn discusses diagrams using the works of Ben Van Berkel as a case point. Lynn makes a point that the type of diagrams he refers to are not representational of ideas, but are conceptual tools. Lynn attempts to prove Van Berkel to be a Proto-Functionalist with regards to how he approaches architecture. Van Berkel's work is described as non-linear with respect to progression from the diagram to concrete constructions. Van Berkel's work looks into the vague influences in architecture. These vague influences being elements such are structural concerns and hidden infrastructures, elements that are not readily quantified in the initial design phases of architectural works, yet have a clear effect on architectural form. These vague influences are anexact in nature, being able to be broken down to their individual influences, yet not readily described as a conceptual whole. The word vague is a cue to how these forces are not so easily quantified and determined. In design, these influences present the chicken and egg problem; design must be carried on with respect to structure, but structure cannot be quantified without a corresponding form. These vague influences have been historical understudied by architects, but Van Berkel's work is done through a process that conceptualizes these influences via the diagram.
What separates Van Berkel's work is his use of abstraction in his work. What separates Van Berkel's version of abstraction from the traditional version is that his version is not reductive of an idea. Rather, his is an abstraction that is generative. These abstractions generate the form, rather than the form generating the abstraction. These abstractions represent the technological aspect; they present an understanding of technology in cultural and social contexts, as opposed to concrete, tangible forms. As the diagrams build on and feed off one another, eventually a diagram is produced that breaks the barrier between this form of technology and the concrete assemblage.
The idea of diagrams being the building blocks in architecture is nothing new. The recovered sketchbooks of architects long past can demonstrate this fact. However, the conceptual, generative diagram is a different approach. These diagrams generate form, rather than impersonate it. In a linear process, these diagrams would be meaningless abstractions. It would be next to impossible to directly generate form from them. Working in a kind of cycle, where the diagram evolves to bring in more concepts in approach to a form brings forth a process that attempts to reconcile the chicken and egg problem between architectural reality and architectural form. This idea of working with these vague influences has the benefit of allowing form to merge with structural necessity, breaking down the wall between drawn ideals and built forms. This method makes it possible to design forms that can be built, as opposed to idealized forms which only server the purpose of falsely advertising the reality of built form.