Rational Design Method

Until the modern era, writers of architectural theory assumed no responsibility for providing a methodical procedure for the creation of a design. For the orders, yes, but not for the building as a whole. A strategy to achieve rational designs was set out by Jean-Nicolas-Louis Durand, in his Précis des leçons, which began to appear in 1802. Although the steps of his procedure were couched in terms of formal development, Durand's method assumed that the process was driven by the requirements of a functional program, even though—compared to later theory—his notion of such a program was confined to the macro level. Accordingly, the simplicity or complexity of the program would determine whether the building would be contained by a single mass or would include dependencies in the form of wings or attached pavilions. It would also determine whether the masses of the composition would be open or closed, accessible to the public or sequestered behind walls; and whether the dependencies would all be alike (if all had the same function) or would be differentiated (if they had various functions). If such a complex is to have multiple functions, which are to be dominant and which subordinate? How are they to be differentiated in size and in location within the complex? Having worked out such a scheme, the designer must then test it by justifying every aspect of it with regard to the program and its formal requirements. However, practical matters such as site characteristics, structural means, and choice of materials—all factors of serious concern to Durand—did not have an integral place in his method of design development.

At least as important as Durand's strategy for working out a design was his method of laying it out on paper (fig. 8-i). The process must always begin with the plan, never with the elevation. The elevation must grow out of the plan, but its regularity is safeguarded by the requirement that the plan be simple, clear, and symmetrical. In neither the plan nor the elevation should anything be included that does not contribute to the usefulness or meaning of the building. To lay out the plan he began with a grid, which established from the outset a modular system. He specified that walls were to be laid out on the grid lines and columns were to be centered over their intersections. If columns play a prominent role, the width of their intercolumniations could determine the dimensions of the grid module. These simple restrictions automatically discipline the configuration of the design with a degree of regularity and increase the likelihood of achieving pleasing proportions.

The urge to impose rational simplicity and regularity of this sort was consistent with the earlier imposition by the regime of Napoleon Bonaparte of uniform weights and measures in the form of the metric system. It expressed a perceived need in architecture for an alliance of art with practicality, at least in regard to function if not to the process of construction itself. This method could be applied equally to the classical tradition or to any of the other historical styles. Given Durand's outlook of cultural inclusive-ness, demonstrated earlier in his Recueil et parallèle des édifices de tout genre (Paris, 1800), design devices from all periods and cultures were fair game for appropriation. His main concern in that respect was that the artistic expres-

Figure 8-1

Design generated in plan and projected into the elevation and section: Durand (Précis des leçons, Paris, 1817).

Figure 8-1

Design generated in plan and projected into the elevation and section: Durand (Précis des leçons, Paris, 1817).

siveness of a building be consistent with its function and the sociological hierarchy of building types. To the extent that it provided an orderly procedure, Durand's design method was broadly useful, especially at a time when changes in society were fostering a proliferation of building types. Accordingly, it was incorporated into French academic tradition, where it remained into the twentieth century.

Was this article helpful?

0 0

Post a comment