Computational considerations of historical architectural analysis: A case study of Chinese traditional architecture

This dissertation is about the fusion of key elements from computation and historical architecture. It builds upon ideas from shape grammars and employs the analytical methodology of historical analysis. This combination can help produce a different experience, through visual and ultimately interactive explorations, of historical architecture, particularly from those cultures where drawings were not the established mode of communicating designs. Grammatical analyses can be used to differentiate between various architectural styles, and explore relationships between styles; these lend themselves to computational models.;There are three parts to this dissertation. The first part is a case study of southern style Chinese vernacular architecture (the domain selected to demonstrate the claims of this dissertation) and deals with concepts from geomancy and its application to traditional architecture. The second part deals with the formal and computational considerations of historical architectural analysis. A shape grammar of vernacular dwellings is developed. This grammar is illustrated for five existing buildings. An extension to this grammar for traditional temples is explored and developed. Lastly, an aspect of shape grammar implementation is tackled. A (programming) language for describing shapes is developed and its design, implementation and use described.;The main contributions of this dissertation are: (1) A new approach to historical architectural analysis, illustrated by way of Chinese traditional architecture. This approach is particularly applicable to designs which have not been documented in any formal way. (2) A shape description language, an interpreted programming language for the specification of shapes and rules augmented with non-geometric information.;There are other implicit contributions of the dissertation to historic preservation; the design of contemporary Chinese architecture; and the teaching of architectural history.;On the whole, this dissertation presents a new approach to the study of historical architecture, one that facilitates investigation from a number of distinct vantage points, integrated into systems of compositional rules for exploration and comparison, and amenable to computer implementation, for the primary purpose of visually and interactively exploring architectural styles.