New peer-reviewed conference publication, “Description Changes in Architecture: Identity Rules in Shape Grammars and Multiplicity in Design Datasets,” in Nexus Conference 2025: Relationships Between Architecture and Mathematics. The conference takes place at the ICEA – Civil, Environmental and Architectural Engineering of the Università degli Studi di Padova and Kim Williams Books, at the Palladio Museum, in Vicenza, Italy, in June 2025.

Long Abstract

Architectural design often evolves through changes or differences in object descriptions of architectural elements. Over the past half century, in particular, architects’ increasing focus on difference has driven paradigm shifts in architectural design (Ellinger, 2021). Francisco González de Canales describes a ‘Mannerist attitude’ in Renaissance artists, Robert Venturi, and contemporary practices, suggesting that architecture and art evolve by integrating new notions of difference in response to intellectual and cultural disruptions (González de Canales, 2023).

At the same time, there is an increasing use of data-driven Artificial Intelligence (AI) methodologies in architecture, art and design, which involve the curation, production or reuse of design datasets. Mario Carpo, in ‘Every Dataset is a Canon’ (2024), argues that, although architects since modernism have increasingly avoided mimicry, there is a rising interest in Generative Adversarial Networks (GANs) – generative modelling tools in machine learning widely used for data driven architectural image synthesis – that are primarily a “technology of imitation.” In this sense, advances in data-driven AI reintroduce architecture to the Renaissance approach of mimicking canonical forms, rather than continuing the trajectory towards “more difference.”

Today’s emphasis on difference in architecture invites a transhistorical examination of how it has been produced. This paper presents an original classification of case studies that illustrates how architects produce difference by altering the descriptions of the objects they mimic or study in their creative design work. These descriptive changes range from the formation of the Doric order to Romanesque, Gothic, Modern, Postmodern and contemporary architectural languages.

We represent description changes as rule-based computational processes with identity rules, a class of shape rules within the design computing theory area of shape grammars. Identities are used to systematically analyze description changes that stimulate important generative processes; from forming the triglyph to crafting a Gothic pinnacle or spatially and structurally organizing Terragni’s Danteum. The diversity of the examples demonstrates the relevance of descriptive change to all aspects of architecture: from purely structural to purely ornamental.

By raising the importance of description changes in architectural innovation, this paper argues for adopting a new lens for representing architectural knowledge in data. Specifically, one that embraces the possibility and necessity of changes in the description of architectural elements and acknowledges the multiplicity of their meanings.

Co-authors: Nicolaos Moustroufis (Architectural Association, MArch | History, Theory & Criticism track)

Downloads

Download preprint (Check back Soon)