Behavioural Architecture: Reaction, Conversation, Interaction

This article examines behavioural architecture capable of response, conversation, and two-way interaction, particularly through self-learning mechanisms. Such systems include adaptive systems that can exhibit complex behaviour. As architecture becomes more like a machine and gains “intelligence,” it often demonstrates non-trivial and unpredictable emergent behaviour, based on relatively simple mechanical (body structure) and electronic designs, as well as interaction instructions/algorithms (mind structure). The evolution from early attempts at creating communicative systems in the 20th century within the realm of early cybernetics is traced, focusing on works such as Gordon Pask’s projects, including MusiColour and Colloquy of Mobiles (which are discussed in the article with regard to their relevance to architecture), and Walter Grey Walter’s mechanical “intelligent” creatures, to modern applications, including projects by NOX, Minimaforms, and Aranchii Architects. The primary focus is on the method (algorithm) of interaction, typically grounded in the scientific and technological advancements of its time, particularly in the field of machine learning, as well as on the value this interaction brings – a value that is largely humanistic (aimed at meeting mental needs: engaging individuals, creating new experiences, and managing uncertainty, or optimising the physical properties of the environment to improve the quality of life) and, more recently, sustainable in nature. Examples of both experimental and practical implementations around the world are analysed, and pathways for the development of interaction in architecture are proposed, involving the engagement of space (adaptation of the skin/envelope, reconfiguration of structures, and the evolution of interior elements) for the benefit of both people and the environment, including urban and city planning contexts.