Architecture and Surrealist Automatism: From Historical Models to AI Speculations

Abstract

This paper develops a theoretical framework for understanding contemporary architectural automation through the historical lens of surrealist automatism. Tracing a shift from André Breton and Philippe Soupault’s automatic writing to René Magritte’s problem-pairing and Salvador Dalí’s paranoiac-critical method, it argues that surrealist automatism progressively moved from motor execution toward perceptual,
associative, and interpretive forms of cognition. On that basis, the paper identifies three modes of architectural automatism—architecture built with automatist construction, architecture modelled on the inner world, and architecture persuasively interpreted as surreal—using them to reassess contemporary AI-assisted design. Rather than treating such systems as neutral instruments of efficiency, the paper asks whether they remain tools governed by intention or whether they reorganise design thought in ways that exceed simple instrumentality. Surrealist automatism, it argues, offers architecture a critical framework for understanding automatism/automation as an opaque, generative and speculative force.