Zixuan Sharon Deng

MArch I

On Vision: The Science and Cultural History of Spatial Perception and Imagination 

Advised by Antoine Picon and Andrew Witt 

This thesis investigates the relationship between evolving modes of visual perception and the space of architecture. The design of architecture is intimately linked to the way space is imagined to be seen, and thus to the scientific and cultural understanding of vision. And yet, the scopic regime of retinal projection as embodied by the camera obscura has dominated architectural imagination, even as empirical understanding of vision drastically changed.  

The monocular observer is accurate only when the observed object is far enough that the two eyes’ horizontal separation—parallax—can be neglected. Whereas the space of retinal projection is single, continuous, and homogenous, the space of binocular vision is split, ambiguous, and haptic. A more recent but definitive shift in the understanding of human vision came with the discovery of orientation-selective cells in the primary visual cortex of cats. The algorithmically assembled image-space is pixelated, fragmented, and noisy. In contrast to the model of retinal projection, which assumes that we passively perceive objects through their image in the retina, vision has come to be understood as the active extraction of relevant information in hierarchical stages: from edges to textures to patterns to objects. 

How architectural space can engage with each and all these spatial schemas is then, briefly, the central question underlying my project. I think it is urgent and necessary for architects to understand and engage with the biology of seeing. Only then can the design of space become relevant, just, and humane. This project is conceived as an effort in that direction. 

Stereoscopic photograph of two pointed Archways on yellow background.