Moving through Caillois and away from a rational understanding of games and play, we can get a better understanding of the social instability that the Pruitt-Igoe projects exhibited.
The social vertigo is a major point in most discussions of PI; in areas of 20 or more families, residents had difficulty distinguishing fellow residents from intruders--cooperators from potential defectors.
Such space itself is disorienting, but in a specific way. Turlington Hall, with its multiple entrances, twists, turns, double-backs, and no windows, is a complex and disorienting space because of its lack of repetition; PI is disorienting precisely because of its total repetition.
The quantification that invariably accompanies grids, and the process of gridding, rationally compensates for this--we know that number 312 will be on the third floor, and between 311 and 313.
Now, lets imagine a grid without numbers--a grid without the very thing that makes it rationally comprehensible. The space would have to take on qualitative features as a means of facilitating navigation.
A Turlington without numbers: navigation would involve spatial direction and identifiable landmarks. "Go in through the north stairwell to the top floor, turn right through the double doors, look for the door with William Blake and Donald Duck posters." Relatively simple in a space that is qualitatively heterogenous. But how would one navigate extremely homogenized space like PI without recourse to simply counting off floors, doors, and corridors?