This makes it clear that different subcultures need their own activities, their own environments. But subcultures not only need to be concentrated in space to allow for the concentration of the necessary activities. They also need to be concentrated so that one subculture does not dilute the next: indeed, from this point of view they not only need to be internally concentrated—but also physically separated from one another. . . .
We cut the quote short here. The rest of the original paper presents empirical evidence for the need to separate subcultures spatially, and—in this book—we consider that as part of another pattern. The argument is given, with empirical details, in subculture BOUNDARY (13).
BUILDINGS
always be obtuse angles between 80 and 180 degrees. (We say 80, because a few degrees less than a right angle makes no difference.)
7 f* r f f / $
M> -- y<j3 -^ W
The range of possible corners.
 |
| Only angles that are nearly right angles fack successfully. |
And one further word about the angles. Most often rooms will pack in such a way that angles somewhere near right angles (say between 80 and 100 degrees) make most sense. The reason, simply, is that other obtuse angles do not pack well at corners where several rooms meet. Here are the most likely typical kinds of corners:
 |
| Polygon) rough rectangle, thick curved avail, exterior curved avail. |
This means that the majority of spaces in a building must be polygons, in plan, with roughly straight walls and obtuse-angled corners. Most often they will probably be irregular, squashed, rough rectangles. Indeed, respect for the site and the subtleties of the plan will inevitably lead to slightly irregular shapes. And occasionally they may have curved walls—either if the wall is thick enough to be concave on both sides or, on an exterior wall, where there is no important social space outside.
I 9 I THE SHAPE OF INDOOR SPACE
A final point. Our experience has led us to an even stronger version of this pattern—which constrains the shape of ceilings too. Specifically, we believe that people feel uncomfortable in spaces like these:
 |
| Rooms whose ceilings can make you uncomfortable. |
We can only speculate on the possible reasons for these feelings. It seems just possible that they originate from some kind of desire for a person to be surrounded by a spherical bubble roughly related to the human axis. Room shapes which are more or less versions of this bubble are comfortable; while those which depart from it strongly are uncomfortable. Perhafs when the sface around us is too sharfly different from the imaginary social bubble around us) we do not feel quite like fersons.
 |
| The shafe of the sface bubble. |
A ceiling that is flat, vaulted in one direction or vaulted in two directions, has the necessary character. A ceiling sloping to one side does not. We must emphasize that this conjecture is not intended as an argument in favor of rigidly simple or symmetric spaces. It only speaks against those rather abnormal spaces with one-sided sloping ceilings, high apexed ceilings, weird bulges into the room, and re-entrant angles in the wall.
BUILDINGS
Therefore:
With occasional exceptions, make each indoor space or each position of a space, a rough rectangle, with roughly straight walls, near right angles in the corners, and a roughly symmetrical vault over each room.
 |
| *$♦ |
You can define the room with columns, one at each corner— columns at the corners (212); and the shape of the ceiling can be given exactly by the ceiling vault—floor and ceiling layout (210), floor-ceiling vault (219). Avoid curved walls except where they are strictly necessary—wall membranes
(218). Where occasional curved walls like bay windows do jut out into the outside, place them to help create positive outdoor spaces (106). Make the walls of each room generous and deep -THICK WALLS (197), CLOSETS BETWEEN ROOMS (198) J and
where it is appropriate, make them half-open walls (193). For the patterns on the load-bearing structure, engineering, and construction, begin with structure follows social spaces (205). . . .
ig2 WINDOWS OVERLOOKING
LIFE*
889
. . . this pattern helps to complete the earlier patterns which give each room its shape: light on two sides of every room (159), CEILING HEIGHT VARIETY ( 190) , and THE SHAPE OF indoor space (191). Once these patterns are clear, this pattern helps to place the windows rather more precisely in the walls. It defines just how many windows there should be, how far apart, and what their total area should be.
*£♦
Rooms without a view are prisons for the people who have to stay in them.
When people are in a place for any length of time they need to be able to refresh themselves by looking at a world different from the one they are in, and with enough of its own variety and life to provide refreshment.
Amos Rapoport gives written descriptions of three windowless seminar rooms at the University of California. The descriptions —by teachers and students of English who were asked to write descriptions of the rooms as part of a writing exercise—are heavily negative, even though they were not asked to be, and in many cases refer directly to the windowless, boxed-in, or isolated-from-the-world character of the rooms.
Here are two examples:
Room 56+6 is an unpleasant room in which to attend class because in it one feels detached and isolated from the rest of the world under the buzzing fluorescent lights and the high sound-proofed ceilings, amid the sinks, cabinets, and pipes, surrounded by empty space.
The large and almost empty, windowless room with its sturdy, enclosing, and barren grey walls inspired neither disgust nor liking -y one might easily have forgotten how trapped one was. (Amos Rapoport, “Some Consumer Comments on a Designed Environment,” Arena—The Architectural Association Journal, January 1967,
pp. 176-78.)
Brian Wells, studying office workers’ choice of working positions, found that 81 per cent of all subjects chose positions next to