RM Schindler’s Elmer House, 1952 (unbuilt), Siting, part 2 of 5

Siting

“Most of Schindler’s later houses were designed for hillside lots, and were magnificently tucked into the surrounding greenery. This was partly due to the preliminary sketch he drew directly on a surveyor’s map. The preliminary design was to him the vital one; …he wrote that it was “the very crux of the architect’s contribution, his main creative effort.” “ 1 (fig 1)2

Special thanks to Jocelyn Gibbs and the Architecture & Design Collection at the University of California at Santa Barbara for permission to use Schindler’s Elmer House preliminary sketch, drawn on a blueline print of the site survey.

The siting of Schindler’s houses can be seen as a series of logical decisions leading to the final design. His decisions may not be the only ones possible, but it is interesting to see how they lead to his final designs.

For the Elmer House, like most hillside designs, you have to start with the car. Large and requiring relatively flat areas, they can limit your options. This site drops down at a roughly 45 degree angle to the north, but also rises along the street to the west. For most of its width along the street, the site slopes up a few feet then drops quickly. The street and the lot meet at the southwest corner, where the downhill slope is a little flatter. This is the most logical place to drive onto the lot.(fig 2, middle right side) Decision 1, take access from the southwest corner.

There is no level spot on the lot. You need a level (or almost level) area to park a car. So any garage (or carport) will need a floor that is at least partially above the slope. Cars are heavy and require strong (and expensive) supports if they are built above the ground. You want to keep the carport and drive as close to the access point, and as much on the dirt, as possible. There are two ways to place the carport. (fig 3) The first is the standard straight-in location (red). If placed right on the property line, this keeps the driveway to a minimum and locates the carport close to the slope. A second not-so-obvious option is to turn the garage at an angle to the street (green), which is Schindler’s solution. Because he can tuck the carport into the curve of the street, this location places the carport slightly closer to the slope (about 10” closer) than the straight-in option, and so slightly reduces the expensive supports. It also has the advantages of turning the carport opening, a big hole, away from the street and turning a designed wall towards the street. The triangular drive this creates does many things; it allows you to back out and then enter the narrow, curving street going forward when you leave, it creates an off-street guest parking space, and it brings part of the house out from behind the garage – creating a more pleasant and open entry sequence. (fig 4) Decision 2, turn the carport.

The next decision, where to place the house, has many different solutions, none inherently better than the others. This decision shapes the design in many ways, the designer’s intent comes in here. Some of the options include attaching the house to the garage (used in Schindler’s Wilson and Walker Houses of 1935), bridging from the garage to a free-standing tower (Janson House, 1948), extending the house sideways along the top of the slope (Lechner House, 1946) and separating the house from the carport and placing the house and terrace at a lower level (Kallis House, 1946).

For the Elmer House, Schindler chose a similar approach to the Kallis House. At the Kallis House the terrace between the carport and garage is wood, part of it is raised above the slope to the same height as the main living level. At the Elmer House, the terrace is lowered to the ground. I think that this, in part, was done to reduce the cost of the terrace and of the house. A terrace on the ground would be cheaper than one in the air. Lowering the house also reduces the height of the supports beneath it. The Elmer terrace is narrow due to the steepness of the slope. It is created by cutting the already steep ground at the back of the terrace and putting that dirt on top of the steep slope at the front, to create a narrow ledge. This might not meet current grading standards, but it is dramatic. Decision 3, separate the house from the carport, create a terrace between, lower the terrace and house to the slope. (fig 5)

The house doesn’t need to be directly behind the carport, it could slide along the edge of the terrace. (fig 6) However, if the house is moved away from the carport, the two shapes become separated and they don’t combine into that one large, powerful form. (fig 7) The initial sketches show the east (terrace) sides of the house and carport aligned. (fig 1) I can’t say if Schindler had the final rotated box form in mind before he started sketching, although I think he did, but he did have the idea of aligning the carport and house to combine them into one large composition. Decision 4, align the house and carport on the side facing the terrace so they can be designed as one building.

But the terrace on the east side of the house is not the best location. The house and carport cast shadows on the terrace in the afternoon, when it is most likely to be used. Couldn’t you flip everything around, putting the terrace and the big form facing it on the west side of the lot? But the driveway access point stays at the southwestern corner (decision 1) and the big form is created by lining up the carport and house (decision 4). To move the terrace to the other side and realign the house and carport you have to move the house and the carport east. To do all that and keep everything lined up, the terrace gets shorter and the expensive elevated driveway gets longer. You are working against the site –  more cost and less terrace. (fig 8 ) Decision 5, terrace is on the east side of the house.

The house is lowered to terrace level, and only touches the ground along the back edge of the terrace. To get to the terrace from the living room, you have to use a bridge next to the kitchen. Although the house is small, the site is so steep that it drops 23’ by the time it gets to the back of the living room. How should the house sit on this slope? Again, there are many choices. The house can extend down to the slope as a solid mass with walls, like the Wilson House.(fig 9) The house can hover above the ground on stilts, like the back of the Rose Harris House. Or the house could cantilever out over the slope and rest on a solid pier like the Tischler House. The Elmer House does none of these. Although closest in approach to the Harris House, where only part of the Harris house was supported by stilts, all of the Elmer House sits on stilts. And it hangs out over them. This mushroomed upper floor minimizes the stilts, when viewed from the terrace. (fig 10) The Elmer house appears to be floating unsupported above the slope. The playroom, taking advantage of the space between the living room floor and the slope, is so open that it does not make the lower supports appear any more solid. The living level hangs over the stilts on the sides visible from the terrace, hiding them from the terrace, but not on the less exposed west side. (fig 11) Decision 6, float the house above the slope.

The terrace to the east becomes the best place to see the entire house, undoubtedly where Shulman would have taken his picture. As Schindler often did, the design drama is aimed at the best vantage point, the terrace in this case.

Clear decisions about the site and the building lead to an apparently inevitable design. (fig 12) The design concept is shaped by and shapes the site.

It is surprising how little was changed between the concept sketch and the final design. (fig 13) The bathroom was relocated, the bedrooms adjusted, but it’s all here at the beginning.

Next: Space

Footnotes
1  Ester McCoy, Five California Architects (New York, Preager Publishers, 1975) pg 182

2  David Gebhard editor, The Architectural Drawings of RM Schindler (The Architectural Drawings Collection, University art Museum, University of California at Santa Barbara, 1993) volume 1, pg 855