A pattern language

There is a fundamental conflict in the nature of materials for building in industrial society.

On the one hand, an organic building requires materials which consist of hundreds of small pieces, put together, each one of them hand cut, each one shaped to be unique according to its position. On the other hand, the high cost of labor, and the ease of mass production, tend to create materials which are large, identical, not cuttable or modifiable, and not adaptable to idiosyn-cracies of plan. These “modern” materials tend to destroy the organic quality of natural buildings and, indeed, to make it impossible. In addition, modern materials tend to be flimsy and hard to maintain—so that buildings deteriorate more rapidly than in a pre-industrial society where a building can be maintained and improved for hundreds of years by patient attention.

The central problem of materials, then, is to find a collection of materials which are small in scale, easy to cut on site, easy to work on site without the aid of huge and expensive machinery, easy to vary and adapt, heavy enough to be solid, longlasting or easy to maintain, and yet easy to build, not needing specialized labor, not expensive in labor, and universally obtainable and cheap.

Furthermore, this class of good materials must be ecologically sound: biodegradable, low⁷ in energy consumption, and not based on depletable resources.

When we take all these requirements together, they suggest a

207 GOOD materials

rather startling class of “good materials”—quite different from the materials in common use today. The following discussion is our attempt to begin to define this class of materials. It is certainly incomplete; but perhaps it can help you to think through the problem of materials more carefully.

We start with what we call “bulk materials”—the materials that occur in the greatest volume in a given building. They may account for as much as 80 per cent of the total volume of materials used in a building. Traditionally, bulk materials have been earth, concrete, wood, brick, stone, snow. . . . Today the bulk materials are essentially wood and concrete and, in the very large buildings, steel.

When we analyze these materials strictly, according to our criteria, we find that stone and brick meet most of the requirements, but are often out of the question where labor is expensive, because they are labor intensive.

Wood is excellent in many ways. Where it is available people use it in great quantities, and where it is not available people are trying to get hold of it. Unfortunately the forests have been terribly managed; many have been devastated; and the price of heavy lumber has skyrocketed. From today’s paper; “Since the end of federal economic controls the price of lumber has been jumping about 15 percent a month and is now about 55 percent above what it was a year ago.” San Francisco Chronicle,, February 11, 1973. We shall therefore look upon wood as a precious material, which should not be used as a bulk material or for structural purposes.

Steel as a bulk material seems out of the question. We do not need it for high buildings since they do not make social sense— four-story jlimit (21). And for smaller buildings it is expensive, impossible to modify, high energy in production.

Earth is an interesting bulk material. But it is hard to stabilize, and it makes incredibly heavy walls because it has to be so thick. Where this is appropriate, and where the earth is available, however, it is certainly one of the “good materials.”

Regular concrete is too dense. It is heavy and hard to work. After it sets one cannot cut into it, or nail into it. And its surface is ugly, cold, and hard in feeling, unless covered by expensive finishes not integral to the structure.

CONSTRUCTION

And yet concrete, in some form, is a fascinating material. It is fluid, strong, and relatively cheap. It is available in almost every part of the world. A University of California professor of engineering sciences, P. Kumar Mehta, has even just recently found a way of converting abandoned rice husks into Portland cement.

Is there any way of combining all these good qualities of concrete and also having a material which is light in weight, easy to work, with a pleasant finish!* There is. It is -possible to use a whole range of ultra-lightweight concretes which have a density and compressive strength very similar to that of wood. They are easy to work with, can be nailed with ordinary nails, cut with a saw, drilled with wood-working tools, easily repaired.

We believe that ultra-lightweight concrete is one of the most fundamental bulk materials of the future.

To make this as clear as possible, we shall now discuss the range of lightweight concretes. Our experiments lead us to believe that the best lightweight concretes, the ones most useful for building, are those whose densities lie in the range of 40 to 60 pounds per cubic foot and which develop some 600 to 1000 psi in compression.

Oddly enough, this particular specification lies in the least developed part of the presently available range of concretes. As we can see from the following diagram, the so-called “structural” concretes are usually more dense (at least 90 pounds per cubic foot) and much stronger. The most common “lightweight” concretes use vermiculite as an aggregate, are used for underflooring and insulation, and are very light, but they do not usually develop enough strength to be structurally useful—most

Currently available concrete mixes.

207 good materials

often about 300 psi in compression. However, a range of mixed lightweight aggregates, containing vermiculite, perlite, pumice, and expanded shale in different proportions, can easily generate 40-60 pound, 600 psi concretes anywhere in the world. We have had very good luck with a mix of 1-2-3: cement-kylite-vermicu-lite.

Beyond the bulk materials, there are the materials used in relatively smaller quantities for framework, surfaces, and finishes. These are the “secondary’’ materials.

When buildings are built with manageable secondary materials, they can be repaired with the same materials: repair becomes continuous with the original building. And the buildings are more apt to be repaired if it is easy to do so and if the user can do it himself bit by bit without having to rely on skilled workers or special equipment. With prefabricated materials this is impossible, the materials are inherently unrepairable. When prefabricated finish materials are damaged they must be replaced with an entirely new component.

Take the case of a garden patio. It can be made as a continuous concrete slab. When the ground shifts slightly underneath this slab, the slab cracks and buckles. This is quite unrepairable for the user. It requires that the entire slab be broken out (which requires relatively heavy-duty equipment) and replaced—by professional skilled labor. On the other hand, it would have been possible to build the patio initially out of many small bricks, tiles, or stones. When the ground shifts, the user is then able to lift up the broken tiles, add some more earth, and replace the tile—all without the aid of expensive machinery or professional help. And if one of the tiles or bricks becomes damaged, it can be easily replaced.

What are the good secondary materials? Wood, which we want to avoid as a bulk material, is excellent as a secondary material for doors, finishes, windows, furniture. Plywood, particle board, and gypsum board can all be cut, nailed, trimmed, and are relatively cheap. Bamboo, thatch, plaster, paper, corrugated metals, chicken wire, canvas, cloth, vinyl, rope, slate, fiberglass, non-chlorinated plastics are all examples of secondary materials which do rather well against our criteria. Some are dubious ecologically —that is, the fiberglass and the corrugated metals—but again,

CONSTRUCTION

these sheet materials need only be used in moderation, to form and finish and trim the bulk materials.

Finally, there are some materials which our criteria exclude entirely—either as bulk or secondary materials. They are expensive, hard to adapt to idiosyncratic plans, they require high energy production techniques, they are in limited reserves. . . . for example: steel panels and rolled steel sections; aluminum; hard and prestressed concrete; chlorinated foams; structural lumber; cement plaster; immense sections of plate glass. . . .

And, for any optimist who thinks he can go on using steel reinforcing bars forever—consider the following fact. Even iron, abundant as it is all over the earth’s surface, is a depletable resource. If consumption keeps growing at its present rate of increase (as it very well may, given the vast parts of the world not yet using resources at American and western consumption levels), the resources of iron will run out in 2050.

Years at which various metals will be depleted assuming current usage rate continues to increase as it did betweeni960 and 1968.

Therefore:

Use only biodegradable, low energy consuming materials, which are easy to cut and modify on site. For bulk materials we suggest ultra-lightweight 40-60 lbs. concrete and earth-based materials like tamped earth, brick, and tile. For secondary materials, use wood planks, gypsum, plywood, cloth, chickenwire, paper, cardboard, particle board, corrugated iron, lime plasters, bamboo, rope, and tile.

GOOD MATERIALS
tile& adobo 1" plonks

ultra-light weight concrete or organic or earth-based materials

In gradual stiffening (208), we shall work out the way of using these materials that goes with structure follows social spaces (205) and efficient structure (206). Try to use the materials in such a way as to allow their own texture to show themselves—lapped outside walls (234), soft inside walls (235). . . .

961

208 gradual stiffening**

962

. . . in STRUCTURE FOLLOWS SOCIAL SPACES (2O5) and EFFICIENT structure (206) we have set down the beginnings of a philosophy, an approach, to construction, good materials (207) tells us something about the materials we ought to use in order to meet human and ecological demands. Now, before we start the practical task of making a structural layout for a building, it is necessary to consider one more philosophical pattern: one which defines the process of construction that will make it possible to use the right materials and get the overall conception of the structure right.

The fundamental philosophy behind the use of pattern languages is that buildings should be uniquely adapted to individual needs and sites; and that the plans of buildings should be rather loose and fluid, in order to accommodate these subtleties.

This requires an entirely new attitude toward the process of construction. We may define this attitude by saying that it is desirable to build a building in such a way that it starts out loose and flimsy while final adaptations in plan are made, and then gets stiffened gradually during the process of construction, so that each additional act of construction makes the structure sounder.

To understand this philosophy properly, it is helpful to imagine a building being made like a basket. A few strands are put in place. They are very flimsy. Other strands are woven in. Gradually the basket gets stiffer and stiffen Its final structural strength is only reached from the cooperation of all the members, and is not reached until the building is completely finished. In this sense, such a process produces a building in which all parts of it are working structurally—see efficient structure (206).

Why does the principle of gradual stiffening seem so sensible as a -process of building?

To begin with, such a structure allows the actual building process to be a creative act. It allows the building to be built up

CONSTRUCTION

gradually. Members can be moved around before they are firmly in place. All those detailed design decisions which can never be worked out in advance on paper, can be made during the building process. And it allows you to see the space in three dimensions as a whole, each step of the way, as more material is added.

This means that since each new material that is added in the process must adapt perfectly to the framework that is there, each new material must be more adaptable, more flexible, more capable of coping with variation, than the last. Thus, though the building as a whole goes from flimsy to strong, the actual materials that are added go from the strongest and stiffest, to the gradually less stiff, until finally fluid materials are added.

The essence of this process is very fundamental indeed. We may understand it best by comparing the work of a fifty-year-old carpenter with the work of a novice. The experienced carpenter keeps going. He doesn’t have to keep stopping, because every action he performs, is calculated in such a way that some later action can put it right to the extent that it is imperfect now. What is critical here, is the sequence of events. The carpenter never takes a step which he cannot correct later; so he can keep working, confidently, steadily.

The novice, by comparison, spends a great deal of his time trying to figure out what to do. He does this essentially because he knows that an action he takes now may cause unretractable problems a little further down the line; and if he is not careful, he will find himself with a joint that requires the shortening of some crucial member—at a stage when it is too late to shorten that member. The fear of these kinds of mistakes forces him to spend hours trying to figure ahead: and it forces him to work as far as possible to exact drawings because they will guarantee that he avoids these kinds of mistakes.

The difference between the novice and the master is simply that the novice has not learnt, yet, how to do things in such a way that he can afford to make small mistakes. The master knows that the sequence of his actions will always allow him to cover his mistakes a little further down the line. It is this simple but essential knowledge which gives the work of a master carpenter its wonderful, smooth, relaxed, and almost unconcerned simplicity.

In a building we have exactly the same problem, only greatly magnified. Essentially, most modern construction has the character of the novice’s work, not of the master’s. The builders do not know how to be relaxed, how to deal with earlier mistakes by later detailing; they do not know the proper sequence of events; and they do not, usually, have a building system, or a construction process, which allows them to develop this kind of relaxed and casual wisdom. Instead, like the novice, they work exactly to finely detailed drawings; the building is extremely uptight as it gets made; any departure from the exact drawings is liable to cause severe problems, may perhaps make it necessary to pull out whole sections of the work.

This novice-like and panic-stricken attention to detail has two very serious results. First, like the novice, the architects spend a great deal of time trying to w'ork things out ahead of time, not smoothly building. Obviously, this costs money; and helps create these machine-like “perfect” buildings. Second, a vastly more serious consequence: the details control the whole. The beauty and subtlety of the plan in which patterns have held free sway over the design suddenly becomes tightened and destroyed because, in fear that details won’t work out, the details of connections, and components, are allowed to control the plan. As a result, rooms get to be slightly the wrong shape, windows go out of position, spaces between doors and walls get altered just enough to make them useless. In a word, the whole character of modern architecture, namely the control of larger space by piddling details of construction, takes over.

What is needed is the opposite—a process in which details are fitted to the whole. This is the secret of the master carpenter; it is described in detail in The Timeless Way of Building as the foundation of all organic form and all successful building. The process of gradual stiffening, which we describe here, is the physical and procedural embodiment of this essential principle. We now ask how, in practice, it is possible to create a gradually stiffened structure within the context defined by the pattern GOOD MATERIALS (2O7).

Facts about materials give us the starting point we need.

1. Sheet materials are easy to produce and make the best connections.

965

❖ 4* *J«

The scattered work itself can take a great variety of forms. It can occur in belts of industry, where it is essential for an industry to occupy an acre or more between subcultures—subculture boundary (13), industrial ribbon (42); it can occur in work communities, which are scattered among the neighborhoods— NEIGHBORHOOD BOUNDARY (15), WORK COMMUNITY (41); and it can occur in individual workshops, right among the houses— home workshop (I 57) - The size of each workplace is limited only by the nature of human groups and the process of self-governance. It is discussed in detail in self-governing workshops AND OFFICES (80). . . .

CONSTRUCTION

In traditional society there are few sheet materials. However, factory production tends to make sheets more easily than other forms of material. As we move into an age of mass production, sheet materials become plentiful and are naturally strong, light, and cheap. Gypsum board, plywood, cloth, vinyl, canvas, fiberglass, particle board, wood planks, corrugated metals, chicken wire, are all examples.

And sheet materials are the strongest for connections. Connections are the weak points in a structure. Sheet materials are easy to connect, because connections can join surfaces to one another. Anything made out of sheets is inherently stronger than something made of lumps or sticks.

2. Ultra-lightweight concrete is an excellent fill material— it has the density of wood, is strongs light, easy to cut, easy to refair, easy to nail into—and is available everywhere. This is discussed fully in good materials (207).

3. However, a?iy kind of concrete needs formwork; and the cost of formwork is e7LOrmous.

This makes it very expensive indeed to build any complex form; and within conventional building systems, it more or less rules out the kind of “organic” structure which we have described. Furthermore, in regular concrete work, the formwork is eventually wasted, thrown away.

We believe that the finishes in any sensible building system should be integral with the process of construction and the structure itself (as they are in almost all traditional buildings) — and that any building system in which finishes have to be “added” to the building are wasteful, and unnatural.

4. We therefore frofose that ultra-lightweight concrete be foured into forms which are made of the easily available sheet materials: and that these materials are the?i left in flace to form the fnish.

The sheet materials can be any combination of cloth, canvas, wood planks, gypsum boards, fiberboards, plywood, paper, plastered chickenwire, corrugated metals, and where it is possible, tile, brick, or stone—see good materials (207). For the ultra-lightweight concrete we recommend a perlite, expanded shale, or pumice aggregate. Tamped earth, adobe, nonchlorinated foams, may also do instead of the concrete, if loads allow it.

2,oB

_graD^uA

_h STIFFENING

-^v * • * -

One ‘version of gradual stiffening, tising one inc/i -planks, gypsum board and burlap as sheets, anith ultra-lig/it'weight

concrete as fill.

"The drawing above, shows one particular realisation of this kind of gradual stiffening. But the principle is far more general than this particular use of it. Indeed, it occurs, in one way or another, in almost all traditional forms of building. Eskimo igloo construction and African basket structures are both gradually stiffened structures, where each next step copes with the existing framework, adds to it, and stiffens it. The stone buildings of Alberobello in southern Italy are examples. So is Elizabethan half-timber construction.

967

CONSTRUCTION

Therefore:

Recognize that you are not assembling a building from components like an erector set, but that you are instead weaving a structure which starts out globally complete, but flimsy; then gradually making it stiffer but still rather flimsy; and only finally making it completely stiff and strong.

We believe that in our own time, the most natural version of this process is to put up a shell of sheet materials, and then make it fully strong by filling it with a compressive

fill.

soft skin formwork compressive fill

Choose the most natural materials you can, for the outer shell itself—thin wood planks for columns, canvas or burlap for the vaults, plaster board or plank or bricks or hollow tiles for walls— GOOD MATERIALS (2O7),

Use ultra-lightweight 40 to 60 pounds perlite concrete for the compressive fill—it has the same density as wood and can be cut and nailed like wood, both during the construction and in later years when repairs become necessary—good materials (207).

Build up the columns first, then fill them with the ultralightweight concrete; then build up the beams and fill them; then the vaults, and cover them with a thin coat of concrete which hardens to form a shell; then fill that shell with even lighter weight materials to form the floors; then make the walls and window frames, and fill them; and finally, the roof, again a thin cloth vault covered with a coat of concrete to form a shell— BOX COLUMNS (2 I 6), PERIMETER BEAM (2 I 7), WALL MEMBRANE (218), FLOOR-CEILING VAULTS (2I9), ROOF VAULTS (220). . . .

within this fhilosofhy of structure₎ on the basis of the flans which you have made_y work out the comflete structural layout; this is the last thing you do on fafer_} before you actually start to build;

209. ROOF LAYOUT

210. FLOOR AND CEILING LAYOUT

21 I. THICKENING THE OUTER WALLS

212. COLUMNS AT THE CORNERS

213. FINAL COLUMN DISTRIBUTION

209 ROOF LAYOUT*

970

. . . assume now that you have a rough plan, to scale, for each floor of the building. In this case you already know roughly how the roofs will go, from cascade of roofs ( i i 6) and sheltering roof (117) ; and you know exactly where the roof is flat to form roof gardens next to rooms at different floors—roof garden (11 8). This pattern shows you how to get a detailed roof plan for the building, which helps those patterns come to life, for any plan which you have drawn.

V

What kind of roof plan is organically related to the nature of your building?

We know, from arguments presented in the shape of indoor space (191), that the majority of spaces in an organic building will have roughly—not necessarily perfectly—straight walls because it is only then that the space on both sides of the walls can be positive, or convex in shape.

And we know, from similar arguments, that the majority of the angles in the building will be roughly—again, not exactly— right angles, that is, in the general range of 80 to 100 degrees.

We know, therefore, that the class of natural plans may contain a variety of shapes like half circles, octagons, and so on—but that for the most part, it will be made of very rough, sloppy rectangles.

We also know, from sheltering roof (117), that entire wings should be under one roof whenever possible and that the building is to be roofed with a mixture of flat roofs and sloping or domical roofs, with the accent on those which are not flat.

We may therefore state the problem of defining a roof layout as follows: Given an arbitrary flan of the tyfe described above, how can we fit to it an arrangement of roofs which co?iforms to the cascade of roofs (r/d) and sheltering roof (117) and

ROOF GARDENS (Il8)?

Before explaining the procedure for laying out roofs in detail, we underline five assumptions which provide the basis for the procedure.

CONSTRUCTION

I. The “pitched” roofs may actually be pitched, or they may be vaults with a curved pitch, or barrel vaults—as described in roof vaults (220). The general procedure, in all three cases, is the same. (For curved vaults, define slope as height-to-width ratio.)

The “fitch” of a vaulted roof.

2. Assume that all roofs in the building, which are not flat, have roughly the same slope. For a given climate and roof construction, one slope is usually best; and this greatly simplifies construction.

The same slope throughout.

3. Since all roofs have the same slope, the roofs which cover tire widest wings and/or rooms will have the highest peaks; those covering smaller wings and rooms will be relatively lower. This is consistent with main building (99), cascade of roofs

(1 16), and CEILING HEIGHT VARIETY (190).

Wide roofs are highest.

4. Any place where the building helps to enclose an outdoor room or courtyard needs an even eave line so that it has the space of a “room.” An irregular roof line, with gable ends, will usually destroy the space of a small courtyard. It is necessary, therefore, that roofs be hipped in these positions to make the roof edge horizontal.

5. In all other positions, leave the ends of buildings and wings as gable ends.
973

CONSTRUCTION

We shall now discuss the rules for roofing a building by using an example of a house designed by a layman using the pattern language. This building plan is shown below. It is a single-story house and it contains no roof gardens or balconies.

We first identify the largest rectangular cluster of rooms and roof it with a peaked roof, the ridge line of which runs the long direction:

Then we do the same with smaller clusters, until all the major spaces are roofed.

Then we roof remaining small rooms, alcoves, and thick walls with shed roofs sloping outward. These roofs should spring from the base of the main roofs to help relieve them of outward thrusts; their outside walls should be as low as possible.

209 ROOF layout

Finally, we identify the outdoor spaces (shown as A, B, and C), and hip the roofs around them to preserve a more continuous eave line around the spaces.

We shall now discuss a slightly more complicated example, a two story building.

IF-

TIT

We begin with the top story, roofing the entire master bedroom and bath under one peaked roof with the ridge running lengthwise:

Next we move to the lower story, roofing the children’s wing under a flat roof to form a roof garden ( i i 8) for the master bedroom, and the larger living room under a pitched roof, again with the ridge running lengthwise.

10 MAGIC OF THE CITY

58

CONSTRUCTION

Then we bring the roof over the master bedroom down over the interior loft.

Finally, we smooth the living room roof ridge line into the side of the roof over the loft. This completes the roof layout.

It is very helpful, when you are laying out roofs, to remember the structural principle outlined in cascade of roofs ( i i6). When you have finished, the overall arrangement of the roofs should form a self-buttressing cascade in which each lower roof helps to take up the horizonal thrust generated by the higher roofs—and the overall section of the roofs, taken in very very general terms, tends toward a rough upside down catenary.

Therefore:

Arrange the roofs so that each distinct roof corresponds to an identifiable social entity in the building or building complex. Place the largest roofs—those which are highest and have the largest span—over the largest and most important and most communal spaces; build the lesser roofs off these largest and highest roofs; and build the smallest roofs of all off these lesser roofs, in the form of half-vaults and sheds over alcoves and thick walls.

major roofs

*J* *F

You can build all these roofs, and the connections between them, by following the instructions for roof vaults—roof vaults (220). When a wing ends in the open, leave tire gable end at full height; when a wing ends in a courtyard, hip the gable, so that the horizontal roof edge makes the courtyard like a room— COURTYARDS WHICH LIVE ( I I 5) .

Treat the smallest shed roofs, which cover thick walls and alcoves, as buttresses, and build them to help take the horizontal thrust from floor vaults and higher roof vaults—thickening the OUTER WALLS ( 2 I I ) . . . .