A Place of My Own: The Education of an Amateur Builder

by Michael Pollan

  The shingles themselves, colored an earthy red shading to tobacco, surprised me with their inconsequentiality. At the business end they were less than half an inch thick, and they dwindled to paper at the other. It was nothing to break one of them in half. Yet layered and woven with enough care, these aromatic slips of cedar made a sturdy shelter, could withstand even a New England nor’easter. It wasn’t until I’d handled a few hundred of them that I fully appreciated the design of Charlie’s roof, the way he’d underscored with his fat rafters and fine lath the delicate weave of wood that a shingle roof is. Building it, you knew this was a roof designed by someone who’d probably once done some roof work himself, someone who had given a lot of thought to what a cedar shingle is.

  The architect Louis Kahn used to talk about interrogating his materials in order to learn what they “wanted to be”—that is, what the distinctive nature of a material suggested should be done with it:

  You say to brick, “What do you want, brick?” Brick says to you, “I like an arch.” If you say to brick, “Arches are expensive, and I can use a concrete lintel over an opening. What do you think of that, brick?” Brick says, “I like an arch.”

  The stuff we make our buildings out of—our bricks and cedar shingles, our concrete and stucco and even plastic—is perhaps the first way that nature expresses itself in our architecture. Working attentively with their materials can draw the architect and builder into a kind of dialogue with the material world; you learn a lot about a shingle—and about red cedar—watching how it responds to your handling. Of course the architect doesn’t have to honor his materials in the way Kahn describes; he’s free in designing to interrogate a philosophical conceit rather than a brick or a shingle, or to strive for a “dematerialized” look in his surfaces. But he probably runs certain risks in doing so. Sooner or later a stain will expose the materiality of his stucco or plastic, if the rain doesn’t undermine it first. Try as it might, no building ever transcends the stuff it’s made of.

  Materials are so essential to our physical experience of a place that to disregard them—to ignore the coldness of steel, the dumb strength of concrete, the sympathy of wood, whose temperature never startles—is to throw away a great deal of architecture’s expressive power. Is the nature of that power linguistic? Certainly my shingles signified specific things to my mind (“New England,” for example), but they also addressed my senses more directly, with their aroma, their delicateness, with the impression they gave my hands of wanting to be layered and woven for strength. It seemed to me you wouldn’t seriously argue that architecture was a language unless you’d forgotten the specific heft of a cool brick, or the smell of fresh cedar warmed in the afternoon sun. The reality and presence and Hereness of these things, the sense they give us of wanting to do one thing and not another, exert a worldly pressure on building that an architect would have to go out of his way to ignore. As for a builder, he wouldn’t even think to try.

  While we shingled, Joe and I talked, mainly about “ice dams.” Ice dams are probably the most serious threat facing a roof in the northern latitudes. It seems that when snow on a roof is melted by the heat radiating upward from the interior of a building, meltwater streams down the slope until it reaches the much-colder eaves, where it’s apt to freeze again, building up in a heavy block along the lower edge of the roof. This is why Charlie had spec’d two-by-six planking under the first three feet of shingles, instead of the relatively light straps we were using higher up.

  But weight is apparently not the only danger an ice dam presents. A thick one will block the flow of meltwater in the spring and actually force it to back up the slope of the roof and then under the shingles, where it is liable to infiltrate the building and drip onto one’s head. One reason roofs are steeper in snowier climates is to prevent ice dams, since a steep roof will shed precipitation more quickly than a shallow one; the steeper a roof’s pitch, the less likely it is meltwater will travel back up its slope. Perched near the peak of my own roof, holding on to the slicked straps for dear life, I had the testimony of my own body to these facts, as the muscles in my legs registered the slope’s distinct wish to shed my weight.

  The easiest way for an architect to avoid problems with his roof is to pay more attention to vernacular practice. Vernacular roofs—most of which happen to be pitched, the precise pitch varying with the latitude and local snowfall—reflect the experience of thousands of builders over hundreds of years; they represent a successful adaptation to a given environment, a good fit between the human desire to keep dry and the predictable behavior of water and wood under specific circumstances. There’s nothing inherently wrong with attempting something more adventurous, but, as in the case of an evolutionary mutation, there’s a greater risk that the novel design will fail. When Frank Lloyd Wright declared that “If the roof doesn’t leak the architect hasn’t been creative enough,” he had a point. There seems to be an inherent tension between architectural novelty and sound construction.*

  The vernacular roof suggests another way that nature finds its way into architecture, sanctioning one solution and rendering others suspect. After long trial and error, builders discovered that the pitched roof worked best at keeping the rain out of buildings built in wood. So it means something to say that, under certain circumstances, a pitched roof is more “natural” than some other kind—that is, more in keeping with the way this world we’ve been given seems to work. And we can say this without having to say anything categorical about “nature.” All we’re saying is that, whatever nature really is, it seems to behave in one way in the case of a pitched roof, and another in the case of a flat one. We may not know nature directly, as the deconstructors never tire of reminding us, but we do have long experience of what works and what doesn’t work in nature. My own hard-won experience of right angles, for example, has convinced me that, whatever the deconstructivists might think, ours is indeed a ninety-degree world. Lloyd Kahn, once a leading advocate of dome-shaped houses, came to a similar conclusion after actually building and living in one:

  What’s good about 90-degree walls: they don’t catch dust, rain doesn’t sit on them, easy to add to; gravity, not tension, holds them in place. It’s easy to build in counters, shelves, arrange furniture, bathtubs, beds. We are 90 degrees to the earth.

  The subject of leaky roofs also suggested to me that there might be certain architectural conventions that “mean” in a less arbitrary way than signs do. Geographers tell us they can infer the climate of a region from the steepness of the roofs found there; the steeper the slope of the typical roof, the more snow the region receives. The pitch of a vernacular roof may be conventional, but it is not arbitrary; it represents something more than caprice or fashion or a “social construction.” Put another way, the roof derives at least one of its meanings—the one about climate—not from the agreement of a group of people in architecture or the field of geography, but from certain facts of nature. Its form is less like a combination of letters in a language than a body part or camouflage device that, far from being arbitrary, exhibits a specific fitness to its environment.

  Even Robert Venturi and Peter Eisenman would grant this much to nature: Architecture’s roofs should not leak. Oh yes, and one other thing: gravity—architecture has got to stand up too. Eisenman allows that a building must work as structure (it should stand up) and shelter (it should stay dry)—though he insists it needn’t look like it stands up and stays dry. But after that, anything goes. Venturi adds that, since anyone can make a shed stand up and stay dry, the really good minds should occupy themselves with the signs and ornamentation.

  And this is how big-time architecture is practiced today, at least by stars like Eisenman and Venturi. Read the credits on important new buildings and you will invariably find two architectural firms listed: one you’ve heard of, and another you probably haven’t. For example, on the Wexner Center for the Arts in Columbus, Ohio, perhaps Peter Eisenman’s most famous actually constructed building, he shares credit
with an obscure local firm by the name of Richard Trott & Partners. The famous architect’s firm will give a building its signature look—deconstructivist, postmodern, whatever—and then a second, unheralded firm is called in to flesh out the design in such a way as to make sure the building will stand up and stay dry and pass muster with the building inspector. In effect, the whole leaky roof problem has been subbed out.

  As this division of architectural labor suggests, the work of construction and the work of design have drifted some distance apart, and you can’t help but wonder (especially when you’re groping for a reliable foothold on a slicked roof) whether this gulf might not help explain the increasingly abstract and literary quality of so much contemporary architecture, not to mention a great many leaky roofs. The history of architecture is the history of the widening of that gulf, from the time when master builders designed and built buildings themselves; to the Renaissance, when architects began designing buildings but left decisions about construction and ornament to craftsmen on site; to our own time, when celebrated architects concentrate on the skin of the building, the details of construction fall to local engineering and design firms, and the craftsman, the one with his hands on the thing itself, has been reduced to an unconsulted laborer. It stands to reason that the greater an architect’s distance from the actual work of making buildings, the more likely he would be to embrace what Venturi has called an “architecture of communication over space.”

  By now, a tendency to emphasize signs at the expense of space, or physical experience, is probably built into the way contemporary architecture gets practiced and judged. The architect is bound to stress There in favor of Here when There is where the architect works. The arena in which a great deal of the work of architecture is performed today is on paper: in the articles and photographs used to disseminate and comment upon buildings and chart the rise and fall of architects’ careers. Building buildings is no longer even a prerequisite to a successful architectural career, as Peter Eisenman, John Hejduk, Robert Venturi, and a great many other current and former paper architects can testify. (Eisenman is probably right to suggest, as he once did, that the actual building of House VI was all but incidental to the project.) In the case of buildings that do get built, since it is very often not the client’s but the media’s opinion that really matters, architects will naturally tend to emphasize those elements in their designs that can be communicated effectively in the relevant media, and these are inevitably going to have more to do with two-dimensional signs than with three-dimensional space, more with images and information than with the tactile qualities of materials and the experience of space. This kind of work has acquired a name: “magazine architecture.” Of course, it never rains in magazines.

  Rain and gravity: Are these really the only facts of nature architecture has to worry about? Is structure and shelter as far as the architect’s obligation to reality goes? For a long time it seemed to me that this might in fact be the case; that Venturi and Eisenman had driven nature into a very tight corner indeed, and, after taking account of those two irreducible basics they could safely ignore it. The rest was all culture and fashion and taste: Anything goes. I couldn’t see any way out of this tight corner—until, that is, I chanced to talk to Charlie about a somewhat unusual commission he was working on the summer I raised my roof.

  It was, of all things, a design for a birdhouse, and what he told me about it made me wonder if the place of nature in architecture might not be more extensive and subtle than a postmodernist would think. Specifically, Charlie was designing a wood duck house for a man who had built a pond to which he was eager to attract wildlife. The wood duck is a threatened species that is apparently quite choosy about its nesting sites. Charlie started from the assumption that his client was in some sense a duck, even though he understood he had to design a structure that would please the eye of his human (and fee-paying) client as well. So he spent a couple of afternoons at the library of the comparative zoology department at Harvard, learning all he could about the needs and nesting habits of Aix sponsa.

  To succeed, Charlie’s little building would not only have to stand up and shed water (the postmodern shed’s bottom line), but must exhibit a whole series of other characteristics necessary to win the attention and ensure the comfort of wood ducks—characteristics that don’t fit under the rubric of ornament; no “decorated shed” was likely to do the trick here. The entrance, for example, had to be four inches in diameter, an aperture large enough to admit the female but not the male; this is an arrangement a nesting wood duck evidently insists upon. The opening should be several inches deep as well, to prevent a raccoon from reaching in to snatch the eggs. Beyond that, the interior needed to be a well-ventilated vertical space dropping down below the entrance tunnel. Lastly, the house had to be sited either directly over a pond or no more than a few feet from the shore, so that the mother could conduct her ducklings to the safety of open water soon after their emergence from the nest.

  The basic idea, as Charlie explained it to me, was to recreate the characteristics of a fairly large woodpecker hole in a dead, hollowed-out tree near a pond or in a swamp—the wood duck’s natural habitat. Charlie was free to design the building to look any way he wanted—vernacular, postmodern, deconstructivist, whatever—but in a few key respects it had better remind a wood duck of a woodpecker hole in a tree or no wood duck would ever come near it. What struck me as significant about this was that Charlie was attempting not to fool the wood duck, who would understand perfectly well that this gabled house on stilts (it wound up looking a lot like a Charlie Myer house) was neither a tree nor a woodpecker hole, but to somehow evoke those things. In a sense, Charlie’s wood duck house was an acknowledged piece of artifice designed to symbolize the wood duck’s natural habitat; as one thing that referred to another, you might say it was a kind of duck metaphor.

  I know; I’m talking about ducks. Yet Charlie’s wood duck house made me appreciate that, even to a duck, the landscape brims with meaning. Certain formations in it imply certain qualities: To a duck, a deep hole set high over water connotes safety and convenience. This suggests a couple of things that seemed at least potentially relevant to human architecture. Meaning is not always a function of language or even communication; to wood ducks at least (who by the way can also communicate among themselves in the usual manner, by quacking), the things of this world are not mute but sometimes speak to a creature directly, carrying meanings of shelter, of danger, of nourishment, of sexual opportunity—all meanings that don’t depend on a sign system or culture of any kind. The meaning of a four-inch hole set high over water is the product not of an agreement among wood ducks—of cultural consensus—but of the species’s evolution. It came into the world whenever it was that wood ducks first figured out that, given the shape and size of a wood duck body and certain facts about the species’s reproduction, this particular formation denoted a superior shelter; in the case of this species, “symbolism”—perhaps even in some sense “taste”—is a by-product of survival: of what works.

  And yet there’s no denying the existence of countless symbols and conventions that are entirely arbitrary and cultural. Even Charlie’s wood duck house featured symbols that almost certainly meant nothing to a wood duck, that were strictly part of a system of signs, a language you had to learn. There were a series of details, for example, that signified a human home: the gable roof, a trio of tiny windows along each side, and some ornament around the entrance that heightened the sense of ceremony there. These things were obviously directed not at ducks but at people.

  What this suggests is that very different orders of symbolism can coexist in a building. Some symbols are patently just as arbitrary as the postmodernists say. How else to account for the fact that, in the first half of the nineteenth century, great white fluted columns on the front of an American house symbolized republican virtue in one part of the country and a slave-holding aristocracy in another? Had Charlie put fluted white columns on the façade of
his duck house, they would have been nothing more than a sign, as meaningless as ng to a duck and, for that matter, to anybody else not versed in that particular human cultural system. So how could you have it both ways: fluted columns that were wholly arbitrary and four-inch holes (or, closer to home, pitched roofs) that were clearly fitted to the facts of nature? The birdhouse suggested a simple hypothesis: Maybe architecture speaks in more than one voice, the first grounded in meanings at least partly given by nature and another trafficking in meanings determined mainly by culture.

  Soon after formulating this hypothesis, I found some human backing for it right in my own human building. Joe and I had finished shingling our roof, capping it at the peak with two well-caulked, -glued, and -screwed-together cedar ridge boards, and we’d turn our attention to closing in the rest of the building. We nailed four-by-eight sheets of three-quarter-inch plywood to the frame, whole ones first, and then smaller sections cut around the rough openings where the windows and the door would go. A layer of house wrap and then shingles would later be stapled and nailed, respectively, onto the plywood sheathing to complete the building’s walls.

  No other single step in the whole construction process had so swift and dramatic an impact on the building as the nailing up of that plywood cladding. After just a couple of hours of work, the building, which before had stood open to the weather on all sides, had acquired a skin and with that an interior; what had been merely a wooden diagram of a structure was suddenly a house. Until now, Joe and I would always “enter” the structure willy-nilly, stepping in between any two studs wherever we pleased. But as soon as we had nailed up the last sheet of plywood, the only way in was through the door.