A building is often easiest to praise at completion. Walls are clean, services concealed, and a photographer can choose an angle that almost makes us forget the place must accommodate work, lessons, or ordinary life. Years later, cables and cupboards have appeared, partitions moved, and an awkward awning been added. The photograph may be less elegant. For regular users, the building may finally work better.
Stewart Brand's How Buildings Learn concerns those years usually absent from architectural photographs. Published in 1994, with the subtitle What Happens After They're Built, it asks whether buildings can continue changing after people receive them. Brand, the founder of the Whole Earth Catalog, approaches the subject through a longstanding interest in tools and use rather than a plan for an ideal house.
Photographs taken at different times make blocked windows, relocated entrances, and additions legible. Such alterations may have no distinguished designer, but they preserve problems users encountered and responses they devised. A building's history does not end with its architect's name and completion date. Other people keep working on it.
A Building Nobody Had to Admire
One much-discussed example is MIT's Building 20. Constructed as a temporary wartime building, it remained in use for decades, accommodating diverse research and activities. Its unremarkable appearance helped make it adaptable. Users were not constantly required to preserve a complete architectural image.
Returning to it in a 2022 interview, Brand connects low initial cost with the freedom to experiment. Affordable space makes failure easier to bear. An experiment need not wait until somebody will commission a prestigious building for it.
We usually assume good research deserves excellent facilities, and often it does. But before results arrive, the exact needs may be unknown. Rooms and equipment positions fixed too soon can require an inquiry to shrink to fit the premises. Building 20 interests Brand because the order can be reversed: begin with something worth trying, then make the room accommodate it.
Freedom does not depend only on how readily a wall can be altered. An institution must allow decisions, and users need tools, materials, and people able to work with them. A cheap building can remain rigid if every change needs layers of approval. Equally, durable construction need not dictate permanent internal use.
Brand calls inexpensive, inconspicuous, readily rearranged buildings low road. The name is not a poor grade. It identifies an advantage conventional respectability can hide: when people are less afraid of spoiling an original appearance, they have more scope to make their needs real.
This is not the only path to longevity. The book also examines well-resourced, carefully maintained high-road buildings such as Chatsworth. One path relies on sustained care and investment; another benefits from being affordable and forgiving of alteration. Brand compares how they accommodate continued use, rather than merely preserve the day of completion.
Why Must Replacing a Pipe Involve Demolition?
One of the book's most useful distinctions separates parts of a building by their rates of change. Drawing on architect Frank Duffy, Brand identifies site, structure, skin, services, space plan, and stuff. They share an address but age on different schedules.
No specialist training is needed to recognize the difference. A chair can move today; another resident may require a different division of rooms; plumbing and wiring need eventual repair; the load-bearing structure should last much longer. Sealing something relatively short-lived inside something difficult to change can turn maintenance into demolition.
Planning for the future therefore need not mean predicting exactly what residents will want twenty years from now. We will often guess wrong. A more dependable aim is to avoid making one guess prohibitively expensive to revise. Accessible services and adaptable rooms give later users ways to address needs we cannot yet name.
Thorough design does not always assign a purpose to every inch. Sometimes it makes room for uncertainty. But that room still requires thought: what can change, what cannot, and how safety survives alteration. Handing problems to successors differs greatly from giving them means to solve them.
Brand has revised his own model. In a 2022 Scope of Work interview, he acknowledges insufficient attention to landscape. Trees, sunlight, wind, and rain are not cosmetic additions after construction; they affect use from the beginning. He also reconsiders treating the site as almost unchanging.
That revision suits the title. The world need not remain tidy after a diagram is drawn. Heat, environmental change, and inconvenience can call the theorist back to relationships missed earlier. A book about adaptation must itself be open to amendment.
Whose Opinion Matters When the Roof Leaks?
Brand is outspoken about celebrated architects, particularly when photographic effect displaces use. To an occupant, a leaking roof remains a leaking roof. An inventive curve does not make the incoming rain worthwhile.
He also admits failures in a design he once helped promote. The Whole Earth Catalog enthusiastically featured geodesic domes, whose striking geometry carried hopes of technologically transformed living. Here he revisits the practical trouble with waterproofing, dividing interiors, and extending them later.
This admission has more weight than ridicule directed only at architects. When an idea attracts us, we may treat inconvenience as users' failure to adapt or assume technology will soon fix it. Residents must get through tonight while the promise remains unfulfilled. Recognizing that a favored design works poorly places their experience before our enthusiasm.
Maintenance likewise becomes more than paying to repair damage. The people repeatedly handling doors, roofs, and pipes know where wear occurs and which sensible-looking plans create trouble. Their knowledge may arrive as recurring repair jobs rather than a finished theory. If the next design never consults them, buildings cannot readily accumulate experience.
Existing failures do not justify refusing every experiment. Brand's confidence in traditional forms can make some judgments too quick. Old solutions can fail new needs, and beauty need not conflict with usefulness. The strongest demand is not a permanently reliable list of forms: designers should return to see what happened to their work.
Adaptable to Whom?
Readers who enjoy modifying their homes may now be eager to begin. Others will first think of a landlord. Even a wonderfully adaptable house invites little patient investment if a tenant cannot be sure of staying next year.
Rereading the book in 2024, Colin Marshall raises the simple obstacle of ownership and cost. A building available for gradual improvement is already beyond many readers' reach. The objection brings Brand's hopes into contact with different circumstances. One person wonders how to make a place more congenial; another must first get a necessary repair taken seriously.
Applying the book to rented homes, offices, or schools therefore requires thinking about participation. Users may not undertake construction themselves, but they can report difficulties, join discussions, and know who is responsible. Regular occupants may understand stairs, light, or noise better than someone making one inspection. When that experience enters decisions, adaptability becomes more than a material property.
Shared life also imposes boundaries. One household's alteration can affect someone else's safety or peace. Rules are not merely obstacles to creativity; they can protect people with little power to press their own needs. The problem is making reasonable changes possible while maintaining responsibility to neighbors and successors.
How Buildings Learn encourages a second look at awkward seams in an old room. Why is that door there? Where did this passage once lead? Why were those windows changed? Some traces record careless patching. Others preserve serious solutions. The building's present appearance contains their accumulated judgments.
The photographs deserve time. Even after the text moves on, two images separated by years can invite comparison: how much survived, and what did later users add? Visually inconsistent parts may help explain why the place was not abandoned. The next unattractive alteration we notice may prompt a better first question: what problem did it solve?