The joinery
problem of a
pre-war apartment.

A pre-war Manhattan apartment is not square. It is almost never square. The plans on file at the Buildings Department say one thing and the rooms say another, and a closet drawn against the plan will land an inch and a half short of the wall on one side and a quarter inch proud on the other. This is the first thing the survey teaches you, and it is the reason a system out of a box cannot work here.

The buildings we are talking about — the Park Avenue limestone-fronts, the Beresford and the San Remo on the Upper West Side, the Sutton Place co-ops, the side-street brownstones on the Upper East — were built between roughly 1900 and 1939. They were built on the assumption that the people who lived in them would never move a wall, and that the people who installed cabinetry would scribe everything to fit. The interior partitions are plaster and lath: two coats of brown coat, a finish coat of lime, troweled onto wooden lath strips nailed to vertical studs. Over a century the lath dries out, the lime keeps moving, and the wall acquires the gentle wave of a settled bedsheet. It is beautiful and it does not accept a flat panel.

A modern catalog closet system is engineered for a different building. It assumes drywall on metal studs, ceilings at eight feet, corners at ninety degrees, and tolerances of an eighth of an inch. The hanging towers are sold as discrete modules and the trim is generic, sold by the linear foot. In a new-construction condo this works; the building was made for it. In a pre-war apartment the towers will tilt forward because the floor slopes a quarter inch toward the front window. The crown will reveal a one-inch shadow gap on the left end and nothing on the right. The drawer fronts will look fine until they sit, at which point the eye finds the angle in two seconds. We have been called in to take out two such installations in the last year, both of them executed by reputable firms.

The other complication is the ceiling. A pre-war primary bedroom on Park Avenue is rarely under nine and a half feet, and many run to ten or eleven. A catalog closet, configured for eight, will leave a two-foot band of dead air above the tower. The instinct is to fill it with a soffit, which makes the closet look short and the room look squat. The correct answer is to draw the whole elevation to the actual ceiling — a top run for off-season storage, a deep crown that respects the original cornice profile, and a tower height that reads as architecture rather than as furniture pushed against a wall. To do that you have to know the actual ceiling, not the assumed ceiling, which is what the survey is for.

A typical survey for one of these apartments takes us most of a working day. We arrive with a laser distance meter, a small tripod, a steel tape, and a digital plumb that reads to a tenth of a degree. We measure each wall at three heights — fourteen inches off the floor, mid-wall, and twelve inches off the ceiling — and we capture the diagonals corner to corner. We probe the substrate in the corners where the cabinetry will land. We find the studs (which in pre-war framing are not on sixteen-inch centers; they are wherever the carpenter felt like putting them in 1922). We pull the baseboard back a quarter inch with a putty knife and look at what is behind it. We photograph everything. By the end of the day we have a survey document of forty to sixty pages, and the dressing room is no longer a rectangle on a floor plan. It is a particular volume of air with a known geometry.

From that survey the shop draws what becomes a real dressing room, not a closet that grew up. The distinction matters. A closet is hardware-led — rods, drawers, shelves, all in service of storage density. A dressing room is room-led: it starts from the proportion of the wall, the height of the ceiling, the location of the window, the way the door swings, and only then does it ask where the shoes go. In a pre-war apartment the room-led approach is the only one that survives the build, because the room itself is the constraint that the system has to bend around.

Scribe-fitting is the next part. Every cabinet that meets a plaster wall has a scribe rail — a long strip of the same timber as the carcass, left wide and trimmed on site to follow the actual wave of the plaster. The fitter holds the cabinet plumb, slides a sharpened compass along the wall transferring its profile to the scribe, and then takes the rail to a sanding block and works it down until it disappears into the wall. It is slow work; a single scribe can take an hour for a tall cabinet. It is also the only honest answer to a wall that is not flat. We will not caulk a gap closed and call it done. Caulk is a record of a fit that did not happen.

The hanging substrate is the last problem and the most important one. A walk-in closet on the Upper East Side that is fully loaded with men's suits and women's daywear can carry six hundred pounds of clothing on the rod alone. That weight cannot live on plaster and lath. We install a back panel — usually a half-inch plywood substrate ourselves, screwed through the lath into the original studs wherever we can find them, with a small backing field where we can't — and we hang everything from the substrate rather than from the wall. The cabinet hides the substrate; the substrate carries the load; the wall keeps its dignity. This is invisible work and it is the reason the dressing room is still hanging straight in year ten.

A finished project of this kind, in a pre-war primary suite on Park Avenue or Beekman Place, generally runs between four and six weeks on the bench and a further two to three weeks on site. Most run in the $80,000 to $180,000 band, depending on the run, the material — white oak or soap-finished oak are the common requests, walnut for the darker registers, sometimes a hand-rubbed lacquer for the city palette — and whether the room takes an island. A small reach-in inside the same apartment, properly scribed and hung, starts at about $3,500 in basic materials and runs into the teens once the substrate and the finish are accounted for.

None of this is exotic. It is what cabinetmakers did for a century before flat-pack arrived, and it is what these buildings were drawn to receive. A pre-war apartment is one of the most generous frames a closet can live in; it has the height, the room depth, and the wall thickness to make the millwork look as if it had been there from the first day of the lease. It just will not accept a closet that was made anywhere but in the room itself. That is the joinery problem, and the joinery answer is the survey, the scribe, and the patience to draw the room the room actually is.