The house had good bones and a terrible plan.

When the clients bought it, the kitchen was in the darkest corner of the ground floor, the back porch had been enclosed in the 1980s to make a windowless den, and a load-bearing wall had been removed at some point by somebody who did not replace it with anything. The staircase was beautiful. Almost nothing else had survived intact.

Our first proposal was to move the kitchen to the south-east corner, where the morning light is, and to open the enclosed porch back up. Both of those meant structural work, and both of them were more expensive than the clients had hoped. We spent two months making the case, and in the end the thing that persuaded them was not a drawing but an afternoon: we asked them to sit in the dark corner at 8am and then in the south-east room at 8am.

Taking the house back

Demolition turned up what demolition usually turns up. The 1980s work had been done directly over the original lath, the chimney had been capped without being vented, and there was a complete second layer of flooring hiding underneath the kitchen vinyl. Some of this was good news: the original pine floors were intact under two later layers, and once lifted, cleaned and finished they became the single best material in the house.

We kept: the staircase, the original floors, three of the five fireplaces, the pantry door, and the window proportions on the whole south elevation.

We removed: everything added after 1960.

The kitchen argument

The kitchen took eleven months and roughly forty drawings, and the disagreement was never about style. It was about whether to put an island in it.

The clients wanted one, because every kitchen they had admired for ten years had one. We thought the room was too narrow, and that an island would turn a generous space into two corridors. Neither side was obviously right, so we did the only useful thing: we taped the island out on the floor at full size and asked them to cook dinner around it for a week.

They called on the fourth day and asked us to take it out.

What went in instead was a long work table that can be moved, a deeper run of counter on the north wall, and the stone sink under the window, which is where the person washing up actually wants to be. The room is better for the argument, and we would not have got there by being right at them in a meeting.

Materials

• Lime plaster throughout the ground floor, tinted warm white
• Reclaimed pine, original to the house, cleaned and oiled
• Honed Vermont soapstone for all wet counters
• Unlacquered brass, left to go where it wants to go
• Painted poplar joinery, hand-finished on site

What we would do differently

We under-specified lighting in the back hall and had to add two fittings after the plaster was finished, which is an expensive way to learn something we already knew. And we would push harder, earlier, on the mudroom: it was the last room designed, it got the least attention, and it is the room the family uses most.

The site did most of the design work before we arrived.

The lot drops nearly forty feet from road to shoreline, which meant the architects had already resolved the big questions by the time we came on: the house steps down the slope in three plates, and almost every room has water on one side and rock on the other. Our job was the inside, and specifically the problem of a house with a spectacular view in every direction.

That sounds like a gift. In practice a house like this can become a corridor of windows that nobody wants to sit in, because there is nowhere to put your back. Most of what we did was build places to sit with your back to something solid.

Warmth against a cold lake

The palette came out of the site in the most literal way. The timber is Eastern hemlock, milled from trees cleared for the driveway, and the stone in the hearth and the bathrooms was quarried about ninety minutes away. Both were cheaper than importing something more refined, and both anchor the house in a way an imported material would not have.

Against that we kept the textiles soft and slightly imprecise: undyed wool, heavy linen, a lot of it washed before installation so that nothing looked tense. The clients have two large dogs and a boat, and a house like this needs to survive wet things being put down on it.

The joinery

Nearly everything built-in was drawn by us and made by a cabinetmaker in Glens Falls who has now worked on four of our projects. On a house this size, custom joinery is usually where the budget goes to die, so we were deliberate: we spent it on the three pieces people touch every day and bought the rest.

• A twelve-foot bookcase wall in the main room, hemlock, oiled
• Window seats in both lake-facing bedrooms, with storage under
• The mudroom, which on a lake house is the most important room in the building

A lake house is really a machine for dealing with wet things.

Materials

• Eastern hemlock, milled on site
• Local grey granite, split face and honed
• Lime-washed walls in a cool white
• Undyed wool, washed linen, oiled leather

What we learned

Washing the textiles before install was the single best decision on the project and cost almost nothing. We now do it as standard. We also over-thought the lighting plan for the main room and ended up removing a third of the fittings during commissioning, which the electrician was very gracious about.

We started the studio because we wanted to stay on projects longer.

Nadia came from a large architecture practice in New York, Reuben from twelve years of furniture making. Both of us left for roughly the same reason: the work we cared about was the slow part, and the slow part was the first thing cut when a practice got busy.

So the studio is built to stay small. Six people, four or five projects a year, and a principal on every one of them from the first conversation to the day the furniture lands. We do not have a business development department. We have never pitched for work in a competitive bid, and we do not intend to.

What we are good at

Old houses, mostly. Buildings that have been altered badly and need somebody patient enough to work out which parts are original, which parts are worth keeping, and which parts are simply in the way. That work is slow and unglamorous and we genuinely enjoy it.

We are also good at the long middle of a project, which is where most of the value is and almost none of the photographs come from. Schedules, specifications, site visits, and answering a builder's question the same day instead of three days later.

What we are not

We are not decorators, in the sense that we rarely come in at the end to furnish a finished house. We are not fast. And we are not the right studio for anyone who wants a house that looks like a particular magazine, because the honest answer is that we will spend your money arguing with you about it.

The best compliment we get is that a room looks like it was always that way.

The people

Nadia Hollis leads design. She trained as an architect, spent nine years at a large practice in New York, and is the person who will make you sit in a dark room at eight in the morning to prove a point.

Reuben Vane leads detailing and delivery. He made furniture for twelve years before the studio, which is why our drawings are more specific about joinery than most, and why our cabinetmakers tend to stay with us.

The rest of the studio is four designers, one of whom has been with us since the second year. We are in a converted feed store on Garden Street in Rhinebeck, with a workroom full of samples that visitors are welcome to handle.

A short history

• 2014 Founded above a bakery in Kingston, two people, one project.
• 2017 Moved to Rhinebeck. First full architecture-and-interiors commission.
• 2020 Nine Mile completed, our first project to be published.
• 2023 Opened the sample workroom to clients and trade.
• 2026 Six people, and still four or five projects a year.

Lime plaster is the most oversold material of the last decade, and we still use it.

Both of those things are true, which makes writing about it awkward. It has been photographed so relentlessly that clients now arrive asking for it by name, usually having seen it in a room shot at the exact hour of the exact day when it looks best. So in the spring we went back to four houses where we had specified it, between ten and twenty months after completion, and asked the people living in them what they actually thought.

What it genuinely does

The thing lime plaster does that no paint does is hold light badly, in the best way. The surface is faintly irregular, so instead of reflecting evenly it gathers light in some places and loses it in others, and the wall changes noticeably through the day. In a north-facing room this is the difference between flat and alive.

It is also breathable, which matters enormously in old houses with no vapour barrier. At Ridgeview the walls are solid, there is no cavity, and a modern impermeable paint would have trapped moisture in the fabric. That is not an aesthetic argument, it is a building one, and it is the reason we first started using the material.

And it ages in a way most people like. Small knocks burnish rather than chip. After a year, the wall around a doorway in a busy house has a soft patch where hands have been, which reads as use rather than damage.

It gets better at exactly the rate paint gets worse.

What it does badly

Now the honest part, because this is where clients get surprised.

It cannot be touched up

This is the big one. If you damage a painted wall, you repaint a patch and in most cases nobody can tell. If you gouge lime plaster, the repair is visible, because the finish is a product of how one person worked one surface on one day. In two of the four houses there was a repair that the owners could point to immediately. In one they had stopped noticing. In the other they had not.

It is not uniformly waterproof

Sealed and specified correctly it is fine in a bathroom. Specified casually it is not. We had one shower surround where water was sitting on a horizontal detail and the finish softened. That was our detailing error rather than the material failing, but it is the kind of error the material punishes and paint forgives.

The cost is in labour, and labour varies

The material is cheap. The finish is entirely in the hands of whoever applies it, and the gap between a good plasterer and an adequate one is enormous and invisible until the wall is dry. We now work with two people and will delay a project to get one of them rather than take whoever is available.

Who should not use it

• Anyone who needs to repaint on a schedule. Rentals, in particular. The maintenance model does not fit.
• Anyone who wants an even wall. If variation reads to you as a defect rather than a quality, you will be unhappy, and no amount of explanation at the sample stage fixes that.
• Anyone on a tight budget who wants it everywhere. Put it in two rooms that get good light and paint the rest well. That is a better house than lime plaster throughout and nothing left for the joinery.

What we do now

We still specify it, on about half our projects, and we have changed three things. We sample on site rather than in the studio, on the actual wall, in the actual light, and we leave it up for a week. We write the maintenance reality into the handover document instead of mentioning it in a meeting. And we budget a small sum for a plasterer to return at twelve months, because by then the clients know which three spots bother them.

That last one has been the most useful change by some distance. It converts the material's main weakness into a scheduled visit rather than a slow irritation, and it costs less than the arguing it prevents.