Renovating a Lake Como Villa Without Ruining What Made You Buy It

Ask an owner what made them buy and you will get an answer about the view. The view is not what they bought; it was available from the house next door. What they bought, whether or not they can name it, is proportion, sequence and light: the shape of the rooms against their height, the order in which those rooms are met walking through the house, and the way the building takes a horizontal light off the water.

Those three are the whole of the asset. They are also invisible on a plan, uncosted in any schedule of works, defended by nobody in the room, and easy to destroy while doing something sensible. Every item below is a decision somebody competent will bring you with a straight face, that answers a real problem, and that is the wrong answer.

One thing this piece is not about. What you are permitted to do here, in what order, and what it costs is a separate and serious subject, set out in our guide to renovating a historic villa. Assume for the next ten minutes that everything is allowed. The question is what you should do, which is harder, and which nobody makes you answer.

The first proposal in almost every project is to knock two rooms into one, and it usually ruins both. A room here is proportioned against its own height; two good rooms at four metres tall become one very long room at four metres tall, which is a corridor with furniture in it. It is lit from one wall, so the far half goes dim and stays dim.

It costs sequence too, the loss owners notice last and mind most. These houses are a run of connected rooms with the doors aligned, so that from one end you see through the length of the building. That alignment is why a house of moderate size feels considerable. Take out the middle wall and the run has nothing left to run through.

Ceiling height goes more quietly. Somebody needs to run ducts or drainage across a floor and the cheap route is horizontal, in a new soffit. A dropped ceiling does not simply lower a room: it removes the upper wall, where the cornice is and where the reflected light off the water lands. Anyone who has stood in the same house before and after knows the second version is a hotel bedroom.

The alternative is to move services vertically and accept fewer of them. Redundant chimney flues, a disused back stair, a cupboard sacrificed on each floor, a floor build-up that is not itself historic: those are the routes, and they cost more. A house with four properly serviced rooms is better than one with nine rooms and a soffit in each.

The kitchen and the bathrooms defeat more of these projects than anything else, and both problems come from one source: the house was organised around a household with staff, and the two rooms a family now cares most about are the two that household kept out of sight.

A masonry wall of this thickness is a thermal flywheel: slow to warm and slow to give the heat up, which is why the rooms are pleasant in a heatwave and why a system trying to take the house from cold to comfortable in an afternoon is fighting the building rather than using it. Heat delivered slowly into mass works here. Heat delivered fast does not, and it makes the surfaces unhappy.

That second point is the one owners resist and the one that matters. These interiors spent their lives in a mild, damp, seasonal regime, and taking a house that was cold and empty every winter to a constant modern temperature changes the moisture in every plaster surface, timber floor and panel of joinery in it. The response is shrinkage, movement and cracking, a year or two after the works, by which time nobody connects the two. Bring the house up gradually, and stop heating all of it to one standard: zone it, and heat a smaller winter house inside the larger summer one.

Windows are the other half of this, and where the most damage is done for the best reasons. What replacement costs is not sentiment: it is the section of the glazing bar, slender in a way modern units struggle to match; the unevenness of old glass, which is why reflections on these faades move; and the relationship of frame to reveal, built as one assembly. A window fifteen per cent too heavy in its section does not look fifteen per cent wrong. The whole elevation looks wrong, because a faade like this is a rhythm and the eye reads the rhythm first. Repair and draught-seal the frames, add secondary glazing that can be taken out again, or reglaze restored frames with the slimmest units that fit.

Light in a room here arrives horizontally off the lake and lands on the ceiling and the upper walls, which is why those are the brightest surfaces in it, and why the plaster is pale and burnished and the floor is dark.

Then somebody puts in downlights, which do the exact opposite. They light the floor, leave the ceiling black, throw shadows downwards onto faces, and flatten the surfaces that were meant to be the room's source of brightness, and they require dozens of holes in the one element most likely to be original and least possible to reinstate. Light these rooms the way the lake does instead: wash the ceiling and upper walls from below, use lamps for working light, and light the depth of the room rather than the middle of it.

The same discipline settles the technology. Plant, tanks, controls, racks and cabling belong in the service world of the house, which was always ugly and is used to being altered: the base storey, the back rooms against the hill, a redundant stair. Cabling runs in existing voids, not in new ones cut through historic plaster. All of it will be obsolete within a decade and ripped out twice more in your ownership, so none of it should be cut into something that has lasted two hundred years.

Some decisions are recoverable at the cost of money and embarrassment. Others are not recoverable at any price, and terrazzo alla veneziana is first on the second list: ground and polished in place, repairable by people who still exist, and once it is in a skip there is no version of it that can be bought and installed. The same is true of old cotto, which is why so much of it has quietly left this lake for other people's houses. Original joinery is a one-way door in the literal sense: a matched copy is a copy, and it is obvious in a room full of originals.

The least understood item is plaster. Stripping interior walls back to bare stone has become a default gesture of authenticity and is close to the opposite of one: these walls were plastered from the beginning, the plaster is part of how they handle moisture, and the rough stone interior is a twentieth-century invention. Once it is off, whatever was painted, scored or dated on it has gone too. Decorative schemes turn up under later emulsion more often than owners expect.

The garden belongs on this list, because it is architecture rather than setting. The hedges were the walls of outdoor rooms and the axes were laid out to be walked, so a hedge taken out to open a view removes a room, and a lawn laid where a formal layout was is a demolition with a mower. A cypress marking a turn is part of the design in a way a specimen grown into the view is not.

One question resolves nearly every argument here, and it is not about taste, budget or authenticity. It is whether the thing you are about to do can be undone: an intervention that can be reversed is in a different moral category from one that cannot.

A freestanding kitchen, a secondary glazing panel, a partition scribed to a cornice rather than cut into it, services in existing voids, light on the walls rather than through the ceiling: your successor can take a different view about all of them. Cutting a new opening in a principal elevation, lifting a terrazzo floor, dropping a ceiling, stripping plaster, taking out a wall: those you decide for everyone who owns the house after you, on the basis of how your family wants to live for fifteen years.

This is not an argument for doing nothing. These houses have been altered continuously for as long as they have stood, and the alterations we now admire were confident, expensive and thoroughly of their moment. The point is not timidity but asymmetry: being wrong costs far more on one side of the line than the other, so the reversible option deserves to win every close call. If the answer would take a week to reverse, decide it now. If it would take a generation, it should not be made standing up.