Skip to content

The Uncompleted Apartments

Post date:

These are some semi-detached apartments in India. The surface area of the paired apartments and the gaps between pairs of apartments ensure cross ventilation which is good if you’re cooking Indian food.

The semi-detached pairs are pulled closer together by the shared access node. The open bridges don’t impede cross ventilation and are observable in varying degrees by the eight kitchens on every floor. It’s good, but would be better if the access node were designed to enhance the experience of living in an apartment building even when it is not being used for access. This configuration is the opposite of say, Mies van der Rohe’s 1949 Lake Shore Drive Apartments with their perimeters reserved for habitable rooms only ,and which has validated every speculative/exploitative apartment development since. We need a model that is more living-enhancing.

As in India, much cooking also takes place in Hong Kong where the humidity is also high. Apartments in towers have a high degree of separation for the same reasons but the access core is integrated into the structure, is more dependent on artificial illumination and ventilation, and is invisible to residents.

This next example uses two stairwells to separate apartments that are probably as detached as apartments in a tower can be.

The “Hong Kong typology” denies the existence of the shared access. Once inside, occupants completely turn their back on it. Although the shared access can be observed in the “Indian typology”, this is just a consequence of it being open and in-line with kitchen windows for reasons of airflow.

I’ve made no secret of my admiration for Ricardo Bofill’s 1975 Walden 7 and the proposal I made in Living on top of one another attempted to show how virtues of observable access could be brought into the mainstream as a good thing in itself, and in addition to the advantages for daylighting and ventilation. With that proposal, I was concerned that two elevators for four apartments per floor might seem excessive and so I wanted to explore a low-rise linear variation as an alternative to corridors recalling the less appealing aspects of ocean liners.

But wherever they are and however large they are, all apartment buildings are characterized by a shared structure and a degree of shared access to apartments that, individually, are unlikely to have any architectural invention. Other than Walden 7, I know of no other apartment building that uses the shared nature of apartment access as an occasion for architectural invention, let alone the essence of that architectural invention being the comfort of being a part of something that is greater than oneself – inasmuch as where one lives is an indicator of that.

The previous proposals in Living on top of one another and The Landscape Within went some way towards replicating the advantages of Walden 7 in a mainstream apartment building. The proposal now, does the same for a linear low-rise apartment building having two elevators. The previous lobby spaces were social spaces that showed activity in terms of illumination and perhaps sound happening within individual apartments but also on the stairs and going to and from the elevators. Now the building is stretched, there’s more lobby level activity.

This was my first attempt at an “inflexible” apartment module to extend the previous round tower. Anticipating better use of external wall area, I shifted the internal stair to the inside of the building. If the basic apartment type is the 1-bedroom apartment, then a 2-bedroom apartment can be configured by appropriating the bedroom of a 1-bedroom apartment above or below. The green 2-bedroom (or possibly 3-bedroom) apartment upstairs would have a small bedroom and corridor where the bedroom is downstairs. In this way, apartments having different numbers of bedrooms can be configured as for the tower in Living on top of one another.

Although the position of the kitchen window was good, I didn’t like the bathroom not being naturally lit or ventilated. Also, that in-side living room window opening onto the access corridor wasn’t nice.

Moving on …

Bathrooms now have windows and, with the intention of maximising external wall window area, I tried placing the stairs in the middle.

The problem with this arrangement is that two staircases are needed to configure apartments with three or more bedrooms. The living room window problem remains. It’s possible to remove one of the staircases and have larger bedrooms, and to configure a 2-bedroom apartment by simply making the staircase into a corridor to access a larger bedroom where the green one currently is. The bedroom doors in this layout are symmetrical about the axis of symmetry. As long as they stay in that position, there’s no reason why the second bedroom can’t be on the floor above or below. It doesn’t matter if the corridor runs horizontal or is inclined (i.e. stairs).

This is not my discovery. The non-disadvantages of inclined corridors have been mentioned in this blog before, in 1928: The Meeting when, in response to the plenum committee’s comments on the Type F apartment, Moisei Ginzburg responded “Staircases take up area but so too would the corridor they function as.”

However, when the basic apartment is a 1-bedroom apartment [red, below] that can have a bedroom appropriated to create a two bedroom apartment [blue] and a studio [yellow], it is better to keep everything on the one level as the length of corridor to access the second bedroom is less than that of a staircase, and the saved area can be diverted to the studio and produce a better studio, but that second bedroom door is a variation and the object of this exercise is to have as few variations as possible.

There are the following consequences.

  • The number of 2-bedroom apartments will always be the same as the number of studio apartments.
  • Equal numbers of apartment types can be obtained by building four storeys of the S+2 layout for every two storeys of the 1+1 layout.
  • It isn’t possible to configure apartments with three or more bedrooms.
  • The living room window to the inside remains a problem. The kitchen window may be a matter of personal preference but the bathroom windows on the lowest of the three levels are not. Even if sills are set at 1.8m … it’s still a bit nasty.

Next up.

  • This next configuration does without the staircase and the bathroom windows are now about 1.8m away from the access corridor.
  • This outdoor area is envisaged as a laundry drying area but also doubles as shaft for utility pipes. Decorative CMU screen it from the access corridor. Human activity is less intrusive and at night the corridor is illuminated largely by borrowed light from kitchens and bathrooms.
  • If the stairwell is included then it must be used efficiently and this means splitting the landings and having apartments on both sides access an additional bedroom as often as possible. The example above shows a 3-bedroom or possibly a 4-bedroom apartment. [Extension is in principle infinite but, realistically, is limited by the living room area.]
  • One-bedroom apartments can’t be paired as before. Every three stacked 1-bedroom apartments is equivalent to a 3-bedroom apartment over three floors, plus two studio apartments.
  • Without wanting to get too geeky, two 2-bedroom apartments with living rooms on the same level can access their second bedroom from opposite sides of a split landing, producing a 1-bedroom apartment and a studio apartment on the floors above and below. This produces equal numbers of studio, 1-bedroom and 2-bedroom apartments.

Q: But why would one want to do this, when equal numbers of the three types of apartment can be easily generated using the previous layout?

A: Because an apartment seems larger when its full extent can’t be comprehended instantly from a single position. [c.f. The Inscrutable Apartment

Consider Shinohara’s 1966 House of Earth.  The bedroom offers a very different spatial experience – a different place to be – that’s all the more enhanced by it not sharing any walls with other parts of the house. With less art and more ingenuity, the bedrooms in André Devin’s circa 1960 Cité Frais Vallon apartments are not above or below the living areas.

  • If the starting point is not two one-bedroom apartments separated by a staircase but a one-bedroom and a two-bedroom apartment then a wider range of variations are possible. 
  • The screened laundry drying areas will make the corridor more pleasant to walk through and may even have a microclimate advantage.
  • Kitchen windows have long views down access corridors for belongingness and security.
  • Habitable rooms have windows facing outwards while those of non-habitable room face inwards.
  • Those in-side living room windows now look across the lobby and glimpses of sky will be seen through the external stairs. The windows on the lowest level will either have to be omitted, be transluscent, be secure, or be high-level. There are also the historic solutions of inside shutters, roller blinds, venetian blinds, net curtains, café curtains …
  • This configuration provides better daylighting and ventilation and, accordingly, a reduction in energy used for them. 
  • 35% less area is used for access when compared with a conventional configuatio for apartment access, and the percentage of GFA used for access is 13.0% as opposed to 18.5%.
  • A building with two elevators servicing 72 apartments could be either a six-storey building with the six-stair layout shown below, or a nine-storey building having a four-stair layout.

• • • 


Linking the triple-height lobbies produced spaces that reminded me of Shinohara’s Uncompleted House.

Nobody knows what Shinohara meant by giving this house that name. We may want to think he saw the space as being completed by the addition of people but if he did he never let on. I doubt he did for it would’ve been contrary to how he encouraged us to comprehend not just Uncompleted House but all the others as well. It doesn’t matter as it’s all history now. What matters is if it’s fine for a lobby and a corridor to make an architectural statement in a private house then why not in an apartment building? So here goes.

I’m not saying this is my idea – I’ve just appropriated it and upscaled it. There’s nothing stopping us from mining the history of architecture for ideas that can be taken out of the context of their architect’s intentions and appropriated and applied to new solutions to the same problems. It’s a relatively unexplored aspect of architectural creativity. I think it’s worthwhile to continually scan history for things we might missed at the time but that, perhaps in combination with other things, might be part of a solution to a new problem.

This proposal is thus an unashamed hybrid of at least four projects.

  1. First and foremost, it makes the shared nature of structure and apartment access the drivers for why the building is configured the way it is, as with Walden 7.
  2. It borrows the enigmatically symbolic core configuration of Kazuo Shinohara’s 1970 Uncompleted House, and presses it into service to give meaning to the shared spaces in a multiple occupancy building.
  3. Apartments of different sizes are configured using the same principle in my own Inflexible House proposal of last year (which is itself an appropriation from Yemeni vernacular houses [c.f. The Buildings of Yemen]).
  4. Finally, the simplicity and rationality of structure and construction owes something to Krantz & Sheldon’s Perth apartment buildings. [c.f. Architectural Misfit #27: Harold Krantz]

Here I’m not championing history as anything other than a repository of useful resources. Unless history is taught for the purpose of creating one’s own knowledge resource in order to make connections on the fly in response to some problem, then students are quite right to ask what’s the point of knowing it. For the same reason, we can all raise the same doubts about our newer, online repositories of architectural solutions that, exactly like history, are limited by their indexing. Their preference for novelty reflects our preference for novelty, probably to mask a declining ability to connect items of existing information for ourselves and synthesize them into something new. This would explain why never before in the history of the world have we had so much architectural information yet so little appetite to apply any of it. I’m filing this one under Education anyway.