Skip to content

The Forgotten Function

Post date:

There’s more than one way to incorporate the eclectic and often random stuff of everyday life into the supposed ideal of an architectural whole.

1. Design it all!

This is the default position for architects. For one, it means increased commissions as the rascally perfectionists can charge a client to design the door handles, for example, and then launch their own product ranges. Win win.

2. Build it in.

At one time, built-in furniture offered a solution. Just hide all your stuff in a big cupboard you can’t see. This is what Japanese did traditionally. Sometimes you can’t tell whether your bedding or another room is behind that door. Dedicated rooms called nando were devoted to the storage of seasonal stuff.


There are limits to what can be built-in and for how long. A lot of messy stuff gets stored in kitchens. Kitchen storage, for that matter, creates the illusion of being built-in but the reality is that the entire kitchen is a consumer item only slightly less prone to replacement than sofas.


Sofas are free-standing items we like to think of as a thing for us to choose and not some architect. Some architects beg to differ.

3. Let go! Move on!

The third way is to refuse to design things into the building elements that enclose the space, and to leave the items used within it free to be what they are – personal effects that enable people to do something inside that space. If a space doesn’t contain anything, then all you can do inside it is a) appreciate its “spatial qualities” and that soon gets boring, b) curl up on the floor and sleep, or c) pass through it. This third way, circulation, is The Forgotten Function.

A recent post ended on these three images, all of which show a division between the bits that are functionally linked to the provision of space and the bits that are functionally linked to the occupant doing something in that space.

This division isn’t binary – it’s graded. Internal partitions, for example, are a part of the building and there for sake of the occupant. Doors and windows are more “personal” than walls but still more “architectural” than a cushion, chair, or artwork. This is a useful idea because clear “lines of responsibility” enable buildings to be buildings and focus on what they have to do. It also enables occupants to get on with their lives instead of spending time and money trying to find the thing they think a particular wall or corner is “needing”.

When shopping for such things becomes recreation then, of course, we have a problem. Architects are complicit in this problem by designing walls that “need” sofas, Or corners or niches that “appear empty” without something. An architecture dependent for its effect upon meticulously curated interiors is a weak one.

The possibility to free both buildings and their occupants so they can respectively focus on what they have to do is the appeal of an architecture that doesn’t indulge this nesting desire. Drawing a line between the building and its occupants and their stuff is a good thing but decisions have to be made on where to draw that line. Here’s a traditional Japanese staircase. It’s basically a substantial ladder that allows a person to go between floors via a hole in the ceiling.


The colour and pattern of this next staircase/bookcase isn’t so different from the one above but is now more obviously a part of the building. If the occupants were to move house, they wouldn’t take it with them.


Similar architectural vs. personal decisions have to be made for fireplaces, doors, windows and all the other stuff that makes enclosures liveable. An informal agreement usually exists between architects and occupants, specifying what the architect is expected to provide for in terms of the building side of things. Formal contractual agreements regulate exceptions.

With office space, you lease some space that’s partitioned off from the remainder of the floor. Whether or not you want to divide it further is up to you and how you want your office to be.


I often mention Kazuo Shinohara’s 1976 House in Uehara as having a plan that’s almost incidental.

The structure is busy enough supporting the enclosure – it doesn’t suggest or facilitate any particular arrangement of internal partitions.

But what happens when the structure is the walls themselves? When walls are pressed into service as structural elements and freed of their duty to divide the space into chunks of requisite sizes for our functional convenience?

(Mauricio) Pezo (& Sofia) von Ellrichshausen Architects have been exploring the idea of the corridor-less house. This is their house in Catalonia, Spain. It’s more of an inhabitable structure than an arrangement allowing a program.

© cristobal palma, courtesy of solo houses

Notice how the bathrooms and kitchen are absorbed into a sub-module in a manner similar to Oswald Mathias Ungers’ Haus Kämpchensweg? Ungers, however, goes as far as to absorb the stair.


This house, again by Pezo von Ellrichshausen Architects, takes it a bit further.

Casa Meri

Again, the kitchen has been compromised for the sake of this idea but everything else works more or less. As with the Type B, the size of the habitable rooms are the same size as the non-habitable rooms. This is either its drawback or its charm. There’s no lack of ways to get to some other room since each room has at least two openings. You can get back to the same room by passing through 10 different doors and nine different rooms. This would make a great party house. It may even be possible to get back to the same place after passing through each internal opening once like the Seven Bridges of Konigsberg.


Pezo and von Ellrichshausen do have an obsession with doors, as their 2003 installation 120 Doors shows.

We could combine the idea of this door installation with the idea of Casa Meri, above, and arrive at this next plan. It’s not really a plan – it’s another structural idea that can be lived in but more easily in this case. The “courtyard” could in fact be shrunk to one square metre with four doors opening onto it, considerably simplifying things but altering the proportions of all rooms.


This house is a bit like Haus am Horn in that it’s possible to get to everywhere one needs to without having to go through the central space.

Allplan: 10 WEHAH v14 n0000031.p

Here’s H ArcquitectesCasa Gualba. Like the two plans above house, the central space also has a structural function but this time there are no interconnecting doors between any of the rooms. The central space isn’t a courtyard or a habitable space but a corridor. It works and it works because of the structural triangulation that happens to produce a space that can be used as a corridor. The circulation space, the rooms themselves, and the structural system are one.


Actually, if we take the four-sided pinwheel plan and offset the courtyard, we arrive at one of those beautiful solutions our culture has produced – the one-bedrom apartment plan. True, the corridor could be shrunk some more, but then those convenient niches for the kitchen and bedroom cupboards would disappear. We’d still have to walk by the bedroom cupboards after we entered the room. Nothing would be gained.


Something similar occurs in this next apartment where there’s a 1.2m x 1.8m circulation space.


To the left is a storage cupboard that has to be passed by and on the other side is a 1m x 1.2m column that isn’t going to move. Behind the storage cupboard is the bathroom. One sixth of the corridor is actually in the kitchen but is used to get past the refrigerator against the column. This 1.2m x 1.8m space can’t be used for anything else but it’s not dead space and it’s certainly not unwanted space. It makes everything else work better for other things and this is something it shares with the three pinwheel apartments above.


As Moissei Ginzburg and his Types Team found, the amount of circulation space in an apartment depends not only upon the rooms connected but upon their proportions and arrangement within the apartment.


It’s possible to enclose space efficiently to create a set of spaces. It’s also possible to arbitrarily assign any set of functions to those spaces. However, once this is done, circulation space is needed if the entire space is to be efficiently used.

Circulation space is not a wasteful thing. It is space with the very definite function of making the other spaces work better. This is one of the lessons of The Japanese Machiya and The Types Study. Circulation space needs to be given more respect. Less of it is preferable to too much, but to eliminate it completely is counterproductive and ignores the important function of this useful type of space.