CNU 21 and Type-based Zoning Codes

Copyright Opengate

Copyright Opengate

Recently, particularly at CNU 21, there has been a great deal of discussion of building types in codes. 

Last August, I posted on codes and building types. As I mentioned, there are two main ways of doing codes. One is a matrix-based approach of clean categories, and the other is type-based. The type-based one tends to branch. A townhouse is a kind of house: a branch off the “house” trunk. A major advantage of this approach is that the code can refer to, in that instance, all houses, or just townhouses. It can prohibit all types of houses except townhouses. It’s an extremely supple and flexible way of working. It dovetails with the way that we talk and think about buildings and places.

Of course there are lots of types: types of frontages, types of building, types of use. Andres Duany says that a building type is a combination of three factors:

  1. Use or “function,”
  2. Layout or “configuration,” and
  3. Siting or “disposition.”

But is that the best way to think about it, and how are some other architects thinking about it? There are a number of different code-writers working on type-based codes, including Sandy Sorlien, who edited the current version of the SmartCode — version 9.2. Her “Pocket Code” is extremely short, and it includes some very basic building types. As noted back in August, Moule & Polyzoides uses-based codes extensively, and the Form-Based Code Institute showcases a few building-type based codes. And types are generally compatible with Andrew Burleson’s exciting Adaptive Code. That code begins with street types. In each case, there is a basic problem of specifying the siting or disposition, as Andres calls it, versus the combination of use/function and layout/configuration. 

It seems reasonable to consider the way the building is sited on the lot to be a different matter from the building type itself. We can consider it a detail. For instance, if we say “house,” we know something about the uses and layout of the structure. It is distinct from “office building.” However, we don’t necessarily know whether or not it is attached. Is it a “single-family house?” Is it a “rowhouse?” We don’t know. We normally consider the disposition — the frontage, whether it is attached, whether it has a detached garage or alley access — to be a slightly different matter than the actual type of the building. We say “row-house,” so that “row” modifies “house.”

This suggests that in ordinary use we might really have two sets of types:

  1. Disposition types; these may consist of:
    1. Frontage,
    2. Side and rear yards,
    3. Location of any vehicular access, and
    4. (Possibly) height. 
  1. Building types; these may consist of:
    1. The mix and intensity of uses,
    2. The interior layout, and
    3. The number and type of out-buildings on the lot.

This sort of arrangement of basic concepts might suggest an easier way of organizing codes. It is more in-line with the way that people already talk about buildings. If we’re already accustomed to saying “rowhouse,” it may not be such a great leap to think “row+house.”

Copyright Burlingham   

Copyright Burlingham


The question of use remains somewhat open in this arrangement. Andrew Burleson’s Adaptive Code wouldn’t specify use at all, but would limit “externalities.” This is laudable in a libertarian sort of way, but it leaves us with the question of what counts as an externality. Certainly noise and odor, overshadowing, and so on are traditional nuisances or negative externalities. Presumably pedestrian and vehicular trips would count too. Would a paucity of “eyes on the street” count as a (negative) externality as well?

At CNU 21, or actually in the “Peery Project” space, Matt Lambert of DPZ and Susan Henderson of Placemakers discussed the upcoming SmartCode v. 10.0. Susan Henderson remarked that although they had thought hard about incorporating building types, they had seen no need for it. In fact, they saw scope for a lot of complication. If a code is organized according to matrices and clear categories, the type-and-subtype nature of building types can add complication. (Full disclosure: I have been working with both of them on a project for Phoenix.)

Also at the CNU, Daniel Parolek of Opticos Design presented “missing middle” housing types. These are the types of housing that blend easily in-between full-blown houses and multistory apartment buildings or shops. They include duplexes, triplexes, walk-ups, and walk-ups. They also include “mansion apartments,” which are small apartment houses designed to look like large houses. All of these types would use specific design methods and elements in order to blend better. The same basic building type, such as a 4-flat walkup, can be designed to look almost like a house or almost like a larger apartment building. These types have the distinction of being sub-types of the more general types. That mansion-looking apartment building is a subtype of apartment building: a twig off the branch”walkup apartment building.” And “walkup apartment building” is a branch off of “apartment building.” For this reason, I was intrigued when Daniel Parolek said that in their codes they actually recommend against coding according to building types. This despite the fact that Opticos’s Cincinnati Form-Based Code uses building types. I’m hoping that Daniel will clarify in a comment below.

This quest for a more innovative way of looking at codes seems likely to continue.


on 2013-06-14 20:33 by Bruce Donnelly

Building types as disguises

Someone reminded me that I had once said building types are a great way to handle unusual functions. The best example I know of is at 58 Joralemon Street in Brooklyn, NY.

Creative Commons license: Autopilot

Creative Commons license: Autopilot

This is a subway exit and air vent disguised as a rowhouse. It’s not a perfect fake, but it does its job of protecting the neighborhood from too much intrusion. 

In a code, we might consider this a sub-type of rowhouse. It has the form of a rowhouse, including culturally meaningful features such as a front door, windows, and stoop. Its use deviates from the standard rowhouse, though. 

If we were to write a process for it, we might call it a “disguising” process. That is, a “disguised” subway exit/vent takes a building type and modifies it for a novel use. It is an operation performed on a building type to produce a new building type. The first time it is done, it would require great scrutiny to make sure that it works. Each successive time it is used, though, it would presumably require less scrutiny. 

So among the uses of building types is the ability to disguise novel uses. This supports the original idea of a form-based code: to regulate form more than use. If a subway facility can be disguised as a rowhouse, then most uses such as daycare, small nonprofits, and so on can be too.