FROM TYPOLOGY TO FORM
What has stunted the evolution of the contemporary Megalopolis, and rendered it opaque or inaccessible as a design problem, is a largely unconscious reliance on urban typology. While modern architecture broke away from typology a century ago — primarily through the plastic innovations of the 1920s — urbanism has yet to catch up with this transition from type to form. In the Fifth Ward project, we wanted to demonstrate what such a transition from urban type to urban form might look like. By urban form it is simply meant something rather more fluid or plastic than a type. Type resides in a singular “essence” that exists as a strictly platonic ideal. As such, type is never realized, it is instead only approximated in instance after instance of its routine application. Each approximation of the type is referred to as a “model.” The fundamental distinction between type as ideal and model as approximation of that ideal accurately describes, not only the urbanism of the past, but the typologically driven urbanism of Megalopolis.
The type that has dominated urbanism for centuries is the simple urban grid and the block structure that it supports. The grid and the block are so fundamental to our definition of urbanism as to be practically transparent to any sustained analysis; we literally do not see how indebted we are to this specific type form. Ladders was an attempt to break through that transparency and see the grid as a historically determined urbanism as opposed to seeing it as the timeless essence of urbanism. To that end, the book took up the history of the grid typology and described an unprecedented episode of grid transformation that took place over the past century. The book described how an evolution of the traditional grid type led to the formation of a new typological form through the processes of systematic erasure or erosion. The typological figure that emerged out of these processes was the cul-de-sac spine or the ladder. The ladder as a unit of urban aggregation was revealed most clearly in the degree-zero urbanism of Ludwig Hilberseimer. His ideal type-form — the Settlement Unit — provides the decisive reading of the typological substratum of contemporary urbanism. That megalopolitan “sprawl” has a typological substrate is still news to many yet, to my knowledge, the idea that the spine type is the basic unit megalopolitan aggregation has never been contested.
What was not given adequate emphasis in Ladders was how the megalopolitan spine sustains a typological tradition bringing both the benefits and consequences of that tradition along with it. Primary among those consequences is typological reduction. The more that a model is reduced to its constituent form, the more closely the typological idea is approximated. The replication of the spine unit in Megalopolis today is exceedingly abstract, almost as minimal and reductive as Hilberseimer’s first description of it. In Megalopolis, the distinction between type as ideal and model as approximation of that ideal is reduced to almost zero. The infinitely small range of variation displayed in cul-de-sac typology today has reduced Megalopolis to an impoverished diagram of habitation. Said another way, the historical movement from the nineteenth century grid to the twentieth century cul-de-sac has been an exercise in extremes. In moving from the endlessly open street system of the gridiron to the closed terminal condition of the cul-de-sac there have been few, if any, stops in between. What caused this reductive swing from one extreme to another was not the evolution of the forms themselves, nor the superiority of one form over the other, but the manner in which we classify forms altogether. This idea is predicated upon a kind of formal succession that is seen not as a transformative process, but the trading of one form for another. Hilberseimer’s Settlement Unit — while an accurate and useful reflection of megalopolitan aggregation — has become our nightmare if only because it has been translated so literally, with so little variation. What the restrictions of type and model have brought to the world of contemporary urbanism is that formal variation has been preempted.
The purpose of the design strategies outlined here is not to advocate the grid or cul-de-sac nor is it to extol the virtues of open or closed typologies. What the project’s strategies attempt is to define the many stops as exist in between the two type forms. These stops define the frontier of contemporary design invention as it is specifically related to Megalopolis. What we attempted to establish in this project is an urbanism that is not bound to the extremes of the grid or the cul-de-sac, the continuous or the discontinuous, the open or the closed. In order to do this a shift must be made, not from one type to another, but a shift to a mode of form-making that is not typologically based. To this end, what we seek is a long overdue transition from traditional urbanism of typology to a modern urbanism of form.
SEARCH SPACE
Variations based on type are tied to essences whereas variations based on forms are based on a theoretically infinite number of variations. In this regard a single form can give rise to infinite variations with each variant having no greater validity (as approximating the essence) than the next. What form and type do have in common, however, is a degree of autonomy. They each carry within themselves the logic of their own reproduction as well as a procedure for variation and subsequent selection. The Fifth Ward project speculates on how an autonomous urbanism of form might loosen the typological limitation on the spine unit bringing a higher (and more sophisticated) degree of variations within Megalopolis. To that end, we demonstrated something called a “search space.” Search space was first imagined, not by environmental designers, but by scientists obsessed with an apparent correlation between digital and biological processes. Its primary theorist was Richard Dawkins who, in the nineteen-nineties, expanded on the connection between digital search procedures and the processes of natural selection.
A brief account of this connection would note that computers solve problems, not by “thinking,” but by systematically generating a set of all possible solutions to any given problem and then searching the resulting “search space” for an optimized solution. As is by now well-known, computers do not think, they search. And what they search is a set or “space” of all possible solutions that is created by an algorithmic procedure. Algorithms are abstract instructions that make digital computation possible. These instructions are applied logically and repetitively in order to yield a series or set of results whenever the algorithm is ‘run’. A simple example may help explain. An algorithm or instruction such as “copy/rotate” can turn a single line or I-shape into an L-shape form when it is applied. Apply the copy/rotate algorithm again and the L-shape becomes a U-shape with the original form and the two resulting forms manifest in the three legs of the “U.” Set the I, L and the U next to each other and you have the beginnings of a field or space of all possible variations to the copy/rotate algorithm’s operation. Of course algorithms can get much more complex than copy/rotate. It is also true that algorithms are also used for the act of searching and selecting. Algorithms generating search spaces that are then searched by algorithms lead to the baroque mathematical houses that make up our most sophisticated software. At the root of it all, however, is a rudimentary procedure of generating a field, searching that field, and making a selection from it.
Dawkins was one of the first to experiment with so-called “genetic algorithms” on computers, back in the early eighties writing simple programs on a rudimentary computer. By using simple instructions he created a program that evolved successive generations of "creatures" he called Biomorphs. The generation of Biomorphs was not limited to a linear, evolutionary process, however, for he imagined all the Biomrphs to take up position simultaneously in space. Dawkins notes that there "is a definite set of biomorphs, each permanently sitting in its own unique place in a mathematical space. It is permanently sitting there in the sense that, if you only knew its genetic formula, you could instantly find it; moreover, its neighbors in this special kind of space are biomorphs that differ from it by only one gene.” (Dawkins, p.65) In his influential book, The Blind Watchmaker, Dawkins describes the process as a simulation of natural selection whereby an optimal [algorithmically induced mutation] is selected for survival through searching amongst all possible mutations produced by the program:
When you first evolve a new creature by artificial selection in the computer model, it feels like a creative process. So it is indeed. But what you are really doing is finding the creature, for it is, in a mathematical sense, already sitting in its own place in the genetic space of Biomorph Land. The reason it is a truly creative process is that finding any particular creature is extremely difficult, simply and purely because Biomorph Land is very, very large, and the total number of creatures sitting there is all but infinite. It isn't feasible just to search aimlessly and at random. You have to adopt some more efficient —creative—searching procedure." p.65
This searching can of course be done by more algorithms in the form of a search engine whereby massive amounts of irrelevant data are automatically filtered out. More creative, however, is the wet engine that operates between our ears. To that end we can construct a thought experiment where all possible solutions to a well-defined problem are laid out in something called a search space. This would create, in a biological sense, a “population” of solutions, where the design that we seek issues from a near infinity of possible solutions. (This method would be opposed to a search for a singular essence.) Design space is simply a table of variations in relation to a specific problem. The possible solutions are generated by a simple set of rules that are arranged in a search space by relative degrees of similarity. One may fine-tune this space to meet needs of the problem. There is either a greater or lesser degree of differentiation to the solutions in the space ranging from entirely different variations on organization to near identical copies. This range in variation can be described as the "mutation rate," borrowed again from biology, where high mutation rates produce fields of wide diversity and where low mutation rates produce tables of similarity or apparent identity. Every design solution thus consists of locating a particular pathway, or trajectory, through "an endless but orderly vista of morphological variety, but one in which every creature is sitting in its correct place, waiting to be discovered." p.66.
Simply stated, what we are interested in the rules for generating form, rather than the forms themselves; we are interested in processes, not artifacts. Constructing a search space is a “packet-of-seeds” as opposed to a “bag-of-bricks” approach. Such an interest in the generation of form of is of course not new. Some proponents see it as the basis of all modern design dating back the origins of Darwinism. One of these proponents is Daniel Dennett who bluntly argues that
“...modern design is about designing the rules for generating form [algorithms], not the form itself. It is about designing processes, not essential forms. This comes directly from a Darwinian logic and it stands in marked contrast to creationist theory, or, the “Argument from Design.” (Daniel Dennett, Darwin’s Dangerous Idea, p.28).
I would argue that the design of processes is what we have for a century called modern design. I would further argue that modern design is modern precisely because it is reconciled to a Darwinian world-view. Modernists have long recognized Natural Selection as the process through which the most complex and sophisticated examples of design intelligence — living systems — have come into existence. Over one hundred years ago, this recognition initiated a search for what is called the "autonomy" of architectural and urban form. In the context of Modern architecture, this autonomy has in part been pursued in the simple belief that form (natural or cultural) must contain a logic capable of generating itself. What this means is that design activity no longer concentrates on the projection of final forms, but on the rules or the logic that generates forms that are never singular, final or complete in and of themselves. An autonomous mechanism is sought that is capable of producing not one Prairie House (to take an obvious example), but an entire species called Prairie House that is unlimited in number and variety. The ultimate reference for this autonomous mechanism is Natural Selection.
Discarding essence and embracing process marks the shift from urban type to urban form. The bottom line of an urbanism of form is the ability to structure the urban environment apart from the strictures of typological reduction. Since the dawn of cities, urban form-making has been a typological process — it has been a process of approximating an ideal by creating essential forms. It has been about the subtracting the extraneous, the circumstantial, and the historical in search of the essential. Megalopolis is an entirely unprecedented kind of urbanism that requires a strategy that is not about elimination alternative s but is instead about actively producing new alternatives. Megalopolis requires a design strategy that is not arbitrarily “free” as no complex urban system can be, but is neither about the opposite of freedom: restriction, elimination and subtraction. The construction of search spaces has but a single objective: to replace subtraction with multiplication (end).
No comments:
Post a Comment