back to WRITINGS
Interview: Patrik Schumacher, published in: MAD Dinner, by MAD.exe office, Beijing,
published by Actar Barcelona, New York, 2007
MAD: The Modernist architects of the last century, such as Mies
and Frank Lloyd Wright, created provocative relationships
between manmade structures and nature, but in ways that
accentuated the technologies of their era. One tool that they
did not have access to is the computer. How do you think the
development of digital tools has influenced the relationship
between architecture and nature?
PS: The nature reference is an old trope. It’s not only the
Modernists; Renaissance and Classical architecture also reflected
back and wanted to set its constructs within cosmology and an
understanding of nature.
In terms of the digital world, we see a proliferation of natural
morphologies which are exciting and can be made to work and
enrich the compositional and organizational repertoire - not just
in terms of appearance, but as organization patterns. For us [these
morphologies] are navigation and orientation issues that can give
more order to a complex environment.
Beyond that, there is still the issue of construction, expanding
the material systems and exploiting their inherent material
capacities. And that, of course, relates back to Modernist tropes,
but I think we can do it much better now. When you talk about
Modernism, you have shelves occurring, you have hanging
structures occurring, and these are very interesting.
Building on that, we’re working with form-finding of fabric. We
just created these mushroom-style umbrella structures for Hyde
Park [in London], where we used form-finding tools to stretch and
tailor a fabric across a steel frame. These are interesting domains
where we work with geometric systems and material constraints.
We’re just starting to look at how to apply the intelligence of the
computer model to material constraints, the logics of surface
curvature to manufacturing constraints, etc. I think that’s a more
interesting new domain where you work with the inherent formal
biases of the material rather than using milling and casting to
give a shape to the material, a shape which is often not inherent
in the material.
These kinds of techniques exploit the freedoms that we have
now. But I’m interested not only in freedoms, but in constraints,
and that’s a discourse which I think follows from the Modernists
as well. At the AA Design Research Laboratory, we are going a
step further and working literally with material systems, as an
initial process prior to the design process proper, where we are
experimenting with materials - fluids, liquids, magnetic force...
[ We’re conducting ] all sorts of physical experiments that try to
find material logics and morphologies which, in a second step,
are translated into the digital world. These things are affected,
obviously, by all the new tools, like Maya and similar tools that are
imbedded with what I call “quasi-laws of nature”. You can create
dynamic environments that continuously shift between states as
a negotiation of forces represented within a computer, in terms of
All these animation and simulation tools are simulating natural
systems and phenomena. I think on an aesthetic level what we
react to is a sense of complex order and coherency, lawfulness and
relationships within these systems, rather than a kind of random
and unordered system, which becomes boring and dysfunctional
very quickly. These digital tools offer quasi-nature, but it’s all
neatly prepared and directly applicable to the modeling of design
MAD: What greater clarity does this transition from the
physical to the virtual provide?
PS: The kinds of morphologies which come out of it are inherently
lawful and coherent. There were prior attempts to establish this
kind of coherency and law, either through talent and composing,
where architects attempted to intuitively mimic the dynamic
equilibrium, for instance, you find in nature or to just restrict
themselves to platonic geometries and symmetries - very basic
orders. To have a more complex order, rather than some kind of
ugly disorder, is possible with these new tools and, of course,
there is a whole world of tools that exist for all sorts of purposes
outside of architecture that can be harnessed, like RealFlow,
etc. And their whole sensibility, in a sense, relates back to the
perception of natural systems.
Lately, both at the DRL and the office, we’re going into scripted
logics - iterative, recursive scripts - which, in a sense, also build
up quasi-nature and reflect principles that operate in nature.
They’re similar, but perhaps initially more controllable and simpler
than the animation tools that I talked about earlier. You have
force fields, Newton fields, particle strips, etc., and these scripted
logics are based on components and component iterations and
modulations which follow particular laws of gradual change.
Of course we are in communication with other researchers who
are able to do more, but we’re trying now to build up a scripting
unit within the office. We’ve gathered quite a few people who are
working with Autodesk and other software companies to expedite
this kind of expertise.
MAD: How does your position as a teacher inform this process?
PS: Within the office context, we are much more straightforwardly
moving into projects with the repertoires that come out
of prior office research or come through all the teaching research.
In terms of a systematic investigation, it’s [done in] an academic
context, which we then bring into the office work.
One big topic for Zaha over the years was the concept of an
artificial landscape. Using the landscape analogy to inject a new
range of opportunities and options into architecture - going
away from discreet objects and segmentation of territories with
harsh boundaries into gradient, smooth transitions, etc. That
was happening according to a knowledge which had not been
systematically investigated. But it had a big impact on the office
In the mid ‘90s we taught a studio in Hamburg where we set out
to systematically study landscape formations - oceans or certain
types of mountain ranges and geological formations, glaciers, etc.
These were rather specific topics within the natural landscape that
we then investigated in greater detail by looking at the scientists’
tools and how they investigate, track and trace, and graphically
represent these formations. We studied, for example, how they
were formed - through the concepts of strata and faulting and so
on - then tried to draw principles and inspiration from that based
on this more systematic study and the appropriation of the work
of the scientists.
Then at Yale we taught another studio where we looked at
organic life - quite broadly. It ranged from the microscopic to the
macroscopic, and we encouraged [the students] to challenge the
idea of a single whole by looking at cellular structures, subsystems
of the body. Any phenomena in the animal kingdom was a
possible topic of investigation. We were especially interested in
system-environment relationships, the imbedding of organisms,
etc. We studied, for instance, collective organisms, which break
into individualized parts that have their own autonomy, but when
you look closer break into further subparts, or seemingly form
complete entities that aggregate into what we call a “collective
In general we looked at principles, systems, and morphologies.
But again there was an attempt to gain depth, unlike in a
competition setting where one would maybe pick up an image
and be inspired. We wanted to go a bit deeper and give it a few
weeks of research and investigation in order to discover and
elaborate fertile analogies and transferences into the domain of
MAD: What future uses do you imagine for the knowledge
generated by these kinds of investigation?
PS: Of course we are still working more on the side of visual
analogues. There are obviously efforts within architecture to go
beyond purely visual and morphological input and have more
performative principles incorporated. I think with respect to
environmental engineering there is a kind of overlap where,
for instance, currently we’re looking closely at component
differentiation - the way components can populate surfaces
and differentiate themselves according to scripts, picking up
on the surface curvature and surface orientation. You can feed
in orientation with respect to the environment - sunlight, etc. -
and so we’re looking at shading elements - facades, balconies,
etc. - that would be set up such that they would be adaptive to
orientation conditions and weather conditions.
Here you’re going directly into adaptive systems and responsive
systems, which also would then deliver the kind of morphology
that we would like to see where you perceive that logic. You see a
consistently differentiated field of elements - in this case shading
elements - which you need to couple not with a box but with a
blob-like surface, where, as the surface swings around and into
the sunlight, along the changing surface condition these shading
elements would increase or shift angles as you move from south
to west for instance. This becomes quite beautiful, but also very
effective. You can script very precise laws that determine how
this system responds to the selected factors. This is a more direct
and literal embedding of an architectural system into a natural
environment in a relationship of adaptation, which is in fact how
all natural systems fit with each other.
MAD: Earlier you mentioned doing studies of biology, even
microscopic biology, as a means to derive new forms and
systems. Could you tell us more about what this kind of work
PS: We did a studio in Vienna that was inspired by the work of
Frei Otto, with his investigation of natural systems and their
extrapolation into the domain of building structure. This work
was absolutely seminal and forward-looking all through the ‘50s,
‘60s, and ‘70s. We tried to do something like that with respect
to biological systems, taking up particular concepts, like the
distinction between homology and analogy. Again you don’t
only look at the phenomenon - for instance an ant colony - by
itself, but you also read up on the science and tools that explore
and simulate that phenomenon. You can look at microscopic
photography, but you can also look at computer simulations,
and there’s a whole world feeding into that. Architects are not
the only ones: nearly every design discipline is looking at nature
and investigating natural systems as models and analogues.
And you can also go to this second layer of already digested and
transformed representations of nature.
MAD: How do you then jump from investigations at a micro
scale into applications in macro scale - urbanism, planning,
PS: One of our crucial contributions in terms of the office, DRL,
and some of the other teaching venues is something that we call
“Parametric Urbanism”. The idea behind it is that, when talking
about large sites with hundreds of buildings, how do we script the
genotypes or conceive of genotypes as the typical building blocks
that would populate a site that has different conditions?
It starts with simple geometry: at the center of the site you have
more relationality than along the edges of the site. We’re trying to
use the concept of genotype and phenotype to set up a genotype
that has an inherent adaptive capacity or self-differentiating
capacity, and to set up a field of continuously differentiating
buildings that shift in height, depth, but also in terms of the
interarticulation of their subsystems, i.e. the way the volume is
dissected into floors, the way voids appear within the volume, and
the way a navigational space is driven through that mass. So it’s
an urban massing model plus.
A prototype of this is the Science-Hub masterplan in Singapore,
but we’ve applied it at a larger scale in a big urban masterplan that
we’re doing for Istanbul, Turkey. We have two primary genotypes:
a tower type and perimeter block type, which then differentiate
and interarticulate and create all sorts of interesting conditions
where the perimeter block pulls up at the corner to create a corner
tower condition, the blocks open up to allow for the interlacing
and continuous flow of courtyards, etc. The crucial thing is how
this [system] adapts with urban geometry, because we’re not
talking about orthogonal blocks, but blocks which follow the
flow lines of the existing topography of an urban fabric. We’re
managing now to script those systems.
MAD: So, in this case, scripting allows you to work in
controlled way from an aerial perspective?
PS: We’re using scripting to handle a whole urban fabric where
you cannot be so concerned with an individual building. We’re
trying to create a parametric system which results in a beautifully
differentiated field that adapts to its edge conditions, allows
for the imbedding of singularities, but which, in the end, you
can navigate better, because you have gradients of what we call
“laws of transformation” radiating through that field. So you can
follow these vectors of transformation and navigate that field,
because it’s based on two or three laws or logics of how the field
transforms, and you can track and understand that, even quite
intuitively. And this also appears to have the beauty of nature.
In an experimental setting, we exhibited at the Tate Modern
Museum a big study for the Thames Gateway, a very large
eastern expansion of London, where we showed morphogenetic
animations of the way these fabrics have been constituted and
how they then settle and differentiate in that territory and
also how two or three types can interarticulate. So that’s a big
headline for us: Parametric Urbanism - using the nature analogy
to create a quasi-nature which becomes a kind of hypernatural
urban fabric. On the one side it contrasts with the ‘70s new
towns, which are all based on the order of repetition and seriality,
and on the other the chaos of the ‘80s and ‘90s metropolitan
growth where you have an anything goes condition and a visually
indigestible chaos. We’re developing a new paradigm that avoids
both monotony and chaos, and offers coherency and logic and a
quasi-natural lawfulness instead.
MAD: Do you find that these references to natural forms and
spaces make unusual urban ideas more palatable to their
PS: It’s not necessarily recognizable as being similar to
something, particularly not to familiar landscapes necessarily.
We’re talking about swarm organisms, the swarming and flocking
of buildings by orientating them towards each other. But it’s
not that you would necessarily make this association as a user
or visitor. Although it also wouldn’t be detrimental either. We’re
not forcing that, otherwise it becomes representational, and the
danger is that you have a one-liner.
In nature, so many forms and systems are yet unknown to us.
So it’s not about making it familiar, but the principles follow
through. Even with something quite new and unexpected, you
could navigate it intuitively if it has these quasi-laws of nature.
Not necessarily of a familiar cliche of a winding river or something
like this, something meant to be comforting. That’s not what we
have in mind. MAD: Designing a system of navigation or orientation that is
intuitively readable is a kind of Holy Grail for planners. How
do these parametric systems that you’re describing achieve or
at least attempt to strike that intuitive chord with users?
PS: What I mean by intuitive navigation is: there’s an inborn
capacity to recognize pattern and be curious about pattern. If
they are intricate and complex, it inspires even more curiosity
and alertness to try to find out what is going on, what the hidden
laws are. And that should pay off. There should be such hidden
laws; you should then be able to follow the build up of density to
find central functions, for instance. You should be able to figure
out and find and follow these kinds of logic - not necessarily
through an analytical process. I think if you drift through nature
you have the same capacity. That’s something that we’re banking
on - pattern recognition and the way we respond and are alerted
to systems rather than random agglomerations. If you have a
garbage heap spilled on the ground, you’ll pass by without giving
a second glance, but if there were a certain geometric pattern to it
you would kind of stop and try to find out more.
back to WRITINGS