George Legendre AA School of Architecture, London, United Kingdom 201
only two threads are left to perform an elementary structural role, as you can see in
the foreground.
Using the longitudinal indexical threads to dissociate the structural and skin-
related functions of the surface makes a lot of sense, but it also comes from –and
leads back to- the routine segregation of form and structure. It is difficult to resist
the iconicity of the load-bearing spine, and Kun-Min Kim, who made this beautiful
sample, must have had many ready-made references in mind.
And yet, in this example by my Princeton student Yusuke Okabayashi, it is clear
that such decisions are pure problems of interpretation. If both indicial threads are
given equal prominence, as is the case here, the resulting tectonic arrangement func-
tions as a continuous, distributed and polycentric whole that blurs the boundaries
between formal and structural functions. For this extraordinary example there is no
blueprint, no pre-made reference, and certainly none of the staidness we casually asso-
ciate with parallelism and flatness (and yet everything here is parallel and flat).
If the indicial threads are not two-dimensional, as shown on this math worksheet,
we can use them to print doubly curved members in depth. The threads are effectively
laminated from the ground up, as I described earlier.
The structural and morphological questions raised by this model are of course
vastly different. This is a peculiar case of degenerate weave: a woven arrangement
of indicial threads veering into a disorderly hyperactive pattern (in a seemingly
uncontrollable cancerous mode, if you allow me the uneasy metaphor again). Threads
that used to be parallel begin to intersect (the intersections are flagged in red) and
half the form is becoming potentially self-structural. The other half is then removed
(i.e. the transversal threads) and the remainder is laminated into a self-supporting
structure that shares the morphological characteristics of half and whole.
In the end such choices may be traced back to basic questions of mathematical
modulation: how many indices have you used in your equations? Are the transforma-
tions periodic? and so on. There are various machinery-friendly equations if you like,
some better suited to sheet cutting, others to lamination.
My point here is that materiality does not ALWAYS begin with matter. The pos-
sibility of producing material undercuts, recesses, self-intersections, holes and other
telltale signs of formal complexity is linked to the mathematical model’s ability to
describe variation, and to our own ability –and willingness– to qualify variation
conceptually. Consider for instance the rather impressive medium of rapid prototyp-
ing: unlike the laser-cutter, its technically inferior counterpart, the one-to-one cor-
respondence RP draws between data model and physical artifact eliminates the need
for the kind of tectonic interpretations shown on the screen. Is that good or bad?
As you can see, prototyping with tectonics and flat sheet cutting can be a much
more intense proposition, so we are entitled to ask: is rapid prototyping primitive?
Is laser cutting advanced? Both systems do one thing very well. I love splitting hair
about instruments because deep down, and even on the surface, their limitations
reflect our own.