In pursuing an instrumental understanding of geometry,
the group identified very early on the limits of ‘conventional’
CAD concepts that mimic pen and paper with mouse and
screen, and constrain the architectural language through
libraries of predetermined architectural elements. Robert
Aish explains:
‘There was a direct mapping between what was thought to
be an architectural vocabulary of : “walls, windows and doors”
and a simplified computational equivalent. Maybe this was all
that could be implemented at the time. But the net result, and
disastrous at that, was to entrench this highly limited form of
architecture by making it “more efficient” and excluding to
architecture based on more general geometry or less
conventional components and configurations. What is
different with recent parametric design tools is that the set of
constructs is far more abstract, but at the same time the
system is “extensible”, so that it is the designer who can make
his own vocabulary of components. We have broken the
“hard-coded” naive architectural semantics. We are no longer
interested in “local efficiency” within a restrictive CAD system,
but rather the designer has the opportunity to define his own
vocabulary from first principles, by first understanding the
underlying geometric and algebraic abstractions.’