After a phase during which tools for computer aided design concentrated on simulating the traditional working methodologies of the designer/architect, new possibilities emerge with research into more complex phenomena enabled by the calculation power of computers.
Whilst these emerging possibilities are exciting there is also the problem
"that we are now facing a totally new universe of surfaces and lines for which our western visual culture gives us very little guidance". Bernard Cache in
"Digital Dragons" (version: 21st april 1998 - no longer on-line)
Thus practitioners who want to work with these new possibilities not only need to overcome the technological hurdles but also the cultural inertia and visual conservatism. Cache continues:
"Occidental abstract art has remained too long dominated by the Phytagorian figures just as occidental philosophy is dominated by Platonician ideas." . . . ibid . . .
In addition, early attempts at developing computer tools did not aim at changing the pallette of shape options by including highly curved surfaces. Instead they tried to assist - through something like artificial intelligence - in the creative choice of the arrangement of rooms. The geometric primitive is king, creativity is automated. This misguided kind of approach culminated in the publication of permutationally 'Possible Palladian Villas', Georg Hersey & Richard Freedman, MIT Press, 1992.
In people such as Bernard Cache
"Even if, according to Leibniz, any form, however complicated it be, can be calculated, it doesn't follow that computer will tell us what to do." . . . ibid . . .
a new direction can be identified which focusses on gaining experience with these new surfaces in the built environment.
Cache achieves his complex surface patterns and objects through advanced mathematics and subtractive manufacturing. For the aesthetic evaluation he has relied on cultures with a more sophisticated sense of abstraction, for example Asian sensitivities derived through calligraphic practice. These sensitivities are not limited to the evaluation of the visual product, but extend to the movement that created it. Cache's example describes the drawing of a bamboo tree, yet the same principle applies to the writing of Kanji characters in general:
"Drawing a bamboo tree for example, is not to represent the plant, but consists of a double exercise in types of strokes . A first series of strokes is intended for the segments of the trunk which must have a thick start, then a slightly thinner but regular line before ending thick again, and a second series of strokes is intended for the leaves starting large and then becoming thinner and thinner in a vigorous end. The painter does not represent something but participates in the abstract movement of nature. What is important is not so much the resemblance of the drawing than the quality of the movement of the «elements» that the painter prolongs in his gesture and the breathing which accompanies both the execution and the gaze of the one who contemplates. These abstract exercises start with codified strokes which are then left to free interpretation, in ever more fluid and sophisticated movements so that all kinds of events can be rendered like the rippling of the wind on the water or a sharp ridge in the mountains." . . . ibid . . .
Cache then applies this knowledge to draw parallels between calligraphy and his computer generated shapes.
Personally I favour a more direct approach which would allow the generation of complex shapes through body movement.
Some form of body movement is necessary for almost any creative act. In painting it can range from minute finger movements holding a brush to the expressive squirting of paint in the action paintings by Jackson Pollock. In music a similar range can be observed comparing a flute player with a drummer. More recently developed instruments such as the theremin from the twenties of this century to the latest sensor technology utilised by the group Sensorband enable a unique new way to generate music/sounds through 'full scale' body movement - approaching the domain of dance. To me it would make sense if architecture, as shelter around us, but also as organisation of the space which guides our movement within and through it, could be created through the moving body. First attempts have been made, for example by DECOI Architects (Mark Goulthorpe, Yee Pin Tan and Zainie Zainul) and their project with the Frankfurt Ballet. Whereas shape generation in this case consisted of tracing the movement envelope with no interaction/correction possibility during the actual performance, developments in Virtual Reality should enable generation and modification of the enveloping shapes by the performer in real time. Thus the initial design would be created in the virtual environment on a 1 to 1 scale before the data is passed on to more traditional CAD packages for further processing. I am convinced that such a technology would overcome the anecdotal preference by architects for a 6B pencil which clearly hints at his/her desire to create with a simple tool but plenty of movement rather than positioning a variety of primitives on a relatively small-scale monitor with a fiddly little input device.

Although Frank Gehry relied heavily on computer technology for the realisation of the Guggenheim Museum in Bilbao, Spain, he did not use computers for the initial, creative phase relying instead on pencil sketches and traditional 3D model making techniques.

But Gehry's work proves that I was on the right track, when I stated 2 years ago, that:

" influenced by the forces of intuition and irregularity [I] intend to demonstrate how a combination of virtual reality and automated manufacturing could free architecture from the reductivist geometry of the Age of Lego" Davina Jackson 'Other Geometries' in Architecture Australia, March/April 1997, pp. 44-47
The second part, the manufacturing part is taken care of by simply transfering some of the production technologies from the aeronautical industries to architecture. The challenge that remains is to develop easy and intuitive input/design devices within virtual environments which will assist rather than limit creativity. Then we will have a real chance to add, in Cache's words, a whole new universe of shape options to our architectural vocabulary.
So far the SEMI-AUTOMATIC SPACE DOODLE demonstrator is far from being a design tool - it does, however, give users a basic exprience of seeing, changing and interacting with a complex surface of high irregularity inside a virtual environment.

As the demonstrator gets further developed it is conceivable that the geometry of an existing building could be altered in ways similar to the incisions of Gordon Matta-Clark some 30 years ago, or the incision and definition of a sculptural envelope within the matrix of the building as executed in my "Eisenman Re-visited" project from 1995/96. Back then the project took months to realise intersecting geometry manually within an Animation package - in GMD's virtual environments, if performed on a slightly optimised data-set the alterations could take place in real time.

SEMI-AUTOMATIC-SPACE-DOODLE Benefits to the Virtual Design Community