The most impressive arguers are those who start with a tremendous blow to the conventional wisdom and never lose the initiative.
Take the aggressive Clifford Stoll, for example. By profession he is an astronomer, but by vocation a mental pugilist, a crusader against what he calls 'Internet madness'. By this he does not mean the dot-com bubble of '99. He means the wholesale destruction of education by online teaching and computer panic - the fear that unless the latest software is loaded up, university students and school pupils alike will be left behind in a kind of Third World of knowledge, from which there is no escape. To avoid this terrible fate, money earmarked for teaching staff goes instead to pay for computers and Internet connections, software, maintenance contracts and wages for experts. Whole libraries and laboratories are transformed into computer rooms, chemical experiments are replaced by video simulations and thoughtful presentations take a back seat to slick montages downloaded from the Internet.
In his lectures, Stoll paints a picture of an educational landscape where the Internet hovers over every curriculum, dictating subject matter and teaching methods alike. As a result, subjects such as the important-sounding 'teaching of computer skills' turn out to be no more than word processing, e-mailing and Internet access - 'skills' that could be learned in a few days if not a few hours.
As far as I know, Stoll has not yet turned his attention to the use of computers in architecture, but it is easy enough to see how he might. In his view, the bottom line is the insight that reality is something fundamentally different from simulation, so that images cannot be a substitute for the real thing.
Nowhere is this more conspicuously the case than in the world of computer-generated images occupied by the architect. We have all seen the cluster of professionals studying the digitised images - untouched by human hand, the lowering sky cut by searchlight beams - that present a hair-raising though utterly spurious likeness of the project under discussion. Even the imaging subcontractors themselves - the chaps with a tonne of cutting-edge computer equipment and the machine-minding skills to match - groan when the planning officer or the man from English Heritage calls for another six expensive images from four different viewpoints.
This is meant to aid negotiations whose purpose - although ostensibly to evaluate another new circuit board for a working city - ends up as a parody of the selection process for works of art to be displayed in the Royal Academy Summer Exhibition.
It is in such endless farragoes of nonsense that the world of architecture illuminates Stoll's loss of authenticity more clearly than the world of education ever could. But there are other insights into it. Another professor from California, Mario Carpo of the Guggenheim Research Institute, has a take on the same phenomenon seen from an entirely different standpoint. In his brilliant study, Architecture in the Age of Printing, in which he traces the origins of modern architecture back to Gutenberg and the invention of moveable type, Carpo stresses the importance of the difference between publishing and printing, which have not always been connected.
There was publishing before printing, he says, and now there is, because of the computer, publishing after printing. The 500-year alliance of printing and publishing that Marshall MacLuhan extolled as the Gutenberg galaxy is now over. And with it has gone the necessity for an architecture of interchangeable parts, modelled on the invention of moveable type.
The significance of this being grasped, art history can be stood on its head and modern architecture can be placed at the end of an old process of development, instead of at the beginning of a new one.