|
One of the hard-won epiphanies of my engineering career is that technology is never context-free. Most innovation, as Mr. Nobel showed, has to do with blowing things up; but sometimes technology trickles into the marketplace, where it gets a chance to leaven the GNP. The revolutionary things about most technologies is the way they scar the lexicon with marketing lingo.
Saying XML is revolutionary is like saying ASCII is revolutionary. The analogy goes deep, because XML *is* ASCII promoted from the character- to the block-level. And in fact, ASCII *was* revolutionary in the important but narrow sense that a standard symbol system potentiates leaps in text processing. XML potentiates parallel leaps for document and data processing. The rest is hype.
Kendall Clark's article points an unswerving finger at the unclothed emperor. Phrases like "Semantic Web" and "government-wide XML vocabulary," can't disguise the fact that forty years of AI research and 200 years of epistemological research have failed to produce anything resembling schemas for human discourse. In addition to Eco I suggest fans of the Semantic Web read the Dreyfus brothers -- e.g., "What Computers Can't Do" -- for a sobering counter-argument.
Standards need standards-bearers, but technologists need to balance enthusiasm with skepticism if they hope to build things that don't blow up. Clark isn't some sort of Luddite. He's debunking magical thinking -- the enemy of scientific revolution. All this and he's funny, too. We're lucky to have him.
|
Previous Message
