April 21, 2003

Lessons from Hypertext

Writing in Usability News, Ann Light focuses on part of my Dust Or Magic talk. The headline is fierce (and oddly capitalized): "Web ignored Lessons from Hypertext, says Expert Bernstein."

On the plane to Portland, I spent some time thinking about lessons. It's time, I think, to take stock: what do we know about hypertext? I'm very tired of theories built on air, without real criticism of real hypertexts. And I'm sick of papers and monographs whose authors are too busy (or too lazy) to build on what has gone before, to respond to what we already know, and instead spend their time coining new terms and hauling in new theoretical systems from outside.

The challenge is that, in hypertext research, we have different kinds of knowledge. Some aspects of hypertext are math and science: we know, for example, the algorithmic issues behind one-way vs. two-way links, or the utility of transitive closures. Some aspects are engineering and design knowledge, like Akscyn's Law and colored links; not provable in the same way as science, but still knowable and known. And some aspects are literary knowledge; de gustibus non disputandum, but we know that some works matter and some techniques of craft are effective.

Taking inventory might be a good idea.