Redefining the e-newspaper. David Gelernter is identified as a "professor of computer science at Yale," so maybe I'm just too dense to understand what he's talking about.

But in his cover essay for the current Weekly Standard, he argues that the "Next Great American Newspaper" will be conservative (we'll see about that), published on the Web (he's probably right), and implemented in a far more appealing and useful way than today's electronic papers. He writes:

[T]oday's web-papers are wedge-ins, stop-gaps, crack fillers, with all the character of putty in a plastic spritz-tube; people read them not for pleasure and illumination but to extract a necessary fact or kill time when they are stuck at their desks. Their builders don't seem to have grasped what makes the newsprint newspaper one of design history's greatest achievements.

I'm not going to disagree. Otherwise, why would we have four daily papers delivered to our house every day? But, as far as I can tell, e-papers are implemented about as well as today's hardware (limited portability, good-but-not-great displays, faster-but-not-fast-enough access) will allow for.

So what does Gelernter have in mind as an alternative?

Imagine a parade of jumbo index cards standing like set-up dominoes. On your computer display, the parade of index cards stretches into the simulated depths of your screen, from the middle-bottom (where the front-most card stands, looking big) to the farthest-away card in the upper left corner (looking small). Now, something happens: Tony Blair makes a speech. A new card materializes in front (a report on the speech) and everyone else takes a step back--and the farthest-away card falls off the screen and (temporarily) disappears. So the parade is in constant motion. New stories keep popping up in front, and the parade streams backwards to the rear.

There's more to it, of course, but that's where it begins: with cascading virtual index cards. This is a big improvement? I'm unimpressed.