Art and Engineering: How to Build Better Devices

by Mark Bernstein

It is a truth nearly universally acknowledged that programmers and engineers cannot design good user interfaces. The lore of software management is filled with tricks and techniques for getting programmers to do what you want, not what they want, as if they were unruly children. We talk about the inmates running the asylum, and management becomes "adult supervision." On the Web, we see the same tension in the quest of User Experience experts to rein in The Creatives. Architecture arose from the desire to reconcile engineering with art; Information Architecture tries to keep the artists from getting out of hand.

Computer science and software engineering started in the 1950s, a time when society was very angry at science. The scientists, after all, had just invented the atomic bomb, and that seemed likely to kill us all. It was a Sputnik moment. On the one hand, we gave young scientists scholarships and jobs; on the other, we started to call them names: nerd, geek, techie. And we started to believe that they weren't quite normal, that knowing calculus somehow unfits one's moral compass or that understanding recursion deprives one of the ability to make software that "real" people can use.

I think this is a mistake, and that relying too much on the intuitions of focus groups and investors and salespeople may lead toward stagnation. Casual users are inclined to be conservative. Ask them what they want, and you usually find that they want what they have -- just faster and cheaper and with extra magic. The consequence has been a decade of software that is too often dull and unadventurous. We would be better served by software with opinion and attitude, software that takes a stand. Frederick Brooks, Jr., wrote about "software with fan clubs" in The Design of Design; we don't know exactly why some programs are loved and some are merely useful, but the software people feel strongly about almost always starts out with one or two creators.

My own interest lies in making literary devices that are better than books -- both "printed" books that tell you what someone else thought and empty books that help you write stuff down and discover what you think. We need both and we need them urgently. Our legislators cannot understand the policies on which they must decide, and our voters cannot fathom the issues on which they vote. In place of judgment, we substitute the opinion of experts. Today, experts can be found to espouse any option you like. We no longer ask whether a politician is lying, but rather whether the politician knows what he says is untrue or is knowingly appealing to the base.

We now know the future of serious reading and writing lies on the screen. But how should we build that future? The conservative answer has been to reproduce each aspect of the book on the screen, adding some incremental changes like color pictures or video illustration to compensate for the expense and inconvenience.  I think we can do more.

Mark Bernstein is chief scientist at Eastgate Systems, where he crafts software for new ways of reading and writing.