GUIdebook: Graphical User Interface galleryHome > Articles > “The myth of the interface”
GUIsTimelinesScreenshotsIconsSoundsSplashesApplicationsAdsVideosArticlesBooksTutorialsExtras
Go backArticlesThe myth of the interface

Reprinted from Personal Computing, issue 4/90, pp. 37-38.

Illustration by John Labbe
This image can be zoomed
In computing today, the interface is ever with us.

You’ll read my colleague John Blackford’s assessment of the significance of an emergent, dominant Unix interface for personal computers in the Technology column in this issue. In this issue’s Special report, you’ll also read about certain “executive interface” issues that govern a class of systems used to keep top management’s finger on the pulse of far-flung enterprises. Among them is the insistence by executive users that personal computers must present information in graphic form, to the specific extent that “problem areas should glow or blink, in red.”

No problem. Software companies with the bucks have no trouble hiring brilliant people to design any kind of screen you want. But in a way, the ability of developers to accommodate this kind of demand and then to celebrate the fact that they did it doesn’t represent a progressive triumph of the computing arts – it’s a kind of stagnation.

You need to know where I’m coming from here: I like graphical environments; I’m probably in Microsoft Windows half the time my personal computer is running. From time to time, however, there’s this little nagging feeling that somehow I have actually moved further away from what it was that pulled me into a working symbiosis with this box of circuits in the first place. The other day, I realized this occasional empty feeling was exactly like the delayed letdown I experienced when I was a kid watching television one day and realized I missed radio.

Radio, in its full spectrum heyday of variety, drama, and world events, was like the DOS prompt in that it asked you to come in and meet it halfway – the understanding was that the really important stuff was going to happen right in your head. Television, whose computing analog is the graphical interface, configures your head into more of a dumb terminal. Whether you’re watching TV, doing Windows, or communing with your Macintosh, the subliminal refrain, in the lexicon of Oz, is: “Pay no attention to the man behind the curtain...”

Permit one further stretch of multimedia analogy: The first time I used a cassette recorder in an interview was the worst job of reporting I ever did. I asked no questions; there was no need to draw anything out, I reckoned, since through this small miracle of technology I was getting it all. On playback, however, it was all gross double-speak. Conclusion: The act of writing down notes had become, through training and habit, the self-induced graphical interface through which whatever analysis I did of a speaker’s veracity, content, and intentions occurred. So to effectively use a tape recorder, paradoxically, you have to listen harder.

What I have been postulating, then, while staring through the Window, is that five years, a hundred seminars, and thousands of vapor trails later, this intense collective obsession with The Interface may have made computing easier, but it has not necessarily made it better.

In fact it hasn’t always made it as easy as it could: I look at – play with – a lot of software programs, maybe four or five new ones a month. There’s only one (a scheduling program) I’ve seen (and now use) that let’s me type “NTH” and “4D” and has no trouble interpreting those, respectively, in the context of its application, as “next Thursday” and “four days.” This is, of course, a text-based DOS program, written by somebody so out of it he or she doesn’t realize the qwerty keyboard tops out the 20th Century’s junkpile.

The fastest, most efficient way to commune with a machine never had a chance in the Age of The Interface. Try this fictional scenario: Instead of waiting until the 8th Century to decide that language has to be written, and borrowing Chinese characters in the most grotesque misapplication of graphic interfaces in history, the Japanese pick up the Roman alphabet from Rome. No question they beat us to the moon after, most probably, MacArthur agrees to terms on the deck of the Yamato. In that kind of a Japan, of course, it would be a simple matter of decree that all students spend a few months in typing class to sustain the population’s national keyboard average of 45wpm. Just as it would be a simple matter in U.S. schools today. Simple, but impossible.

But if it took personal computing only a decade to junk the QWERTY interface, perhaps others will be even more transitory. David Liddle, chairman of Metaphor Computer Systems, a company that has certainly upheld its end of the interface obsession, observes: “One should not confuse product identity with usability. For example, there are companies right now fighting in court over features that are demonstrably bad, in usability terms, but important from a product identity point of view.”

This past fall, Liddle told the following story to a conference and with it, admittedly out of his context, I rest my case:

“The real danger in confusing computing abstractions with the abstractions of users’ jobs is illustrated in the story of the Caribou Eskimo. [They] were tremendously efficient bow and arrow hunters; then came the Hudson’s Bay Company. It gave them rifles, which made them much better hunters. Eventually, the Hudson’s Bay Company left. The Caribou Eskimo were no longer good at listening, smelling, tracking, and getting close enough to something to hit it with a bow and arrow.

“They lost all the abstractions of their jobs as Eskimo hunters, immersed as they were in the user interface abstractions associated with pointing rifles... there was no one living among [them] who knew how to hunt with a bow and arrow. Luckily, some anthropologists had filmed and interviewed some tribesmen 50 years earlier; so, a team went up and retrained the Caribou Eskimos in their old ways.”

by Robin Nelson

Page added on 12th December 2004.

Copyright © 2002-2006 Marcin Wichary, unless stated otherwise.