Is it mere coincidence that the name of Apple’s new fetish object, the Newton, anagrams to NOT NEW? This flagship in an eventual flotilla of "pen-based personal digital assistants” from Apple and others isn’t as conceptually novel as the Macintosh was when it debuted in 1984, but maybe that’s too long a yardstick by which to measure new gizmos in an age of increasingly jaded users. Whenever I hear gung-ho predictions that technology X (CD-ROM, pen computing, 8-track tapes, whatever) is the wave of the future! and will change your life!, I think of the words of a 1986 Village Voice review of a Devo comeback album: “There’s nothing older than yesterday’s futurism.”
(The future has a habit of defying expectations; weren’t we supposed to be piloting hydrogen cars with a joystick and enjoying a Jetsons-like fully-mechanized home environment by now?)
The hand-held Newton MessagePad lets you draw and write directly on the screen using a special pen (or stylus, if you want to get fancy). Like other pen computers, the Newton features handwriting recognition, the ability to translate your handwriting into text. What’s exclusive to the Newton is its capacity to process your handwriting intelligently. Write “fax Yeltsin” on a Newton and it will do exactly that, based on its built-in understanding of what fax means and a Rolodex-like listing of Yeltsin that you’ve already entered. Using its standard infrared communications port, you can “beam” information to and from other Newtons and electronic organizers like the Sharp Wizard; with optional modems and network connections, you can link to DOS/Windows and Macintosh computers and E-mail nets, and thence The World.
However, Apple chair John Sculley’s grandiose prediction that the Newton “will be seen as the defining technology of the digital age” sets him up for what could be a very big disappointment indeed. Besides the fact that we live in the digital age already, I doubt that Newton’s impact will be clear-cut. Pen computers (very much including the Newton) evoke mixed feelings; it’s possible to make valid, irreconcilable claims about them. Based on conversations with people working in the field (the pen posse?) and my use of a few pen-computer models – an Eo Personal Communicator; a Sharp OZ-9000 “personal organizer”; and a Grid Convertible, a 386 laptop featuring a screen that flips up like a Polaroid SX-70 camera to reveal a keyboard – here’s what I think of the yin/yang claims and counterclaims about pen computing:
Pen computers are/are not intended for people already using computers. Most computers users (virtually everyone in white-collar jobs) can type, if only by the hunt-and-peck method. These users, while apt to be charmed by the novelty of pen computing, are also apt to be frustrated by the lack of a keyboard.
Moreover, it’s telling that a critical market for the Newton (and the competing AT&T Eo models) is the “mobile executive” – the kind of person who presumably either already has a conventional laptop computer for the road or decided years ago, at the dawn of the laptop era, that he or she simply don’t need a computer away from the office. As for keyboard-phobic executives, that prime target market for pen computers: How many are there? And where are they? And why is their unwillingness to type prompting Apple et al. to replace the keyboard baby with pen-based bathwater?
And even if the interfaces of pen computers hide behind ostensibly familiar “desktop” or “notepad” metaphors, we’re still talking about computers, which implies files, memory, disc storage, batteries, and so on. This stuff is complicated. In my own experience, if something goes drastically wrong, it’s more difficult to dig yourself out of trouble with allegedly E-Z graphical interfaces (Mac, Windows, OS/2) than with command-line interfaces (DOS, Unix) that require you to have a good chunk of a priori knowledge. The Newton is too new to spot the holes in its operating-system safety net, but computer neophytes are the last people I’d want to see staking the organization of their lives on personal digital assistants. A Filofax, a pager, and a cellphone might not be as slick, but they sure are dependable.
However, modern Mac and Windows software increasingly asks you to augment mouse clicks – by holding down the Shift, Command, Control, or some other key; by double- or triple-clicking; by pressing a secondary mouse button – to perform functions like extending a selection or popping up a submenu appropriate to what you’re doing at that moment. You’re pretty much up the creek in these cases if you’re holding a pen when your software expects you to be using a mouse. Worse, using a pen for a mouse demands the hand/eye coordination of a fighter pilot. Even double-clicking can go wrong if you don’t hit the same spot on both clicks.
Handwriting is/is not a more “natural” input method than typing. Typing is “natural” to millions of people, and quicker, too; you can correct your errors as you type, instead of waiting for a handwriting recognizer to kick in. But let’s give Apple credit for enabling the Newton to interpret cursive handwriting at all; it’s a singular achievement. (Other pen computers force you to print, not write.) Still, handwriting recognition is as yet too slow and error-prone to be practical. The Apple claim of better-than-90% correct recognition after a few hours of use (the Newton and its user meet halfway: It learns your script over time, and, like living with a child who’s a picky eater, you come to learn what the Newton will choke on) did not materialize in the demo I saw. A Newton with the benefit of weeks of learning the demonstrator’s handwriting fared so poorly that he put it aside and used another Newton never trained on his handwriting, which worked pretty well.
Ironically, while Newtons (future ones with bigger screens, anyway) should be useful for disabled users – as a communication aid if you can’t speak or a control panel for phones, thermostats, and the like – they will render other users functionally disabled. When your only means of computer input is a pen, you become, in effect, a quadriplegic, many of whom have only a head-mounted pointer or a mouthstick with which to manipulate a computer. For such folks, the task of entering text (the virtual albatross around pen computing’s virtual neck) is particularly easy through the use of proven, reliable onscreen keyboards – quite literally, a picture of a keyboard displayed on the screen on which you can “type” via various means.
The Newton does come with an onscreen keyboard, which you use as a last resort if the handwriting recognizer just can’t decipher your scrawl. Microsoft Windows for Pen Computing includes one, too, but it’s junk, the kind of slipshod product Bill Gates can get away with because he thinks he rules the world. You’re better off with a superior product, WiViK (Windows Visual Keyboard) by Toronto’s Bloorview MacMillan Centre. WiViK lets you size and position an onscreen keyboard (or several of them) to your liking and even specify which characters it will display – a QWERTY layout, or accented characters, or a few hard-to-type punctuation symbols (©, ®, §, curly quotation marks).
Most slick, though, is WiViK’s word prediction. (Actually, it’s an optional, extra-cost module, but WiViK without word prediction is like spaghetti without sauce.) Start typing a few characters and WiViK shows you some likely words that begin with those characters (extracted from a dictionary which WiViK updates based on your own usage). Tap on the word you want and boom, it’s typed in for you; though this still requires effort, it almost always takes less time than typing the word out or using handwriting recognition.
A WiViK-like product for the Macintosh, ScreenDoors, also provides onscreen keyboards (well, one at a time, anyway) along with word prediction, though at a no-frills level, lackign the customizability and good looks of WiViK. But the developers of WiViK and ScreenDoors, Fraser Shein and Randy Marsden, are friendly with each other (the onscreen-keyboard biz is a small one, you see), and an improved ScreenDoors is coming in the fall. Marsden murmurs that he may create a Newton product, perhaps one uniting handwriting recognition and word prediction. Now you’re talking.
Better still, Shein and his posse are developing a system of recognizing the path taken by a pointing device as it types out a word. Think of it as the path your index finger would take if you used it to type out words on an imaginary keyboard hovering in the air before you. You wouldn’t have to click on any letters; you’d just move around “on top of” the keyboard. This technique can’t always pinpoint a word (you’d still have to select from some candidate words now and then), but it could be the fastest way to use a computer that lacks a physical keyboard. But I doubt we’ll ever get the chance: Just as the QWERTY keyboard became a standard by default, handwriting recognition is looking like the future of portable computing. That doesn’t mean we have to like it.
[Originally published 1993 ¶ Updated here 1999.06.24]
← Back to keyboards