Human-Computer Interaction

3 History of HCI


How did the field of human-computer interaction get started?

One good place to begin our story is in July of 1945, when Vannevar Bush wrote an article for the Atlantic Monthly, later reprinted in Life, called “As We May Think”.

Today, technology has mostly augmented people’s physical abilities; Bush outlined a vision for information technologies that augmented people’s intellectual abilities.

Who is this guy? What’s his deal? And what led to his pressing vision?

Bush was vice-president and dean of engineering at MIT in the 1930’s, where, incidentally, he was Fred Terman’s advisor. Terman went on to become dean of engineering at Stanford and in the eyes of many the father of Silicon Valley.

In 1939, Vannevar Bush moved to Washington. He’s a leading scientific policy maker. He directs a lot of the government funding, and indeed creates and is instrumental in setting up large-scale university research. This administrative effort eventually leads to the creation of the National Science Foundation and ARPA, institutionalizing government-funded scientific research.

The goal of this article, written in the final months of World War II, is to ask “What can government-funded scientists do to create a better world in peace time?” and his vision was a strongly human-centred one.

Bush wrote of a future interactive desk; he calls the system “memex”. The idea is that all of the world’s information would be available on the knowledge worker’s desktop. Key to the memex idea was effective user interfaces for information storage and retrieval. Remember, this is 1945, so there aren’t yet practical digital computers — the first room-scale digital computers were just being built — and the idea was to use microfiche — high density film — to store everything!

Even more impressive, Bush’s memex vision invents hypertext: he has this idea that people could author trails through this information store, save them for later use, and share them with others. But you’re not always at your desk, right? You want technology to come with you. And knowledge workers need to produce content as well as consume it. And the world isn’t just textual; it’s also visual. So Bush imagined that, in the future, you’d wear a camera, right in the centre of your head, like a third eye, and use it to capture stuff. And he worked out a design that made it as easy as possible to take pictures, so there’re no dials or settings to fiddle with.

As with the memex desk, the details turned out differently; but the core vision was right on target. Today, for example, there are more than a billion camera phones that people carry with them. The programmable digital computers that soon follow, like the ENIACS on here, were a huge technological lead-forward. But, as you can see from the wires, the user interface left a lot to be desired.

The idea of providing a more effective interface to computers has a long and storied history, beginning with Grace Hopper’s invention in the early 1950’s of the first compiler.

What’s inspirational for me is that she conceptualized how improved tools could provide a much wider audience with access to computation.

In the intervening years, good programming environments for the desktop and Web enabled legions of developers to create the content that helped put a PC on every desk.

It’s a long path from Grace Hopper’s visionary work on the compiler to the graphical user interface. There are three key highlights I’d like to share with you along the way.

The seeds of direct manipulation were sown at MIT at Lincoln Labs by Ivan Sutherland. The key innovation of the graphical user interface is that the user’s input is performed directly on top of the system’s output. This input-on-output directness makes the interface much easier to understand and much more intuitive. This input-on-output directness makes the system much easier to understand and feel more intuitive. In the case of Sutherland’s Sketchpad, the input was a light pen and the output was an oscilloscope. Here’s a short clip of the system in action:

You can see, we have several unusal pieces of input-output equipment here. We have a scope, and these are unusual at the time. And pushbuttons. Toggle switches. We have several other related devices. This made the TX-2 a fine candidate for the Sketchpad devlepments back in 1961. As I draw my art of say… on the scope, it reinforces what I have in mind. This is in general part of the design process.

The next major stop on our journey is the creation of the mouse and hypertext; these are key foundations for the Web.

In 1945, Doug Engelbart was a navy radar technician. Engelbart spent his monotonous years in the Philippines. In the library, he found a copy of the magazine Life; It reprinted Bush’s Atlantic Monthly article. As John Markov writes, the idea of a device that could extend the power of the human mind left Engelbart awestruck. He had a vision. It took a long time, but eventually he got some funding and set to work. And what Doug Engelbart came up with, he showed to the world in his famous 1968 demo.

The research program that I’m going to describe to you is quickly characterizable by saying, if, in your office, you, as an intellectual worker, were supplied with a computer display, backed up by a computer that was alive for you all day, and was infinitely responsive to the reaction you have, how much value will you derive from that? And in a second you’ll see the screen and it’s working. And the way the tracking spot moves in conjunction with movements of that mouse.)

Engelbart’s mouse worked with two orthogonal wheels. Each was a potentiometer, a variable resistor, like stereos commonly have for a volume knob. So you get about 300° of a turn and that’s it. Its usable parameter provided about 5 inches of motion in each direction.

After the 1968 demo, Doug takes a show on the road. He travels the country with a 16-millimetre Bell & Howell projector. Ivan Sutherland had recently joined the faculty at the University of Utah. Doug comes to visit and shows the demo, and in the audience is Ivan’s PhD student Alan Kay. Alan has been dreaming of a personal computer. He sees Engelbart’s video and his eyes bugged out — they have the same dream. After his PhD, Alan moved to the Stanford AI Lab, where John McCarthy’s group has an early time-sharing system, maybe the place in the world where every person had their own terminal. From there, he moves to Xerox PARC, where he fleshes out his vision of a Dynabook. Here is a picture of the prototype that Alan made in the early 1970’s. This isn’t a functioning computer at all; it’s made out of carboard; it’s a prototype designed to communicate a vision.

With this vision in hand, Alan Kay and his colleagues at Xerox PARC start building the foundation of the first real graphical user interface. It took them a decade to get it all together, to get it ready to ship. Xerox released the STAR computing system in 1981. The STAR featured a bitmapped display, a window-based graphical user interface, icons, folders, mice, ethernet network, file servers, print servers, and email.

This next clip shows the fruit of their labour:

The display screen shows your working environment; we call this the desktop. Using the Move key, you can arrange your desktop in any way you like. Making a copy of one of these blank documents is like turning a sheet off a pad of paper. The [inaudible] of users to make their own form pads is one example of the usatibility [sic] built into the system. The screen closely approximates the appearance of a printed page. And I’m going to copy it into this upper document, and the new paragraph will appear here. This is a little window that lets me set the various parameters of the paragraph.

When the STAR shipped, this was almost four decades after Vannevar Bush’s vision, three decades after Grace Hopper’s compiler, two decades after Doug Engelbart’s first functioning system, and a decade after Alan Kay set off to work building this computer, inspired by the Dynabook ideas.

This is an example of what Bill Buxton calls “the ‘Long Nose’ of Innovation”, that the early ideas behind a new technology paradigm are often seeded decades before the major commercial adoption.

I’m sharing this history as we begin this course for a couple of reasons:

First, seeing seminal work reminds me that good ideas are often klunkier early on. Second, as Johnny Lee and Bill Buxton point out, if you are looking for a good future product, the seeds you need may already be out there in rough form, waiting for you to polish it into a diamond. Third, I just think these early systems are totally awesome and it’s really inspiring.