Doug has just passed away. I honor his memory by publishing this interview I made with him 26 years ago.
Conversations with Douglas Engelbart
by George Pór
Computer Currents, 1987
The coming “big bang” of the ’90s will be an explosion of information and knowledge, like nothing we have ever seen. Never have there been as many high-powered means of information production in the hands of as many people as today. The converging knowledge technologies of hypermedia and expert systems, and along with optical mass storage and high-speed data transmission will further the movement towards the “big bang”.
There are two immediately noticeable trends signaling the shape of things to come. One of them is: the number of individuals, groups, and communities that can interact with each other through various electronic media is continuously increasing, and so does the size of shareable communications space. The other trend is: the communication media, modes and channels available to individuals and organizations are continuously expanding, and so does the bandwidth of shareable communications space.
As these trends pick up momentum, they’ll have an irreversible impact on the very fabric of all information-based cultures and economies. How beneficial or harmful that impact will be depends on how prepared we are to co-evolve our thinking and working habits with our rapidly evolving information environment.
The same technologies that will lead to the “big bang” of the ’90s can help us develop the new cognitive and organizational strategies we need to cope with it. Learning those strategies will be a condition to making our electronic “rite of passage.” Douglas Engelbart’s quarter-of-a-century quest for “high-powered electronic tools” to support “augmented knowledge work” was never as timely as today.
Engelbart is one of America’s least-known treasures. Those in the computer world associate his name with the first window on a computer screen, the first interactive telecomputing demonstration, the first mouse, the first word- and outline-processor, the first shared screens for teamwork, and other technological breakthroughs. Let me introduce you to another Engelbart, the one whose ultimate concern is how to develop sophisticated computer and human systems for “augmentation.”
“By ‘augmenting man’s intellect’ we mean increasing the capability of a man to approach a complex problem situation, gain comprehension to suit his particular needs, and to derive solutions to problems… We do not speak of isolated clever tricks that help in particular situations. We refer to a way of life in an integrated domain where hunches, cut-and-try, intangibles, and the human ‘feel for a situation’ usefully coexist with powerful concepts, streamlined terminology and notation, sophisticated methods, and high-powered electronic aids.” (A Conceptual Framework for the Augmentation of Man’s Intellect, 1963, by Douglas Engelbart)
The idea of enhancing our thinking faculties with a “machine” made for that purpose was introduced to the public at the end of WWII by Vannevar Bush, science advisor to President Roosevelt. He was the first to point out the need to create a machine with levers and buttons that would enable us to sort out the myriad pieces of information of all shapes and colors pouring in. He called it “memex.” That was the year 1945.
Forty-two years later, searching for a way to explain the amazing scope of things its new hypermedia program can do, Apple Computer got a hand from Bush. The brochure introducing HyperCard says, “As Bush described it, the human mind snaps instantly from one related thought to another, following an intricate web of associative trails: ‘the speed of the action, the intricacy of the trails, the detail of mental pictures is awe inspiring… Man cannot hope to fully duplicate this mental process artificially,’ he added ‘but he certainly ought to be able to learn from it.’ Bravo, Mr. Bush. Bravo. Finally, somebody has.”
Bravo, Apple. Bravo. We owe it to ourselves to give credit where it’s due and to acknowledge that Engelbart not only discovered Bush over four decades ago, but also was the first to build a working model of the kind of mind-amplifier Bush was dreaming about. Augment, a sophisticated software environment that Engelbart developed, is an enchanted land for the mind where you can reach out, grab a thought, hop on it, and ride in any direction you want, with perfect confidence in the system’s ability to take you within seconds to any nook or cranny of any huge knowledge web of your fancy.
Once recognized and systematically applied in the planning and management of knowledge work, augmentation may well become the single most important factor that will increase intellectual wealth all across the board. All we need is the right model of integrating the right technology with the right methodology. What is right? The one that will enable us to learn what it takes to graduate to the next plateau of human evolution. How can we recognize them? That’s one of the underlying questions of the Conversations with Douglas Engelbart. Here follows the interview.
George Pór: Doug, years before the computer revolution started, you were talking about computer support to ‘augment man’s intellect’? Later, you de-emphasized the ‘intellect.’ Would you explain why?
Doug Engelbart: When I started using the word ‘augment’ as a term, in the 60’s, the word ‘intellect’ didn’t satisfy me, but I couldn’t think of anything better until about l968, when I found Peter Drucker’s book The Age of Discontinuity. I don’t know if he coined the terms ‘knowledge work’ and ‘knowledge organizations’, but his description of them certainly fit, and I said Ah ha! That’s more exactly what I’m trying to do; so I’ll call it augmenting knowledge work instead of augmenting intellect. I want to boost that capability. Then I thought, if what I’m trying to do is to better the environment for the knowledge worker, why not call it a knowledge workshop? So we started calling it the Augmented Knowledge Workshop, or AKW.
George: How does the Augmented Knowledge Workshop support collaborative work?
Doug: I wrote a paper in 1972 about coordinating information services for communities. That was the intent then Ñ to support community. I assumed that a knowledge workshop for a community had some aspects different from what an individual’s would be. An individual needs his augmented knowledge workshop to do his own work. But on top of that there is a set of things that facilitate sharing and collaboration with a community … sort of like a second shell outside the individual one. The individual’s workshop needs to be compatible with what the community is going to use.
George: What features need to be common in the individual and the community workshops?
Doug: The core of what you have to share is the development, production, and control of information modules or knowledge modules, called documents. They are bundles of information containing an exposition or argument or treatise or memo or proposal or whatever. We’ve been calling these bundles documents for years. I felt this was the core we needed to manipulate and work with.
The hypertext nature of the document is almost as valuable to your individual work as it is when you’re sharing with a group of people. A multimedia assumption was there from the outset. Assuming you have the capability, within a community you have to be able to share documents in some standard way and control what you are going to look at and share. We must be able to produce documents that everyone can work with, that are addressable and that we can make citation links with. For me the best structuring allows lots of flexible viewing, so that while you find your way around, you view and generate views appropriate to the need of the moment.
You also need dialogue support. For us that was the mail and the Journal. The mail system has to be able to handle any segment of a document; whatever your documents contain in the way of multi-objects and hypertext, mail should be able to handle both. This has always been a problem because our mail deals with structural addressable objects, which are quite different from the linear text sequential asking, files the rest of the world uses as mail objects.
George: You mentioned ‘addressability.’ I understand that’s a central concept for you, and probably many of our readers aren’t familiar with it. Would you elaborate?
Doug: Can we assume they understand what I mean by hypertext?
George: I don’t think so.
Doug: Well, it’s an outside term for me. I’m happy enough to adopt it, but I’m a little leery of using it to apply to our work until I understand what other people mean by it.
Over the weekend I was playing with HyperCard and yes, you can link whatever you want to, but the user never knows anything about the address. I was glad to see that the cards are given numbers, because it seems to me that every object has to have an address with some text or symbol that you can transmit to somebody as an address of any object within your document.
Every object in the document should be addressable in that sense: if it is, there is considerable value to users if they’re also aware of the addresses. But if you make them implicit and subterranean, as I gather some systems do, you have to use the cursor to point to the object you want to link, or plant a link in one place in your file that points to some other place. The user is never aware of any real addressability; it’s something below the surface that the computer keeps track of. I’m assuming that’s what most people will say is hypertext.
To me, if you make the address explicit, you are going to gain value. A link is something that you should be able to pass around. If I have a link in one of my documents that points to an object in another document, I should be able to pass that link to you in the mail, to say here’s a link to something useful. That’s a requirement.
George: I’m interested in your personal experience of developing your own augmentation system. What motivated you to do what you’ve been doing these past three decades?
Doug: I was going through a very heavy personal exercise between December of ’50 and February of ’51, trying to formulate a long-term professional goal that would maximize my contribution to humanity. I followed many, many leads thinking about it, even considered changing my profession. I was almost 26. One of the things I began to realize, that turned out to be absolutely critical to my making this commitment, was that the problems of mankind are getting both more complex and more urgent very rapidly. The product of these two factors is hard for humans to deal with. I wanted to see what I could do to maximize what mankind has to cope with this. That’s what I meant by saying we need to augment.
George: Do you think we’ve gotten any closer to a solution to the complexity/urgency challenge?
Doug: Sure…. I think electronic mail itself helps an organization. The shared-screen things we’ve done Ñ they’re going to make a huge difference. Bringing in hypertext, with all its cross-linkage, is going to make a real difference. All of these are expanding our ways of thinking and formulating and portraying our concepts and arguments and thoughts, and organizing them so they are more easily studied and manipulated and communicated and argued about. All these things are going to make a big difference. And there are lots more developments that will start pouring forth when people really tackle trying to boost their capabilities and stop just trying to automate what they used to do.
George: Where do you see the next frontier in knowledge work?
Doug: I think the interest emerging in hypertext is really significant, that’s part of the frontier. The way I conceptualize it is asking what can one do for the individual. Starting with being able to form concepts and manipulate them, the individual then symbolizes these concepts so he can abstract them and actually develop a verbal language to think and communicate. And then one externalizes the symbols that represent these concepts and their relationships. These were tremendous steps in the past in our effort to expand human capability. Then I look at all these computers and say geez! It’s just another way to externalize and manipulate symbols.
Then you realize that instead of externalizing symbols on pieces of paper in the linear sense, you can produce any structural relationship you want between the kernels and symbols in the computer memory. If we can learn how, we could make a very accurate map in the computer memory of this structure of concepts in one’s mind. My next thought was that it would be almost hopeless to read, it would get so confusing. But the computer could be at your service in getting the view of the structure at any one time that suits your needs of the moment. If you don’t have to look at all the detail, you can look at different portrayals of it. This concept of viewing and viewing control is very basic, along with the concept of structure. Then too you want addressability, so you can add arbitrary linkages and citations and use addresses and jumping commands and manipulation commands on the structure.
George: Since there are a number of hypertext systems on the market and as you said, an increasing interest in them, can you add anything else about what features people should look for?
Doug: I gather that people are looking for different viewing forms. A lot of them are just various views of the web, to see how the relationships hang together. That is one specific mode of viewing and I guess there will be many other forms of viewing people will find useful. I found it blows your mind if you drift very far away from a map that you can hold in your mind, especially with really arbitrary ways to connect lots of things together. It is very important to have flexible view control. Out of that will probably emerge structural conventions people will need to learn to get a mental map, and methods and procedures to follow when they’re constructing, manipulating, or studying webs of hypertext documents.
George: Thank you Doug for your time. This was very illuminating!