Douglas Engelbart

Interview 3, March 4, 1987

ENGELBART: I got invited to go back and talk in the ASIS meeting in Austin. Of course by the time I looked at my mail and have an answer, it's probably too late. I think they are going to have a session--the 30th anniversary, I think, of the ASIS (American Society for Information Science). Somebody had gone back and found something called the "World Brain." Science fiction. H.G. Wells had written something about the world brain and they wanted to make a session on that. Do many professional library people like you guys participate with ASIS?

Lowood: Everybody has their own slice of the pie; the ones that tend to be on the information side do.

ENGELBART: That's what I gave my very first paper on. It was then called the American Documentation Institute or something like that.

Lowood: Yes, that's right. It's impossible to find the early journal. Okay, where we left off, we got up to 1964. We got up to where you'd finished your first year on the augmentation project and it hadn't been a very successful year. You had just gotten funding from Bob Taylor at NASA, and continuation from ARPA, you had the whole thing with the computer in Santa Monica, and you were about to start again from scratch on what would become the NLS Augment system. So starting there, in 1964, I guess my first question is: at that time, what was the nature of the commitments from NASA, ARPA and I guess you were also starting to get money from ROME Air Development Center (RADC) somewhere around that time? How did these different funding agencies differ in their expectations? Did they themselves collaborate in any way? Did you have a sort of coherent base from which to start?

ENGELBART: At that time I think there was something like $80,000 from NASA and $60-80,000 from ARPA. The main thing was that was enough to let us get started that year and kept us from extinction. It wasn't the Rolm air development center at that time; it was a small office in what was called the Electronic Systems Division of the Air Force, which was a division that was part of the group that was responsible for getting in and installing radar and those kinds of electronic systems. There was a young second lieutenant there who had been in ROTC and was serving his couple of years, just out of university. He'd somehow found a copy of the '62 proposal or report, and he got turned on about it, so he persuaded his major that they could put some money into this. But it ran into trouble in a year or so because the second lieutenant called me up and said, "Well, I've got it all sewn up now for continuing but I'm finishing my term of service, so I'm leaving and the major is taking over." When the major took over, the lieutenant hadn't allowed enough time for things to be worked through and the major found himself in some kind of a crunch. He looked to find out what we had been doing. He was in a world where they were buying things, you see, not investing, and he got all pissed off. He wasn't very talkable-to, and felt I was trying to gyp him or something. It got really bad and what ended up happening was that the people at SRI fired a guy that had been working with me, to placate the Air Force. I still see him once in a while around here. He found other jobs and moved around, but somehow that left a bad taste. For the other two, the NASA and the ARPA contracts, we'd focused right on the central part of getting a minicomputer which had much less power than the first Apple II, I think, and cost lots of money. It probably cost $170,000--that's what it would have cost to buy the machine and its peripherals. It's just amazing. Slow and very awkward to program. You had to use paper tape. If you wanted to store anything between the times you turned the machine off, you had to punch it out on paper tape and then read it back in to fill the memory. So then if you were working with a document file and part ways through your program crashed, you had to load it in again by paper tape and start over. The only way you could get back-up was to punch out a paper tape.

Adams: This was narrow tape? When you say tape, I'm thinking 3/4".

ENGELBART: Yeah, it's about 3/4" wide. I think there were seven or eight holes across. It was coded so you could read it.

Adams: I vaguely remember using something like that in an office. You got to the point where you could actually read it like Braille.

ENGELBART: Yeah. It's almost like the same code.

Lowood: How did that limitation affect you? That seems like an awfully strict limitation on the kind of project that you were trying to pursue. You were concerned at that time with user interfaces, and here you are dealing with technology that is hardly developed.

ENGELBART: No kidding. Well, it's like if you are trying to build an airplane for the first time and it's hard to find a landing field and you don't have any way of getting gas into the tank and you have to pick it up with gallon cans or something. It's just part of the very rough environment as you are trying to stake out some area and learn about it. All those things, through the years - every time something better would be available, for input/output storage or something, everyone would be so thrilled to add that to the system, things that everyone takes for granted now. Each step of the way was a kind of an exciting step.

Lowood: Then both ARPA and NASA were, at that time at least, in 1964, fairly clear that they were funding exploratory work that would not have a direct application.

ENGELBART: Yeah. It went through a big, shaky review with NASA about '66 or '67. What triggered that was that we had made some movies showing the kind of fast interaction and editing on the cathode ray tube, and I had gone to one of the periodic--I'm not sure, they were semi-annual events that became annual meetings--when the ARPA office would get together its principal investigators. This particular meeting--it was probably `65 or `66--was at MIT, and they were fairly informal. We were all sitting down together and for some reason Bob Taylor, who was then running the office, turned and said, "Well, Doug, why don't you start by telling us what you are doing?" Most of their really "significant" work was a timesharing group, and an AI group people were also in there. I was a funny tag-along. So I guess he thought he'd see what I had to say and warm the place up. I had these movies so I said, "I have some movies." So they ran the movies. It really made a big impact because up to that time everybody thought that timesharing was all interacting from typewriter terminals or some displays, but they really weren't thinking about really working fast at the kind of things where you wanted something on the screen changed and it just changed. It really made an impact, which was fortunate. That evening, when we were sitting around the lounge, Taylor said something like, "The trouble with you, Doug, is that you don't think big enough," or something like that. My jaw dropped. He said, "What would you really want to do?" I said, "Get a timesharing system so that we can have a lab or we could build it and use it ourselves and evolve it from there." He said, "Well, let's write a proposal." So that's when we wrote the proposal for getting it. They were just emerging; the first practical timesharing system was one that came out of Project Genie, it was called, at Berkeley, with Butler Lampson and Peter Deutsch. I've seen them periodically through the years. I can't remember the name of the third guy. There was a small computer company down in the Santa Monica area making a commercial computer, the "930". They happened to have one up there and they figured out a way to extend its electronics a bit and make it a very effective timesharing system. So the company decided they would make a commercial version they'd call the "940". It was actually the very first timesharing thing that was available. It's essentially what Tom O'Rourke got hold of to start his
timesharing company that grew to be pretty big. So anyway, we made a proposal that we would get one of those. ARPA needed to have another organization through which it could channel the money for that kind of an acquisition, so they assumed NASA would do it. We talked to NASA people and the boss of the man who had been interacting with us wasn't sure about this at all because timesharing, he thought, was a fad. This was down in Virginia at NASA, at Langley Research Center. So he would pound the table and talk about these "goddamned prima donna university people who are trying to tell the practical people in the world what's going to happen. So alright," he says, "you come down and make a presentation to my people." So three of us went down there, trying to do it. It was just great. I took Bill English and Jeff Rulifson, both of whom are very competent people. It was just great the way they answered all the questions and got everybody interested and turned on. So the boss finally decided, "Well, alright." That was probably '66.

Lowood: You mentioned Bob Taylor. As I figure it, there were four ARPA directors that you worked with: Licklider, Sutherland, Taylor and Roberts. From interactions with them, can you contrast the way ARPA worked under those people and particularly how that affected your operation?

ENGELBART: All of them could very much just go by instinct and they would each come over a particular line and push it. But they didn't reach over the shoulders of the people to interfere, but stimulated more than pushed. Things changed immensely when Roberts left and some civil service people took over. There was a period there where--I am hesitant about naming names--but it was very hard. There were opinionated people who you couldn't get to listen, and who wanted to design the programs themselves and then allocate jobs out to people, rather than letting the researchers, who were professionals at it, do the thing. Support for us just fell apart very rapidly. In fact, three months after Roberts left we got a communiqué saying, "At your anniversary in January what we are planning to do is just to terminate, to drop you entirely." They changed that and came back and gave us various assignments for the next year or so, but it totally pulled away from supporting anything about the spirit of Augmentation.

Adams: Your funding was then just year by year?

ENGELBART: Yeah. And it was to be part of a program that they had formulated themselves. They had assigned different roles in the program to different other people. We played a support role. It was very, very hard because, as it has turned out since in watching the way that the issues about the architecture and approach to things has gone, we had for a long time been on a path that was right dead-center. They had in fact canceled part of our program after about a year because it was too complicated. But it was the one that is now appearing dead-center. It was very hard.

Lowood: When was ARC itself, the Augmentation Research Center, founded?

ENGELBART: It started out as a project we had. We were a project within another laboratory, and the project would have various names, usually acronyms having to do with the sponsors. So it was called the ARNAS project for a while because ARPA and NASA were sponsoring it. I can't remember when, but at some point in there they said, "You are big enough, we'll make you a laboratory." I think at that time I actually called it the Augmented Human Intellect Research Center. A big long name like that, AHIRC. Then I think it was about `68 or `69 that somebody suggested, very embarrassed, "What if we just changed it to Augmentation Research Center?" I said, "Oh, that's fine." Everybody thought I was going to put up a big negative fuss, or something.

Adams: Did you like the image of an ark, I know it's spelled differently.

ENGELBART: It turned out in 1969 I started to grow a beard. I don't know how big a cubit is (laughs).

Lowood: So they fit it into the SRI research organization, and it actually became equivalent to the other laboratories at SRI.

ENGELBART: Right. There are four to six laboratories, generally, in a division and divisions report to a vice-president and there are two or three vice presidents, each with some divisions under them. So that's the hierarchy. Then often within a laboratory they have other kinds of unit breakdowns. Most of the laboratories were job shops. They were working on a certain kind of electronics, a computer thing or something like that, and different members of their staff would be at different times guiding proposals and getting projects of various sizes and doing them. None of them was at all coordinated into one coherent program that stayed long with a consistent strategic direction, year after year. It made it very difficult.

Adams: Did you interact with the other labs, working jointly on projects under some sort of guidance?

ENGELBART: No, there wasn't that kind of coordinating guidance above the lab level. We tried several times, but it was disastrous, because the perception of what we were trying to do was just not communicable.

Lowood: Let me ask that question on a different line. Let's say that there were different groups that we can identify: there were the people in the project, there was SRI management, there were the sponsors of the project and eventually, at some point, there were, I suppose, clients involved with the product, and there were research assistants both outside SRI who in some form were looking at it. Now those are all different groups with possibly different expectations. What kind of interactions were there between those groups in terms of the project? For example, there must have been interactions between you folks in ARC and the SRI management and as well with the sponsors, and in a sense those must have been negotiated to define what was going on. How did that all work out?

ENGELBART: Those are very critical questions which in some sense would bear a lot of examination, because someplace in there is a better explanation than I can pull together about why the collapse and exodus (I guess that's as good a term as any). Anyway, the general way it goes within SRI is if some kind of researcher could wangle a proposal and get a project from someone, the main thing the management is concerned about is if it was something that looked to be to SRI's credit and wouldn't disgrace them and wasn't illegal and all that, and was not going to run into red; they pretty much didn't go beyond that. So they didn't get involved in any kind of assessment at the science level, other than if it looked like it was going to be bad science, somebody might complain and they might not want to do that. I don't remember any cases of whistles being blown, at least not very loudly.

Adams: How did the projects report to that level of management?

ENGELBART: The reporting was essentially all just performance and dollars flow.

Lowood: It was just a superstructure that existed there and kept things together and helped you find people and kept them paid and so forth, but beyond that it wasn't...?

Adams: If you were satisfying the sponsor, you could do what you wanted?

ENGELBART: There was no way to appeal. When they began to get the feeling that what we had been doing was bad science, or something, they had no way in themselves to evaluate it. They would just go ask people. We didn't have much in common with the other groups around. Every once in a while someone would get interested in some kind of interactive thing and sometimes they would approach us to see if there was something we could do in common, but generally, they had such a specific narrow sort of thing they wanted to do that they weren't of any value to us.

Adams: You couldn't bootstrap on each other?

ENGELBART: Yeah. They just were so different. A lot of times they would feel like, well, we ought to give them access to our timesharing computer and all of that. There were just some really disturbing cycles in the early days. For instance, when we were getting enough money, say in `64, `65, to move from this little tiny minicomputer to a much better CDC one, they told us "You'll have to get the support for the whole thing." So we figured out the configuration we could get, and we finally wangled enough money to do the whole thing and we got it. We built an interactive terminal on it and started to do our work. As soon as that terminal got there, there turned out to be a lot of people who wanted to do some experiments with it. So the Institute started giving them time and I'd say, "Well, we got it and we did all the work to get it and we had to guarantee it, so I don't think it's fair for us to get pushed into smaller and smaller amounts of time for these people." They would just act as though I wasn't being fair and wouldn't give them a reasonable chance. I said, "Well, they get a lot more chance than I had to start with." In fact, some of them wanted more memory, so they boosted the configuration of the machine and upped the rates that they were cHRGing and pretty soon, instead of having a machine to myself, I had it about thirteen hours a week, and with one terminal, the only way to do interactive things was to go in and sign up.

Adams: And there was no trade-off that they were giving you for that?

ENGELBART: No. The budget that I had to run it was buying me less and less time. It wasn't a timesharing machine, it was just one facility. So when Taylor offered me a timesharing one, I insisted that it wasn't an Institute facility, it was my lab's facility. I built a wall around it, because people would just keep moving in. They did not understand at all the idea that you have to use it and you have to be there all the time and if other people are playing on that machine that means that your responsiveness goes down. So, the other people around didn't have experiences with all of that yet, and management didn't. I was always the first one that got the machine and the first one to get the timesharing, and caused all these special troubles. What they remembered later were the troubles, even though providing other researchers with access became a matter of course later. There was this sort of negative aura about stubborn, uncooperative me and my lab. I just got a hopeless feeling about trying to communicate to any of them what was different in what I was trying to do.

Lowood: So you were really operating under contradictory expectations; on the one hand, you were to be entrepreneurial in getting these things and yet on the other hand, you were to work for the good of the common institute.

ENGELBART: I really did for quite a while, but they just crowded me off. It's funny, they would just not believe me when I'd tell them what it was doing from our point of view. I wasn't used to having people disbelieve me in a matter of what I see and feel.

Adams: Could one appeal to one's sponsor that the project that they were funding was running into trouble because certain conditions were intolerable?

ENGELBART: I could have, but I wasn't aware of that option at that point. They certainly listened when I told them that I did not want them to pay the Institute to set up a facility and lease time to my project as an Institute facility. I said, "That won't work, I've learned that already." So they went along and said okay.

Adams: You negotiated some kind of contractual situation?

ENGELBART: Yes. That drags up all kinds of indigestion-producing things. Sure, that's part of the history. One thing that would be really interesting would be to go back sometime--I couldn't do it, but other people could--and talk to other people around and look at the records. I was immersed in my own dream about things. I was really naive about a lot of the world, about management issues and problems that they had to face. I really didn't pay enough attention to the communications and to the basic everyday politics of trying to make sure that people understood or that you were putting on a good image. There was a lot, I'm sure, that I could have done to forestall and improve a lot of situations. It would be interesting, anyway, just to look at that.

Lowood: Can you point to somebody who you think was, perhaps, an ideal manager within that system? Or maybe, in an ideal sense, not to name someone, what was the right way to navigate all of those different forces?

Adams: You mean a project manager.

Lowood: Yes. Obviously there were people who were successful at working within that system. What qualities did they have?

Adams: Or qualities their research had.

ENGELBART: The whole structure of what they were doing was so different that right at the outset it makes it very hard to compare. Here I had this long-term perception and dream and went on to pursue it, and my picture was that I was getting the money and I was hiring people who would help me pursue it. There wasn't anything similar.

Adams: Did you have production benchmarks or project benchmarks similar to other projects? By date so-and-so. . . . ?

ENGELBART: A few of them later were getting their own computers and had to get those in, like the AI lab, which was getting its own computers. They would be doing various kinds of projects on there, but they didn't all tie together by any means. One would be studying natural language. Another would be studying how to drive a robot. Each of the people doing those separate things pretty much did what they wanted to do with the money they were getting. Being the manager of a laboratory like that was pretty much more of a question of seeing that they kept busy and that they got the proposals out. One of the problems in a world like that is that right in the middle of a project you're working very hard on the project and there's no time to do a proposal about the next thing you're going to do. Then you get near the end of a project and it is so late that there is going to be a gap of no income. So a lot of managers have got to see that people keep on that, getting the income and trying to build up his laboratory. One success factor is maybe publications, but not nearly to the extent that it is for a university. It's mostly the financial operations in the laboratory, whether it's growing or not.

Lowood: You just described a dual role--the head of the laboratory being the conceptualizer, as it were in your case, but also the manager. At least in the first year, I know that you were completely shunted out of the project, but after `64, were you always in cHRGe of both?


Lowood: Were there any attempts to keep you in cHRGe of the intellectual direction of the project but to give you some sort of support on the management end?

ENGELBART: Probably from mid `74 to `76, I was trying to get them to do that. There was one division head that we had that was very empathetic. They couldn't seem to convince anybody above him, or do much that changed the environment, but he was an empathetic listener. Suddenly one day he and four other division heads just got fired because the president was tired of trying to get them all to spend more time on the bottom line. So one day he just fired a bunch of division heads. We got a new one. The new one was the one that six months later told me, "Well, my solution is to replace you." Because up to that point I had been trying to get memos written to say that it was really too much stress on me at that time. The classical thing they tell you is, "Well, if someone else were the manager, then he has to be the boss and he going to make the decisions."

Adams: So they really saw the management on top and then the conceptual researcher below that.


Adams: Which is a reverse of how you would image that.

ENGELBART: Right. I would say, "I just can't picture why you require that I have to do both." They just said, "Management never does anything like that. If you are the manager, you are the boss." I would say, "Come on! The captain of a ship has an executive who takes care of how things run and is responsible for the management. There are all kinds of examples like that." So they finally said that they wanted to talk to the guy who headed the management sciences part of a management consulting outfit, and he picked one guy to come who was supposed to make a study and try to produce a plan. This guy was young and kind of green. He got all turned on by what we were trying to do. I'm not so sure about whether his personality didn't do us more damage--but he just didn't convince anybody. So then the new guy took over. He told the lab managers in a meeting, "I'm going to spend at least an hour or two a week with each of you until I really learn what your business is, and then we will start making some general plans." In the next six months he spent exactly half an hour with me. I should have just really, really been knocking on his door every week to get going on it. I was pretty busy and then our fire at home happened, and then he called me in his office and said that he had decided that he should derive the solution and not have me telling him what the solution was. His solution was to replace me. Bingo. He never had the slightest understanding of what we did and the guy he put in cHRGe didn't either. People just started leaving us.

SRI Organization vs. Sponsorship

Lowood: That's the crucial difference, really, between SRI's research institution and a university, where someone can continue their research individually. I suppose it doesn't require a lot of support from the institution; you would have protection against this sort of thing. Within SRI there was nothing like that, was there? There are no guidelines in terms of guaranteeing that.

Adams: You bring the project in to them, which enhances their reputation....

ENGELBART: Oh yes. I had brought millions of dollars in.

Adams: You got the sponsorship, and it's not your project in the sense that they could remove the principal investigator from it?

ENGELBART: Certainly, they feel like they are responsible to the sponsors for things like that. If they feel the person running it is not competent enough, they consider it their job to replace him, and generally that's right.

Adams: Was it your option to pull the project away and find another location for it?

ENGELBART: If I had gone and moved to some other place and if I could have talked the sponsor into it, sure.

Adams: Was there an environment that you could have gone to?

ENGELBART: No. I looked around. At that time the sponsor had turned very negative too and this made it bad. Looking around the Institute, there was just no one. I had some good friends who were very concerned personally, but who didn't seem to know what to do. It was very different from the way it is today, in people's perception about what interactive computer support is and can do. In those days a lot of people were getting interested in it and their perceptions would look very primitive today. They looked very primitive to us at the time, with the experience we had. So if the SRI managers asked different people about it, they would get very uniform answers about, "Well, they're off on cloud 9 over there at ARC, it's so complex and baroque, and yes, they pioneered some interesting things in the 60's but they have just not been moving at all fast and they are not up with what everyone else is thinking these days."

Lowood: Before we get to the technical problem of ARC, there's one other question on the managerial side of SRI, which is their patent and licensing policy. That's another aspect of income for SRI. We'll talk about the mouse in a few minutes, but in terms of the mouse or any other device that was patentable and could bring income in, was there a split between you and SRI or did SRI get all of the income? How did that work?

ENGELBART: When I went to work there was a standard form you signed that since they were giving you steady employment, etc. and paying you to do research, that any products of that research belong to them, or to the sponsor. That was standard. Sometime, a year or so after I left, as a matter of fact, after I was fired--I actually was fired--they upgraded that to give the patent person a share. It was graduated depending upon how much income they got from it, so that I would have gotten hundreds of thousands of dollars from the share. As it was, sometime four years ago they invited me and Ballard [Mrs. Engelbart] over for lunch and talked to the president and the patent attorney and some of my old friends, and they gave me a check for $10,000. I was just dumbfounded. We got curious and we checked with patent attorneys later and they said, "Well, if they hadn't given that check to you, you could probably go back and sue under all those conditions. They just did a very neat move."

Adams: Co-opting you.

ENGELBART: That's just business. The funny thing about it is at the time that they were splitting us off, the Tymshare people buying it, we were telling the Tymshare vice president who was doing the negotiating, "You ought to look at that mouse patent. You could probably pick it up for very little. It might very well be worth something." He didn't think it was worth even bothering about. The fact is I probably could have personally bought it from SRI for $5000 or something. That was long before they were getting any money for it.

Adams: I don't want to get us out of your chronology, but you mentioned a fire a couple of times. Is that something we should discuss, or was it personal, not a loss of records or research materials?

ENGELBART: Mostly personal belongings; not very many research notes.

Lowood: Let's go back to the Framework and to the technical work now at this point. Could you go back to 1962 and your report, "Augmenting the Human Intellect" and tell us what you think the core elements of the Framework were at that time?

ENGELBART: OK. Intuitively, I had a picture of all the things that a computer could do for you, with very fast interaction, and what the cathode ray tube display screen could do, and the fact that on a display screen you could have whatever symbols you want, all working rapidly. The computer could keep track of whatever movements you wanted it to keep track of that you were making, and translate those into explicit things you wanted to do; it involved the whole business about the interface, about how rapidly you could probably learn to steer the computer to do the things you wanted to and about new linguistic forms and all of that. The collaboration between people. So I was trying to make a Framework that somehow could make coherent sense of how you would approach going after that and, if you think there's a lot of value, how you could you picture it being based upon some framework. I remember the revelation to me when I was saying, "Let's look at all the other things that probably are out there in the form of tools," and pretty soon focusing on language; realizing how much there was already that is added to our basic capability, and that a lot of it we absorb culturally and a lot of it is sort of trained explicitly in the schools. It amounts to an immense system that you essentially can say augments the basic human being. All I was talking about was bringing in more. It was a real letdown, suddenly, to realize we're just bringing in more in top of something that represents such an immense amount of an already very ingenious invention. I am really especially awed by language.

Lowood: So because of the computer tools alongside all of these cultural tools that existed already--is that why you chose the word augment? The fact that you're adding onto a finite base?

ENGELBART: Yes. You're just augmenting the basic human capability; there already is a fantastic system. We have to augment the basic human capability and the computer was just another artifact. So that really jolted me. But then I began to realize the unusual characteristics that the computer and communications technology were offering, in just plain speed and quantity. I had done enough work on scaling effects to realize that the whole qualitative nature of some phenomenon can change if you start changing the scale of some part of it. I began to realize in how many ways, and how directly, the computer could interact with the different capabilities that we've already got. It began to dawn on me--really a clear picture for me--that the accumulation of all those ways would make a big impact. And a very large thing that came out of that--probably the thing that made the biggest difference in my perspective from what other people were looking at--was the realization that, to really go after the value that was there, you needed to look at all the candidate changes in the existing human system and all the cultural things. They weren't there because that's the God-given way we should think, talk and do things; they evolved co-evolutionary-wise with all of the rest of the stuff going on. So let's just re-examine it all. It's all a candidate. That got me to say, "Alright, what are some of the ways we can start? Here's the explicit structuring you can do and here's the document model out there and the viewing and the moving around. Let's think about the ways to operate and explore faster ways to control it, even though they seem very unusual. But anything goes if you're looking for where the value's going to come from." So it turned out to be exploring how you get the value, assuming that when you really learned, then you could tailor things for the actual cost and pay-off mix that any given application would require. Costs might mean the costs of learning; fine, you don't try to use the more exotic control interfaces so that people don't have to pay so much to learn. That sense of exploring and not worrying whether it seemed unusual made a big difference.

Lowood: Did you use the word automation at all or did you steer away from the word?

ENGELBART: I considered it way back and it didn't seem at all right to me. I wasn't trying to automate the way we do things now.

Adams: Did you have a model for the qualitative change in decision-making, thinking or learning that your system would cause? Was that in the back of your head or did you believe that you would perceive that model once you got the tools working?

ENGELBART: The model in my head about how you go after this was that there would be no way that we, from our current framework, could guess what was going to be the best way to do it. The best thing we could do would be to find a good evolutionary strategy for the co-evolution that you have to do. That was there all the time. I remember making this diagram in the back of the report, a flow diagram of the different stages and the idea of building in the bootstrapping. It was something we were starting; there would be generations of pursuit before it probably would start to level off. Just learning how to harness, to integrate all this technology into our way of thinking and working, would take a long, long time. But some paths that you would follow would get to the benefits and effectiveness much sooner than others. The strategy was important to me. These are things that made so much difference. I'd say, "I'm starting to work with documents, with the support of expositional communication and thinking and developing an argument that can be communicated, and with integrating lots of other people's thoughts and considerations." That would be right at the heart of how you learn from experience and integrate it. And I had to consider all the different kind of stake-holders.

Adams: Was that the first leap into that sort of interactive community? How would you articulate the first leap into a new vision of how the thinking was evolving and how your tools might be changing the way of thinking? As you were going through your little flow chart, when did you say, "Aha! I now see where this is evolving. I have now created the first air breathing creature?"

ENGELBART: That's a good question.

Lowood: I have a question. I was sort of thinking the same thing. I'll turn it a little bit around and maybe make it a little bit easier to answer. I was surprised that some of the specific inventions that came out of the project, the things that you described as very meta-level, as far as the project goes, compared to your broad conception of Framework. Yet some of these very specific things, like the chord key set, I think was '63 and the mouse was `65, actually came amazingly early in the project. How did you get from the Framework to the specific products that came out so early in the project? Why did you, for example, go so quickly to the chord key set and to the mouse?

ENGELBART: Well, I actually practically described them in the '62 paper. I just knew that there were quite a few different levels and that for one thing, if the machine could respond and do the things, then you really needed to give consideration to how you were going to control it. These were just two fairly simple things to start with. I remember thinking consciously, "If I could picture some of the dimensions of the system that are likely to be new, then it would be important to find a way to get started along that dimension." Because then evolution is quite easy, with natural staging, rather than moving to do something radical. So anyway, that was just one step, such as the structured files, to get started in thinking about putting explicit structure into the computer-held representation, trying to make a map of the structure of your concepts out there. There would be a lot to learn about it. Let's just start with this hierarchical thing with the links. I figured, as I was laying that out one Saturday that in two or three years, we'll be moving on to other stuff. I was thinking of more complex structures. Then I thought, "No, let's keep it simple to start." So the mouse and the keyset worked similarly, starting the dimension of control.

Lowood: I understand your logic as far as structured files and such go; I can see how that maps directly from your framework, and in fact you talk about structured information in a general sense of the word. But I don't quite see the path as clearly on the pointing devices and the control over the computer, how that was, at that early point of the project, already a high priority. How did you get to the point where the actual control over the structure was a priority?

ENGELBART: It wasn't a very high priority.

Adams: It was a mechanistic tool?

ENGELBART: Well, one thing was that we were still getting our money from NASA and they wanted something that wasn't just melded into the rest of it. So I said, "Well, let's go after some screen-select devices, that's a good project." Bill English wanted a project, so we went after that.

Lowood: Was that actually Bill English's project then?

ENGELBART: Yes, to organize that. He and a couple of other people made the test set-up. Then I was saying, "Here's one of the devices we could pick. I want it to be in context, so that you're making the selection in a context that we'd be thinking and working with, where you assume that the keyboard is still an important part of it. At the outset it's text. So that steered us into the kinds of things to select. I wanted it so that it wasn't just how fast you could find a spot and get there--this includes accessing a device, pointing, the errors you could make--not just doing it fast.

Adams: There's also a visual sense. If the computer compartmentalizes things in a linear way--you're accessing with a keyboard, text that follows and precedes--the mouse builds on that and gives a visual, a different kind of thought process. I don't know whether it's non linear, necessarily, but I see information stacked on top in three dimensional ways rather than linearly. What would have been the next step, post mouse, or what was the next step?

Lowood: Let's not go past the mouse just yet, because I still want to explore a little more of what you were just saying about the mouse. You said that you had this sort of initiative, perhaps from NASA and were considering a number of different technologies at that point. Did you have something else in mind when the mouse and the chord set were designed in the laboratory or were they really created out of whole cloth?

ENGELBART: Oh yes. With the mouse, we were trying to create something different than the available cursor control like the tracking ball and some kind of odd things that people had, and the light pen. These tablets that people have made now, the tablet is sort of alive and you put a stylus on it and the computer senses it and can control the cursor. Well, the very first of those was being developed at RAND Corporation and we thought that they could probably loan us one. But they said, "Well, we don't have than many." We said, "Alright." For some reason the analogy was the periodic table in chemistry. You have to be able to think of different kinds of approaches. I carried little notebooks around for years. I just have piles of them. Sometime I ought to go and find them all. They're in boxes. They were carried in my front shirt pocket and I would get thinking about things and so I'd have a place to write them. They'll make an interesting collection. When I was thinking about that I said, "Aha, I remember a couple of years ago thinking about all that at a conference." So I dug back into my notepad. Here were just a page or so of a description of what was the mouse. I remember sitting at some graphics conference and just feeling at a wall because everybody was talking and I'm not skillful at all in getting them to listen to me. So a lot of times out of frustration I'd start talking to myself. I remember thinking, "Oh, how would you control a cursor in different ways?" I remember how my head went back to a device called a planimeter that engineering uses. It's a little simple mechanical thing that has a bent arm and the elbow of the arm has a little disc that rides on it and this little disc is out here and you start out following some closed path and when you are all done, you can read the two discs and do something about it and calculate the actual area that's included inside. I saw that used when I was a senior and I was fascinated. One instructor took time enough to tell about the characteristics of the little wheels. If you have a little wheel with a shaft and you push it along like that it doesn't rotate and you push it along like that and it does rotate, and if you say, "I want to go from here to here" and you push it this far and it rotates and this far and it essentially doesn't matter what path you follow between those two spots. It will roll the same no matter what. It rolls only as much as the component of its rolling direction.


ENGELBART: When I first read about those I said, "Oh, I already know about that and that's how they do that.

Lowood: Vannevar Bush was interested in that device when he was designing his first differential analyzer at M.I.T. in the 1930s.

ENGELBART: Yes, it was mechanical.

Lowood: There are pictures of him pushing it around the countryside. It is fascinating.

ENGELBART: Well, anyway, just thinking about those two wheels, soon the rest of it was very simple, so I went and got that and made a sketch. It worked right the first time. So it was just added to the rest of the devices. We have a reprint here of the paper it was written in and about the tests and experiments. The paper was really talking about the screen selection testing. You know, I can say, "This thing we built, called the mouse--it just happened." Once you had your hand on any of them, they were all accurate, with a little bit of skill, at picking something. It would sit there and be where you left it. You didn't have to pick it up and you could put buttons on it, which helped.

Lowood: Was it simple to go from the mouse to the screen at that time?

ENGELBART: Oh yeah. The thing of taking an analog voltage and converting it to digital has been a basic instrumentation device from way, way back; that's all that it was.

Lowood: And the software aspects of it were also fairly trivial in your mind, as far as integrating it in your mind with the rest of the system?

ENGELBART: Yes. It was essentially no different from a tracking ball; in fact, when you start looking at it, it is like a tracking ball turned upside down. That was the way most mouses were built. I didn't arrive at thinking about it that way, but that would have been appropriate.

Lowood: So the Apple mouse actually operates on a different principle.

ENGELBART: Most of them put the ball in and then riding on the ball are the two little discs. They just do that so that the discs don't have to ride on the surface. I was never quite sure why they decided to do that.

Adams: Was it to protect the discs themselves during the tracking?

ENGELBART: Potentially.

Lowood: I think it might be cheaper.

Adams: Who came up with the term "mouse"?

ENGELBART: We're not sure. No one can remember. In the lab, the very first one we built had the chord coming out the back. It wasn't long before we realized that it would get in the way, and then we changed it to the front. But when it was trailing out the back like that, sitting there, just its funny little shape. [Demonstrates]

Adams: Is this the first mouse?

ENGELBART: No. I'm not sure where the first guy is right now.

Adams: Was there ever any thought of doing a remote device that didn't require linking with a cable, or was that beyond the technology?

ENGELBART: Oh yes. When the first guys talked about it I said, "You know what I'd do? I'd get a little chain and tie it on so nobody would steal it."

Lowood: There was in fact such a thing devised for the Apple.

ENGELBART: One guy that worked with this is the engineering vice president for a company called Metaphor, that puts out a very exotic workstation with very high power capabilities. The little mouse has a little infrared piece that's all by itself.

Adams: What generation is this one you are demonstrating?

ENGELBART: This is almost the direct copy of the shape and everything of the very first one.

Adams: It looks like a little telephone.

ENGELBART: There are wheels, inside. You get used to a certain size. This isn't bad.

Adams: Did you have an ergonomic tester designing where the buttons should be--raised, how responsive, etc.?

ENGELBART: People have asked me, "How did you decide on three buttons?" Well, it was all we could put on. That was all there was room for.

Lowood: Did you ever think of combining the handset and this?

Adams: With another set of keys?

ENGELBART: Oh, lots of times. With more keys; in fact we thought of actually make a ten key set. I had the design all worked out. These things worked really very well and got us moving. There were lots of things to invest in and I kept really looking at what was--the whole or part--making a better strategy for us to move ahead. We had gotten to where investing in these improvements got us diminishing returns relative to other areas where I felt the payoff was.

Adams: So a separate device seemed to be appropriate.

ENGELBART: This kind of decision on my part made for a certain amount of friction in the lab, where people would just love to go out and improve this, but it just didn't fit. At some time it might be a valuable thing to do, but I would try to make an analogy to people, "I'm trying to get to Antarctica to explore it, and I know these islands look very interesting and you'd like to stop. But we'd never get there." So anyway, it got pretty hard on some of the guys.

Lowood: I wanted to mention a few of the other aspects or features of NLS and maybe just let you talk about them, in free association terms or in terms of chronology or importance to the project or whatever occurs to you. And I guess one to start with would be hypertext and the whole notion of that kind of structured environment that you were trying to create. When did that actually begin to exist or take shape as a working system?

ENGELBART: Right from the start.

Lowood: Did you have, for example, the outlining, that kind of text processing, that kind of structured outlining capability in the system already in the mid-60s, or right from the start? You obviously had to have software people around to write the code for all of that, and surely they hadn't worked on something like that before. How did you go about creating the algorithms, getting the thing created?

ENGELBART: It was sort of like when the architect sits down with the structural engineer and says, "Well, I want to do it like this." "Oh, you can't do it that way!" "Well, what do you mean you can't do it that way?" So you go back and forth and around and around. I wanted a statement to be its own unit and the very first times we implemented it they were sequential files, like people use now. Then you can have some special character sequences that aren't going to show on the printer, or something. Let's say this is the start of a new statement and if you want the level of the statement in the hierarchy, for a while we were actually putting the statement number there, 1A5 or 1A7. It got hard to do some of these operations you wanted to do. When you wanted to move a whole branch, then you had to run down and figure out where the bottom of the branch was and move that and change all the numbers. But it was all fairly straightforward logic. It didn't require anything like inventing new mathematics.

Lowood: You would, in essence, take the framework--your overarching framework--down a level and say, "This is what I want to see on the screen."

ENGELBART: I could go down inside that and talk about how they would think about determining it, or discuss the logic of what we were trying to do, and I could say, "That's really slow" or "There must be an easier way."

Lowood: There's a book on the "Mythical Man Hour"; did you experience as the project got larger that it was harder to get from what you wanted it to do or that it took longer?

ENGELBART: Well, mostly, the difficulties in that were in personalities. I never was a professional programmer, but the concept about what you are trying to do, lay out, and all of that, somehow is quite easy to grasp. When I work with architects and builders I see that the architect may not know how the builder has to go out building it in detail, or whether it is structurally okay, but he depends on these other people to do it. But pretty soon he gets these instincts, about, "I'm pretty sure I can cantilever this much and I may have to put an awfully big beam in there like that, but here's what I'll do to get ready so if the guy has to put a big beam in I'll put the facade." Then he goes to the structural guy, "Quit complaining, just tell me." Sometimes you go to a structural guy and he says, "You don't want to do it that way." "Well why not?" "It's going to be too expensive." "I didn't ask you to make a design and then tell me how much it'll costs. But I did that all the time to people like that.

Lowood: What was the pool of professional programmers at that time?

Adams: Yes, how did you contact people?

ENGELBART: Well, that's a pretty wild thing, for a while it was hard, until '65 and '66 when we started getting some that were really accomplished. We had one very good detail guy who wasn't terribly imaginative, but really worked well. He did quite a bit of it in '65. But somebody else we hired for a while was just a disaster. Then we started to get guys fresh out of school who really had lots of experience and were going to graduate school and really wanted to learn about everything--a few people like that. So when you are recruiting people, you can size them up a lot about what their expertise is, and also, just talking to them, they can get other people more excited. It was hard until we started getting stuff on the screen that was really unique. I had written things in '65 or something about how if we used the same structure programming, how much a payoff there's going to be. It wasn't until we finally got to the point where these guys started putting in the source code that they said, "Ah!" They got so excited. What now is called structured programming--that term hadn't yet been invented--was just a way to work with the views. So by the time we got to that point and started bringing people in, there was an excitement and a difference that could start attracting better people. I'd always built a lot on my instincts about the guys' competency and not being a fake and just the way people talk--that affects me a lot.

Adams: Did you recruit people from the university directly?

ENGELBART: Oh, I don't know where they would show up from.

Adams: Word got out?

ENGELBART: Yes, somehow. People had friends and they would call. People would show up. We would always sort of find our own. But anyway, by the '70s we had a really good crew. The quality was very high. The thing about doing it, for me, is that there were some personalities who, the instant I would say something, would interpret it in a way that satisfied themselves that I was really dumb and didn't understand. I'd stop and say, "Just a goddamned minute." We'd find out that they didn't realize what they were talking about. They just didn't want me fooling around in their domain. It made it very hard. It was very hard to get some of the novel things. There are things that I gave up on year after year. We could go back through my notebooks. When you say, "How did you build a mouse?" For everything we did do there were dozens of things that I couldn't get people to do or find resources to use. To a large extent it was like if your crew was getting cranky and they didn't really believe you that there was an Antarctica. Well, the best thing you can do after a while is stop and let them chase around, if you can take the time to do it. I had been waiting for years to add some things that the old-timers in the group would joke about now, that I never got built. I'd keep trying to find time we could put it in and build it and they would laugh about it as though it was one of those crazy things. They would be very valuable things in my mind.

Adams: Would they have started a new evolutionary branch?

ENGELBART: To some extent, right.

Lowood: Can you remember some of the things?

ENGELBART: Oh yes. I was going to show you. In physics and engineering, where you are talking about vectors, it can be a force. Or many things can be reduced to a vector. It's a direction and a magnitude. So they have all kinds of operations you can do on a vector; one thing is that if you have two vectors, what's the dot product? It's sort of like this -- how far along your desired vector do you get if you take the other instead? I would say, "Well, if this is what I want to do and I have enough money, I can go there under the ideal situation and say, "Alright, the best I can convince the mixture of people to do might be something like this." So I'd get that much. I would sort of feel what percentage am I getting. A lot of times I would just settle for 30%. Then sometimes it could get up to 70 and 80%, and that was really great, and then back down to 20%. Oh boy! It has been about minus 5% for the last 10 years.

Lowood: Again, you're talking about a management problem, but now you're talking about you as the manager rather than being managed by someone else.

ENGELBART: It isn't simply just managing, it's that you're in the midst of what other people are turned on about, what they want to get out of their own work. They work very, very hard and intensely, so you realize that they need something out of it, to do it their way. On the other hand, you say, "Boy, I really need that." You can't constantly go to somebody and keep wrenching him over to do it your way. Pretty soon, it's too wearying. You wish he could perceive it the way you do or you could replace him easily. You could say, "Go and I'll get somebody else", but it would take three months or so to find a good replacement person, and at least a year before that person would start to really understand it well.

Adams: Yes, you want a competent technician who can do what you ask and find a way to do it, but also someone who is a creative thinker who can look around the problem and find a solution.

ENGELBART: Right. And nowadays, when Hypertext is hot, there would have been no trouble at all to get people to work on that. But in those days, their peers were using different criteria, had different interests.

Lowood: It might be good at this point, since you're talking about a perception of the product of ARC, to talk about the 1968 Computer Conference and the presentation you made there. Maybe you could just start and go over why that happened and what happened.

ENGELBART: I was beginning to feel that we could show a lot of dramatic things. We'd have visitors come by but with artificial intelligence, there are groups of people all over the country working on it. What do you do to get people going on augmentation kinds of things? Maybe what we needed to do was to show a lot of people at once. I got the picture of what we could potentially do. What equipment can do for you, how you can put it together, has always been easy for me to perceive conceptually. I started out in engineering because I was interested in a lot of that. So I could picture how we could put it on. I also had this adventurous sense of, "Well, let's try it, then." It fairly often ended in disaster. Anyway, I just tried it out. I found out that the conference was going to be in San Francisco, so it was something we could do. I made an appeal to the people who were organizing the program. It was fortunately quite a ways ahead. The conference would be in December and I started out sometime in March, or maybe earlier, which was a good thing because, boy, they were very hesitant about this. They sent people twice to a site visit. One time they were going to cancel it all because one of them had been out at Langley, and somebody had proudly shown them a system that could already do what we were talking about. I said, "God, that's our system." Since they had sponsored us, we kept them a copy of it, and they every once in a while showed people. (laughter) So they finally bought it.

Lowood: What organization was it?

ENGELBART: I guess it was AFIP, the American Federation of Information Processing. It was at one of their semi-annual major conferences. They had their fall joint conference-- joint meant between their different sponsoring computer societies. One was in the fall and one was in the spring. Okay, we could do it. Actually, it really never would have flown if it weren't for Bill English. Somehow he's in his element just to go arrange things. Pretty soon, we had video channels from the telephone company all arranged. They'll come and put up the roof and before you know it, there is a co-ax running down, and they'll be up in the skylight with four dishes on a truck. We needed this video projector, and I knew they had one too. I think that year we rented it from some outfit in New York. They had to fly it out and a man to run it. The telephone installers were putting the other in. Pretty soon, by the conference time, I went up there and everyone was swarming all over. They had to make some special equipment, and a guy did it for us. We soon got the cameras out and I was working on how to script it and talking to everybody about how it ran. We knew we could get the video controls so we actually bought them. They weren't terribly expensive. There were boxes that you run two videos in and you turn some knobs and you can fade one in and out. With another one you can have the video coming in and you can have a horizontal line that divides them or a vertical line or a corner, in switching. It was pretty easy to see we could make a control station that could run it. Bill had worked a lot as a stage manager or production manager for theatrical groups and he loved to do that, so he just made a very natural guy to sit there. He built a platform in the back with all this gear. The four different video signals came in and he would mix them and project them.

Lowood: Was that common, that sort of presentation?

ENGELBART: Absolutely not. No precedent we ever heard of.

Lowood: It was at least an hour or an hour and a half, and a lot of equipment required. What did they call it?

Adams: Yes, what was the program description?

ENGELBART: I don't know. It turns out that we didn't think about giving out special publicity. A lot of people said, "Oh God, we would have liked to have seen it!" I don't know how many people were there, but somebody has commented that there are more people who have claimed to have been there than were!

Adams: Well, that's a sign of its success.

ENGELBART: I keep being surprised when I run across people who really were there.

Adams: How many people were at the conference?

ENGELBART: I don't know. The auditorium was big enough that it probably could have held 2,000 to 3,000 people. I just can't remember how many filled it.

Adams: You were too busy to count the heads.

ENGELBART: Boy. And nervous as hell. Gee. So much swung on it and we had all this special technology to get it working in the first place. We had agreed that we would have some trial presentations captured on film, and that we would have that film standing by. But we just knew how the hell would we find our place in that film if everything crashed halfway into it? Every one we did was different.

Lowood: This was expensive; how did you do that?

ENGELBART: That was a big part of the gamble. I was pretty sure that we were getting money from NASA and ARPA. It was a time when you are just sort of on a good friends basis and you interact. How much should I tell them? I got far enough so that they got the idea of what I was trying to do and they were essentially telling me, "Maybe it's better that you don't tell us."(laughs) They could get in trouble if the thing crashed or if somebody really complained about it. We had a lot of research money going into it and I knew that if it really crashed or if somebody really complained, there could be enough trouble that it could blow the whole program; they would have to cut me off and black ball us because we had misused government research money. I really wanted to protect the sponsors, so I would say that they didn't know. So that's the tacit agreement we had between us. As a matter of fact, I think Bill English never did let me see how much it really cost. (laughs) But I know it was on the order of $10-15,000, which would be like $50,000 nowadays, or the equivalent. A lot of money.

Lowood: We know that a lot of people say they were there but probably weren't there, and we know the kind of impact the presentation is supposed to have had, looking retrospectively. Did you have any feedback immediately about it? How did you think it had gone?

ENGELBART: Some people came rushing up onto the stage, one in particular, Butler Lampson, who is a superbly intelligent guy, and at that time was at Berkeley, in '68. I don't think Xerox PARC started until the '70s and he moved down to the Park. But anyway, he was just so excited and that was something pretty great. So I knew that there was a lot of enthusiastic reception about it right there. But basically I really was hoping that it would get other people seriously started in things like this too, but it just didn't.

Lowood: Most people describing their reaction to it seem, I would say, to have been very impressed by what they saw, but saw it more as an example, than to be spurred in a completely different direction. Alan Kay, when he talked about it at the History of Computer Workstations meeting--did you know that he had been there? He talked about how he had been there and how it had been an inspiration, but he certainly went in a somewhat different direction. So it had kind of an odd effect, in a sense. It was a success, wasn't it, but it didn't really validate your project the way that you wanted it?

Adams: Or translate it into energy for your project.

ENGELBART: That's how I felt. How much the sponsors over the next eight years kept supporting us because of the presentation, I have no way of knowing. But everyone else was still using linear files for years and years. The ideas of links were beyond them. I'm still very puzzled why there was sort of the dark ages for ten years where it just wasn't a topic. In the seventies, Bill English was one of the first that went over to the PARC [Xerox PARC]. More people did, and Peter Deutsch, who was one of the very early PARC people from Berkeley, worked for us in the summers, doing programs.

Lowood: Why was it that everything except the mouse was ignored?


ENGELBART: It culturally didn't fit. Especially after the collapse in the mid-seventies, I didn't ever really feel comfortable going any place to visit people who were doing computer work. It would be fun to hear what they were doing, but there was absolutely no forum for talking about what we were doing. There just wasn't any interest or reflection back about any aspect of what we were doing. A very strange effect. There is a lot of precedent for things like that happening in the past, but when it does happen to you, it's a really interesting thing. Pretty soon you feel invisible.

Adams: How did you cope with that, emotionally/professionally?

ENGELBART: I guess by becoming invisible. I ended up sitting in a corner here at TYMSHARE for years, still once in a while trying to publish something, still thinking and working about where to go, and telling myself that it's pretty dumb, and that I should probably try to do something else that is more conducive to a proper career. I'd go through all of it and come out, finally I guess around 1981, just saying, "Well, I think with my particular orientation about the value I think all this would have to society, there is just no way I could give it up." It would blow my karma into small pieces. So I just better live longer. As a matter of fact, I started going to the Nautilus gyms and really worked at it. It was good therapy, anyway.

Lowood: Do you get any encouragement from outlining the programs that are now almost becoming as commonplace as word processors?

ENGELBART: Sure. Oh yeah.

Lowood: Do you see an opportunity there that the parts are kind of coming into place and maybe then the synthesis could be built on to these different parts?

ENGELBART: It really is. It's like the sunshine is coming out. Yes. Things like that channel 54 program--that just surprised me immensely, the sort of staging they gave me. It was very surprising. And last Friday night I got a real jolt. I got a call from someone who said, "I'm with the alumni from Oregon State University " where I graduated. I thought, oh my gosh.

Adams: You thought it was a fund-raising call for the university?

ENGELBART: I slipped by these last years without giving anything. So I was hardly listening. Then he was telling me about some award that they'd been giving annually to the outstanding alumnus and I thought, "Well, does that mean they need money for it, or do they want recommendations or something?" Then in the next breath he said, "And you have been selected for this year's award."

Adams: Did you ask him why?

ENGELBART: Yes, I said, "How in the world?" So he told me that someone had seen that piece that was written in the Examiner about a year or so ago and sent that in, and somebody else who happened to be on the committee, whom I had known back there, thought that was interesting and started finding out more and talked with Ballard. Then they got in touch with you [Henry]. So all this sneaky stuff going around. It just floored me. Ballard sent them up a video copy of that channel 54 thing.

Lowood: You know who you're hard on the heels of? Linus Pauling.

ENGELBART: Yes. How do you like that? He's one of my real heroes.

Adams: This is where Linus and Ava Helen went--the same university.

ENGELBART: Now there's a real hero, boy. I just think of how this stuff is kind of snowballing. A real thrill.

Lowood: One last question, because it is exactly on this theme. I think we'll have one more session where we'll finish up getting you from SRI to TYMSHARE and then talk for a while about present direction, and detach it a little bit from the historical chronology. But I wanted to ask you--you were just talking about people migrating to Xerox PARC and the perceptions of your presentation and now this new sort of historical awareness, that's coming up because some of the things that are hot at the moment fit with some parts of what you were doing. I wanted to ask you about your reaction to the standard chronology--usually this is associated with the mouse, but it has also been associated with windowing or document preparation, or help (it could be any one of those components). Some people say it all started with Apple Computer or with the Macintosh specifically. Other people say, "No, there is a whole historical development behind this. It goes back to Xerox PARC." Then a very few people at this point realize there is even a pre-history before that at SRI and your project. From your point of view, are those historical links valid in any sense? Is there really a valid connection between the kind of work that you were doing and the work that occurred at Xerox PARC and eventually, two generations after that, a computer like the Macintosh? Is there really a connection or is that just hype in a sense?

ENGELBART: I don't know. I felt like they rejected so much. There are people--some of the contributors, like Charles Irby. Once at an office automation conference four years ago, somebody wanted an historical picture of the AUGMENT, STAR and LISA, and they wanted to get that link. Charles Irby, when he left our place, pretty soon was involved in STAR and the software. Another guy that left, was making the STAR hardware. That guy went off to be a Macintosh developer, for Steve Jobs. So they enlisted both of those to try to discuss that situation; there was Charles Irby and Bob Belleville.

Lowood: John Postel, is he also in this category?


Lowood: So there are a lot of alumni.

ENGELBART: There were a lot of alumnus, fourteen or fifteen or more at Xerox PARC, and I'm not sure if any of them are left there now. They went off in different directions. I think the best way to assess that would be to talk to some of them. Certainly the thing of going off to icons and menus is a step backwards. I looked through all that kind of thing in the 60s and rejected it because the assessment I did was that it wouldn't go fast. You couldn't work rapidly enough that way. It would get in the way until you learned what the menu has and could operate it. Going with a keyset would give you so much more power.

Lowood: Did you ever visit them or talk to them?

ENGELBART: I would go occasionally, but it would get very uncomfortable for me. The guys who had worked with me were so excited by the level of technical support they were giving everybody at Xerox, which I could never get. It's a problem I had with support from the government sponsors. They would have their own picture about how much you could afford to spend on the programming end there. That meant we had that much less to spend on the computer support. I could never really get them to accept the proportion of our funds that would go to make the computer support so every programmer could really have his office. It was the strangest bind. Then at Xerox, everybody was on Alto and, oh boy, they were excited about the technology, and the timesharing. It was very exciting for them, they were so busy. Most of them, I think, felt that now that they could go and create the technology there really wasn't any more need for SRI ARC to go on, and in fact, one of them even told us that. And yet, no one was going at the dimensions of real application. Anyway, I just ended up really not feeling welcome. People were so anxious to tell me all this other stuff, which would get me all upset inside, because there wasn't anything I could say.

Adams: Xerox wouldn't buy the idea of that other level of research.

ENGELBART: There just didn't seem to be anybody who would.

Adams: That's where practical applications were going; they could have also been the locus of forward-looking research.

ENGELBART: That's something of a problem for me, anyway, that I am not a very aggressive talker and I just often get absolutely washed away by an excited, aggressive talker. They'll raise their voices and interrupt things. I finally just don't go visit them.

Lowood: Have you had any interaction at all with Alan Kay, because I suppose he would be somebody who looks at the Park operation at its heyday--he would be the equivalent figure to you in that operation, if anyone were guiding it, with any vision.

ENGELBART: Well, it was only one part of the Park that he was in the middle of.

Lowood: Yes, just the Alto project; Flex and all that sort of thing.

ENGELBART: I think doing the engineering in Alto was someone else, other people. I think Alan's stuff was more of a small talk thing and he wanted to build a Dynabook, and I was never very clear on what a dynabook is. Anyway, I didn't know much about what he was doing. I just think it would have been somehow so much better and more productive if somehow I could have formed links and visited and talked with people there and at Stanford, but I just couldn't find any overlap. That was part of it. It was really too bad.

Lowood: So the same was true then of Stanford.

[END OF INTERVIEW] [an error occurred while processing this directive]