Engelbart’s father was an electrical engineer who had a shop in Por...
This was a fundamental difference between Engelbart and many of his...
I believe the paper being referenced here is ["A Conceptual Framewo...
Engelbart was stationed on Leyte Island; a very remote island in th...
> I looked at all the crusades people could join, to find out how I...
For his Ph.D, Engelbart developed a shift register based on a bista...
Artificial intelligence was a very hot topic at SRI. Researchers we...
[A Conceptual Framework for Augmentation of Man’s Intellect](http:/...
## Sketchpad Ivan Sutherland at MIT’s Lincoln Lab, was a contempor...
### Language: tool for communication vs tool for thought The debat...
## CDC-160 The CDC-160 series was a line of general purpose comput...
![Manufacturing Jobs](https://i.imgur.com/4EsOAI5.jpg) It is worth...
### Keyset The keyset never reached the same degree of popularity ...
## Engelbart and the mouse Even tho a lot of people seem to beli...
### The Mother of All Demos The seminal 90-minute live public demo...
### MOL940 MOL940 or simply MOL (Machine Oriented Language) ...
[A Research Center for Augmenting Human Intellect](http://www.douge...
Xerox PARC, with Alan Kay at its head, further refined many of Enge...
16 IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997
Douglas Carl Engelbart: Developing the
Underlying Concepts for Contemporary
Computing
SUSAN B. BARNES
Currently, the World Wide Web is the hottest topic in contemporary comput-
ing and popular culture. The Web’s meteoric rise is difficult to escape notice.
Web stories are plastered in the popular press, and advertisements now in-
clude cryptic strings of letters starting with http://. But what is missing from
the current commercial descriptions of the Web is a discussion about the 30-
year history of research and development that created the underlying tech-
nologies on which the Web is based. Much of this foundation was laid in the
1960s by Douglas Carl Engelbart. In 1968, at the ACM/IEEE-CS Fall Joint
Computer Conference, Engelbart demonstrated his concept of “interactive
computing” to a group of computer scientists, and this is now considered a
seminal event in the history of computing. The technologies Engelbart origi-
nally presented included: windowed screen design, the user interface, hy-
pertextual linking of documents, the mouse, collaborative computing, and
multimedia. His pioneering work in the 1960s influenced future generations
of computer designers and developers. Almost 30 years after Engelbart’s
initial demonstration, many of his pioneering visions are now commonly
found in personal computers and the developing World Wide Web.
Introduction
oday, people do not need to know the name Thomas Alva
Edison to turn on an electric light and read a book. Simi-
larly, they do not need to know the name Douglas Carl Engel-
bart to turn on a personal computer, click on an icon with a
mouse, and access the digital libraries of the world. However,
unlike Edison, Engelbart remains a little-known figure in the
history of technology. Engelbart remains unfamiliar despite the
fact that he invented the mouse, window-style computer screen
displays, and other key concepts that are used everyday when
people operate their computers.
For over 30 years, Engelbart has pursued a vision of interacting
with computers that has remained mostly ignored by the computer
industry. However, without realizing it, the computer industry has
spent the last 20 years implementing ideas Engelbart first demon-
strated in the 1960s. Engelbart’s goal was to develop interactive
computer systems that would match computational capabilities
with human capabilities. He was interested in designing a new
category of computer applications for augmentation not automa-
tion. Utilizing computer technology, Engelbart envisioned a future
in which individuals and small groups would be able to augment
the strengths of their most important natural tools for dealing with
the world—the abilities to perceive, think, reason, analyze, and
communicate. First, he wrote a paper to explain his framework for
developing interactive computing. Afterward, he built a working
model to demonstrate his vision. In the process of creating the
model, Engelbart designed a series of interactive computer tools
that became the foundation for the future development of hyper-
text, computer-supported cooperative work, the graphical user
interface, electronic mail, on-screen editing, video conferencing,
and the World Wide Web.
Early Influences
Engelbart began his engineering career at Oregon State College in
1942. He enrolled in an electrical engineering program because he
tecting and ranging, hence, RADAR) technology. World War II
was in progress, and Engelbart recalls:
By the end of my sophomore year, if you were in college in
engineering, you were getting deferred. But, they dropped
all that, and so I got drafted at the end of my sophomore
year. And when you get drafted, you can opt for different
choices. So, I took a test that the Navy was giving for the
RADAR, trying to screen people to go into this new tech-
nology. So, as an enlisted man, I got to get drafted into the
Navy and go into its yearlong training program.
1
Engelbart was drafted into the U.S. Navy, where he was intro-
duced to a number of new technologies, including wave propaga-
1058-6180/97/$10.00 1997 IEEE T IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 17 tion, antennas, and amplifiers. Following this training, Engelbart was sent to the Philippines. As his ship backed out of its berth in San Francisco, whistles and firecrackers started going off, and peo- ple began cheering. Japan had just surrendered. Instead of going into active combat, Engelbart was sent to the Philippines to help with the demobilization efforts. As a Navy electronics technician, he took care of radios, sonar (then known as sound navigation rang- ing), teletype transmission, and radar. This experience was an intro- duction to all types of new communication technologies. Engelbart’s vision of using the computer as an interactive communication tool began to take shape after the end of World War II. Toward the end of his naval career, he read Vannevar Bush’s article “As We May Think” that was published in the July 1945 issue of Atlantic Monthly. The article inspired Engelbart to pursue a career developing better communication and “knowledge” tools. Bush published the article as a reflection on how technology could help solve problems in a postwar society. During the war, Bush realized that the amount of scientific data being written and collected was growing at an accelerated pace. As a result, he believed that people would need new types of knowledge tools to locate, organize, coordinate, analyze, and navigate through the increasing volumes of research information. In “As We May Think,” Bush described a hypothetical device called a memex (memory extender) as an example of the type of knowledge tools that should be developed. The memex was a work- station conceived to use a variety of analog technologies such as microfilm readers and storage. Bush described it as follows: A memex is a device in which an individual stores all his books, records, and communications, and which is mecha- nized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory. It consists of a desk, and while it can presumably be oper- ated from a distance, it is primarily the piece of furniture at which he works. On the top are slanting translucent screens, on which material can be projected for convenient reading. There is a keyboard, and set of buttons and levers. Other- wise it looks like an ordinary desk. 2 Bush’s memex was envisioned to aid the process of thinking through a mechanized indexing system. Different pieces of infor- mation in the system could be connected by creating individual- ized associative trails. A trail was similar to the trail of mental association in a person’s mind. Used as an augmentation tool, the memex would supplement and extend human memory capabilities and reflect the associative nature of the human mind. In the fol- lowing passage, Bush describes the process of linking trails of information: When the user is building a trail, he names it, inserts the name in his code book, and taps it out on his keyboard. Be- fore him are the two items to be joined, projected onto adja- cent viewing positions. At the bottom of each there are a number of blank code spaces, and a pointer is set to indicate one of these on each item. The user taps a single key, and the items are permanently joined. 3 “As We May Think” was not predicting the future, but creating it. 4 Bush’s compelling vision of information technology influ- enced future generations of computer pioneers. After reading Bush’s article, Engelbart went back to school to finish his BS in electrical engineering and graduated in 1948. Afterward, he took an engineering job at the Ames Research Laboratory in Mountain View, California, with the National Advisory Committee for Aeronautics (the forerunner of the National Aeronautics and Space Administration, NASA). After working at Ames for several years, he got engaged to be married. But the Monday after he became engaged, Engelbart realized that he had achieved all of his life goals—get an education, get a steady job, and get married. Now, at the age of 25, he began to think about a new goal for his life. He later described his decision-making process as follows: I dismissed money as a goal fairly early in the decision process. The way I grew up, if you had enough money to get by, that was okay; I never knew anybody who was rich. But by 1950, it looked to me like the world was changing so fast, and our problems were getting so much bigger, that I decided to look for a goal in life that would have the most payoff to mankind. 5 It looked to me like the world was changing so fast, and our problems were getting so much bigger, that I decided to look for a goal in life that would have the most payoff to mankind. For several months while driving to work, Engelbart contem- plated his skills and the various kinds of goals he could pursue. He was also looking for a cause that could utilize his engineering education and his interest in computers. While searching for his new goal, Engelbart realized that he ran into the same issue time and time again. Like Bush, he understood that humankind was moving into an era in which the complexity and urgency of global difficulties were surpassing the traditional tools for dealing with problems. Engelbart recalled: I had this kooky thing happen to me in 1951 where I de- cided to commit my career to trying to help mankind be able to cope better with complexity and urgency and the problems of the world. I had an image of sitting in front of a display and working with a computer interactively. I had been a radio and radar technician during World War II, so I knew that any signals that came out of a machine could drive any kind of hardware—they could drive whatever you wanted on a display. But I really didn’t know how a com- puter worked. Still, I though, “Boy! That’s just great!” The images of the different symbologies that you could employ, and other people sitting at workstations connected to the same complex, and working in a close, collaborative way. And I just said, “Well, that’s something I can pursue as an electrical engineer, and maybe try to follow as a goal.” 6 Based on his experiences as a radar technician, Engelbart envi- sioned using computers as a technology to interact with informa- tion displayed on screens. At the time, Engelbart had never worked with a computer. But, he had read about them in books. From his radar experience, Engelbart understood that if computers could show information on punch cards and paper printouts, they could write or draw information on a display screen. He saw the Underlying Concepts for Contemporary Computing 18 IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 connection among a cathode-ray screen, an information processor, and a medium for representing symbols to a person. Engelbart began sketching out a system that would enable people to steer through different information domains using knobs, levers, and transducers. It was not until eight years later that Engelbart was able to be- gin his goal of developing interactive computing systems. First, he quit his job and went to the Graduate School of Electrical En- gineering at the University of California at Berkeley. Professor Paul Morton had started to explore computer science and was developing the California Digital Computer. While in graduate school, Engelbart’s ideas about interactive computing were not understood or well-received by his colleagues, who were punch- ing cards to interact with the computer. In contrast to a traditional batch-processing method, Engelbart wanted to connect a keyboard to the computer and have a person interact directly with the ma- chine. Because his ideas were not acceptable to the university community, Engelbart settled on a different research topic to complete his degree. In 1955, he earned his PhD along with half a dozen patents for bistable gaseous plasma digital devices. After leaving academia, Engelbart began to look to the commercial world as a way to pursue his vision for developing interactive computer systems. Initial Concepts of Augmentation In October 1957, Engelbart accepted a job at an organization called the Stanford Research Institute (SRI). SRI was interested in conducting research in military, scientific, and commercial appli- cations for computers. During his first two years at SRI, Engelbart earned a dozen patents, working on magnetic computer compo- nents, fundamental digital-device phenomena, and miniaturization scaling potential. By 1959, Engelbart had enough recognition at SRI to begin to pursue his own research interests. With support from SRI and a grant from the U.S. Air Force Office of Scientific Research (from Harold Wooster and Rowena Swanson), he worked for a couple of years to produce a paper describing his interactive computing concepts. At first, he tried to find an established discipline as a basis for the framework of his design. But people in other disciplines, such as documentation and artificial intelligence, were not interested in his ideas. Finally, he discovered a Rand Corp. report written by J.L. Kennedy and G.H. Putt titled “Administration of Research in a Research Corporation.” The thesis of the report argued that when a researcher starts an inter- or new-discipline project, the researcher would encounter difficulties when approaching indi- viduals in established disciplines. Each discipline has its own unique conceptual framework that new members of the profession begin to learn during the first year of professional school. If a conceptual framework did not exist for a new-discipline research project, then an appropriate framework must be created. After reading this report, Engelbart started developing his own unique conceptual framework for designing interactive systems. In 1962, Engelbart finished his paper titled “A Conceptual Framework for the Augmentation of Man’s Intellect,” and a con- densed version was published the following year in P.W. Hower- ton and D.C. Week’s book Vistas in Information Handling. In his paper, Engelbart defined augmenting man’s intellect as the capa- bility of a man or woman to approach a complex problem situa- tion, gain understanding about his or her particular needs, and derive solutions to the problems. Engelbart’s system was concep- tualized to augment or support people working in a variety of disciplines, including diplomacy, management, social science, programming, law, and design. He wanted to develop a set of core tools that all these various intellectual workers could use. Addi- tionally, he wanted to develop specialized tools to meet particular requirements to help augment the way different types of intellec- tual workers solve problems. In his 1963 paper, Engelbart explained his theoretical approach for developing computer-based augmentation systems. Engel- bart’s conceptual framework attempted to orient people toward the possibilities and problems associated with using modern tech- nology. In developing his design theory, he started with the human perceptual system. An individual communicates to and receives information from the world through his or her motor channels. According to Engelbart, this processing is of two kinds. The first category of processes includes those processes of which a person is generally conscious, such as recognizing patterns, remember- ing, visualizing, abstracting, deducing, and inducing. The second category of processes involves self-generated information, such as unconscious processing, mediating received information, and mediating conscious processing itself. These methods of process- ing information are acquired as individuals grow up and live within a given culture. People are specifically taught conceptual and symbolic ways to process information, including spoken and written language. Moreover, people are taught to use tools and methods for organizing structures. All of these symbolic and or- ganizational methods constitute one big system that augments human capability. See Figs. 1 and 2. Fig. 1. In the development of his chapter “A Conceptual Frame- work for the Augmentation of Man’s Intellect,” Engelbart argued that our basic mental–motor–perceptual machinery cannot do much by itself. Furthermore, Engelbart argued that culture shapes the proc- essing of information by developing techniques for us to organize and utilize our basic human intellectual capabilities so that we can understand very complex situations and attain methods for devis- ing and implementing solutions to problems. In examining the methods and tools people use in the problem-solving process, Engelbart developed the concept of augmentation means. Aug- mentation means are the tools and methods used to supplement human problem-solving capabilities. Engelbart categorized aug- mentation means into four classes: IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 19 1) Artifacts—physical objects designed to provide for human comfort, the manipulation of things or materials, and the manipulation of symbols; 2) Language—the way in which the individual classifies the picture of his world into the concepts that his mind uses to model that world, and the symbols that he attaches to those concepts and uses in consciously manipulating the concepts (“thinking”); 3) Methodology—the methods, procedures, and strategies with which an individual organizes his goal-centered (problem- solving) activity; and 4) Training—the conditioning needed for the individual to bring his skills using the three augmentation means listed above to the point where they are operationally effective. 7 Fig. 2. In the design of the augmentation system, Engelbart treated it as a two-part system. One part was technology, and the other part was human factors. Engelbart created a diagram (see Fig. 3) of a general system to illustrate how human processes interact with artifact processes. He referred to this type of general system as an H-LAM/T system (Human using Language, Artifacts, and Methodology, in which he is Trained). In his paper, Engelbart proposed to design a new computer-based system that would bring together a trained human being with his or her artifacts, language, and methods. In contrast to traditional artifacts, such as paper and pencil, the artifacts in the new system would include computers with computer-controlled information-storage, information-handling, and information- display devices. Engelbart realized that a shift from artifacts, such as paper and pencils, to digital computer-based tools would alter organizational structures, because changes in the augmentation means would alter the way people work. A change in physical artifacts would consequently change the language and the methods used in prob- lem-solving activities within an organization. Simultaneously, while he was designing his augmentation system, Engelbart was aware that this new technology would alter the way people work. Realizing that computer-based systems would change work prac- tices, Engelbart argued that people would require training. Training is necessary to bridge the gap between old and new technological systems. Engelbart used the example of an un- trained aborigine to illustrate this point. An untrained aborigine would not be able to drive a car in urban traffic, because the abo- rigine would not be able to bridge the gap between his cultural background and contemporary society. However, it would be pos- sible for the aborigine to learn step by step how to drive safely. According to Engelbart, human minds learn by a series of small steps rather than by large leaps, and the small steps depend on the previous steps. The structure or organization of these small ac- tions or steps can be called process hierarchies. Fig. 3. A diagram representing the two active domains within the H-LAM/T system. Process hierarchies are broken down into individual small steps or subprocesses. People develop a repertoire of subprocess capabilities or tools that they select and adapt to execute higher order activities. For example, a mechanic must know what func- tions his or her tools can perform and how to use them. Similarly, the intellectual worker must know the capabilities of his or her tools and understand the suitable methods, strategies, and rules for making them work. When an intellectual worker writes a docu- ment, the process of writing is used as a subprocess or tool for higher order activities, such as organizing a committee, creating a report, or preparing a legal brief. The shift from pencils and paper to digital tools, such as word processing, changes information handling and the skills required to use new technological artifacts. Thus, people would need to learn how to work with these new tools. Symbolic Manipulation In Engelbart’s terminology, working with a complex machine involves a cooperative interaction between the domains of human processes and artifact processes. The term man–machine inter- face, now called the discipline of human–computer interaction, has been used for some years to represent the boundary across which energy is exchanged between people and machines. Engel- bart stated that the exchange across this “interface” occurs when an explicit human process is coupled with an explicit artifact or computer process. For example, the finger and hand motions used to type on a keyboard (explicit human processes) activate key- linkage motions on the keyboard (explicit artifact processes). Thus, the computer interface is the point of interaction between the artifact processes of the computer and human-processes. Underlying Concepts for Contemporary Computing 20 IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 According to Engelbart, developing computer artifact proc- esses to match human processes would result in a new fourth stage in the development of human intellectual capabilities. In his framework, he devised a hypothesis to describe each of the four stages of human symbolic manipulation. The first stage is concept manipulation. Humans have the biological capability of generat- ing concepts and abstractions. They can mentally manipulate these concepts and “think” about abstract situations. The second stage is symbol manipulation. In this stage, humans take a great step forward when they learn to represent specific concepts in their minds with particular symbols. In the third stage of development, manual external symbol manipulation, symbol manipulation activity now supplements the individual’s memory and ability to visualize concepts. Manual external symbol manipulation includes graphic repre- sentation and the “means” or devices used to make graphic representations of symbols such as a stick and sand, pencil and paper, and straightedge or compass. He argued, that the “means” or methods used to externally manipulate symbols have principally been associated with the evolution of the in- dividual’s present way of manipulating concepts or thinking. Simply stated, the media we use to symbolically record ideas will influence the way we think. Engelbart asserted that cur- rent patterns of thinking have been shaped by the way we manually manipulate symbolic representations. For instance, the written alphabet influences our ability to conceptualize the world, because it is an abstract method for symbolizing peo- ple, objects, and ideas. The literate mind creates a worldview that is not possible by experiencing the world through our senses, because written concepts can be constructed through logic, deduction, and critical reasoning. In stage four, Engelbart combines symbol manipulation with the use of language. He asserts that in addition to the methods we use to manipulate symbols, language also affects the way we think. Building on the Whorfian hypothesis 8 that argues language plays a prominent role in shaping the perceptual world of the people who use it, Engelbart further argued that a second factor in language’s influence on culture is external symbol manipulation. Using this second factor, Engelbart proposed a neo-Whorfian hypothesis that stated, “both the language used by a culture, and the capability for effective intellectual activity, are directly af- fected during their evolution by the means by which individuals control the external manipulation of symbols.” 9 In other words, the tools or media we use to externally manipulate symbols influ- ence language and ultimately the way we think. Interactive Computing Engelbart’s framework (see Fig. 4) suggested a fourth stage in the evolution of human intellectual capability. Engelbart called this new s tage “automated external symbol manipulation.” Com- puters add the ability to automate the external manipulation of symbols. A shift from manual to automated external symbol manipulation would result in a fourth stage in the development of human intellectual capabilities. This fourth stage would occur because, according to the neo-Whorfian hypothesis, these new tools would alter the human thinking process. In this fourth stage of evolution, Engelbart described his vision for interactive computing: The symbols with which the human represents the concepts he is manipulating can be arranged before his eyes, moved, stored, recalled, operated upon . . . this could be a computer, with which individuals could communicate rapidly and eas- ily, coupled to a three-dimensional color display within which extremely sophisticated images could be constructed, the computer being able to execute a wide variety of proc- esses on parts or all of these images in automatic response to human direction. 10 Fig. 4. A diagram from Engelbart’s framework showing how the mo- tor capability of the abstracted human module drives the computer tools and employs the associated methodological and conceptual- linguistic parts of his augmentation system in different modes. (P = perceptual, M = Motor; see Fig. 1.) Engelbart was aware that moving from manual to automated symbol manipulation would create changes in the human system. In 1988, Engelbart stated: It takes a long time (generations) to discover and implement all of the fruitful changes in the human system made possi- ble by a given, radical improvement in technology. Where, as is the situation now, technology improves by rapid, large steps, it is predictable that the human system will become critically stressed in trying to adapt rapidly in ways that formerly took hundreds of years. There has to be a much- enhanced consciousness about concurrent evolution in the human system. The technology side, the tool system, has inappropriately been driving the whole. What has to be established is a bal- anced coevolution between both parts. How do we establish an environment that yields this coevolution? Well, that’s where the bootstrapping in a laboratory comes in. I said I wanted to do what I knew it was going to be like in our fu- ture. So we had to be more conscious of the candidates for change—in both the tool and the human systems. 11 Engelbart came up with the concept of bootstrapping as a method of analyzing the coevolution of human and tool sys- tems. Bootstrapping is building the tools to build better tools and simultaneously testing them on yourself during the inven- tive process. Engelbart and his team served as both the develop- ers of the technologies and the subjects for the analysis and evaluation of the new augmentation system. To implement the IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 21 bootstrapping strategy, a team of people from diverse disci- plines, including psychology, computer programming, and com- puter engineering, were assembled. People working on the proj- ect were expected to change to this newly evolving design and its emerging terminology. Support From ARPA Once Engelbart’s theoretical framework was completed, the next step was to get funding to build and test a working model. Engel- bart first gained support from Licklider through the Advanced Research Projects Agency (ARPA) and later from Bob Taylor at NASA. Licklider, a psychology professor and researcher at the Massachusetts Institute of Technology, shared Engelbart’s vision of interactive computing. In 1960, he wrote a paper titled “Man– Computer Symbiosis.” 12 In this paper, Licklider asserted that a new relationship between humans and computers was necessary. Martin Campbell-Kelly and William Aspray have described his paper as follows: The single most important idea in Lickliders paper was to advocate the use of computers to augment the human intel- lect. This was a radical viewpoint at the time and, to a de- gree, at odds with the prevailing view of the computer es- tablishment, particularly artificial intelligence (AI) re- searchers. Many computer scientists were caught up by the dream that AI would soon rival the human intellect in areas such as problem solving, pattern recognition, and chess playing. Soon, they believed, a researcher would be able to assign high-level tasks to a computer. Licklider believed that such views were utopian. He argued that computer scientists should develop systems that enabled people to use computers to enhance their eve- ryday work. He believed that a researcher typically spent only 15 percent of his or her time “thinking”—the rest of the time was spent looking up facts in reference books, plotting graphs, doing calculations, and so on. It was these low-level tasks that he wanted computers to automate in- stead of pursuing a high-flown version of AI. 13 In 1962, Licklider became the first head of ARPAs newly es- tablished Information Processing Techniques Office (IPTO). IPTO’s objective was to invest in research projects that would provide more capable, flexible, intelligent, and interactive com- puter systems. Licklider ’s interests in the human factors aspects in complex human–human and human–machine systems made him an ideal director of IPTO. Licklider had come to IPTO in 1962 with a vision of how to increase human–computer interaction and established a pro- gram of breadth and depth to realize this vision. His succes- sors ... all subscribed to this vision and continued to pro- mote it. 14 Licklider moved quickly to get funding for Engelbart’s project. By early 1963, funding was approved, and Licklider requested that the project be connected to the System Development Corpo- ration’s (SDC) AN/FSQ32 computer in Santa Monica, California. SDC was developing a time-sharing system for the Q32. Later that year, an online data link was established between Menlo Park, California, and Santa Monica with a CDC 160A minicomputer at the SRI location. Engelbart describes the computer as follows: Mind you, the CDC 160A, which was the only commer- cially suitable minicomputer that we knew of, even though having only 8K of 12-bit words, and running at about 6 mi- croseconds per instruction, cost well over$100,000 (1963
dollars). It had paper tape in and out; if the system crashed,
and the most recent dump of your working file (paper tape),
before you could continue. A crude, industry-standard
Flexowriter (online typewriter) could be driven; otherwise it
was paper tape in and out.
15
Although the funding Licklider provided was a start, the proj-
ect needed more money to develop interactive systems. Suddenly,
an offer for support came from Bob Taylor, who was working at
and left behind a copy of his framework report and proposal. Af-
The combined ARPA and NASA backing was enough to begin
developing a working model.
The Augmentation Research Center
Engelbart set up the Augmentation Research Center (ARC) at
SRI. The goal of this research was directly oriented toward
augmenting the capability of humans to deal with tough knowl-
edge work. Engelbart’s team wanted to invent new methods to
process effectively the large volumes of information that
knowledge workers use. Peter Drucker had coined the term
knowledge workers. Drucker defined knowledge as the system-
atic organization of information. Knowledge workers were peo-
ple who create and apply knowledge to productive goals. In the
late 1960s, Druckers The Age of Discontinuity predicted that a
social transformation would occur in the last quarter of the 20th
century. The U.S. economy would shift from an economy that
manufactured products to one that was based primarily on the
exchange of knowledge.
Engelbart’s primary research goal was to invent technologies
that would support the knowledge worker. A secondary goal of
the augmentation research was to develop computer systems that
support active collaboration among groups of knowledge workers.
This aspect of Engelbart’s research has become the discipline of
computer-supported cooperative work. His augmentation system
was the first major research work done in office automation and
text processing. In fact, the entire augmentation project was much
more ambitious and broad in scope than the current office pro-
ductivity tools we use in the professional work environment of the
1990s. Today, we are still trying to implement many of the con-
cepts Engelbart introduced in the 1960s.
The system ARC developed was called NLS (for oNLine Sys-
tem). It included such advanced features as electronic mail, com-
puter conferencing, multiple windows on a screen, and a mouse.
NLS was designed to enable any team member to read material
written by anyone else working on the project and make com-
ments on the documents. Electronic documents could be accessed
from any terminal connected to the augmentation system. Docu-
ments were “shared” and distributed among members of the group
during the development process. This project was the first proto-
type of a paperless office. No paper changed hands during the
tually immediate. The ability to create, share, and store documents
Underlying Concepts for Contemporary Computing
22 IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997
was achieved by connecting the text-editing system with a special
kind of electronic filing organization system that would serve as a
combined memory, record, and medium for unifying individual
research efforts.
A software journal was established for individuals and groups
to have access to a shared electronic space for thinking and com-
municating. The journal enabled individuals to insert comments
into the group record of the augmentation experiments or browse
through them. Moreover, the augmentation system allowed pro-
grammers to trace the way system features had developed and
evolved. The journal was part of the overall dialogue support
system designed to make more effective the decision-making
process in small group interactions. Additionally, the journal or-
ganized the dialogues, notes, and publications created during the
process of developing the augmentation system and testing how
the NLS software worked. NLS was the central software applica-
tion, and it focused on the composition, modification, and study
of structured textual material. Computer interfaces were devel-
oped to enable users to interact with the textual information dis-
played on screens and terminals.
Engelbart and his team designed unique methods for interact-
ing with the system. In addition to a traditional keyboard, the user
could input commands with a five-key, one-hand chord keyset.
Simple one-character commands were input through the chord:
for example, I (insert), D (delete), and M (move or rearrange).
After the command is typed, the user points with the “mouse” to
tell the system where on the screen the command should be exe-
cuted. Engelbart’s mouse was a box with hidden wheels under-
neath. Ted Nelson described working with the mouse as follows:
The Engelbart Mouse is a little box with hidden wheels un-
derneath and a cable to the terminal. As you roll it, the
wheels’ turns are signalled to the computer and the com-
puter moves the cursor on the screen. It’s fast and accurate,
and in fact beats a lightpen hands down in working speed.
16
In tests, the mouse outperformed the lightpen and joystick in
placing a cursor in random screen positions. Engelbart states:
“The actual invention of the mouse was the result of analyzing the
various characteristics of other pointing/input devices.”
17
tion to hand-held devices, they experimented with knee and head
cursor controls. But, muscles would cramp. As a result, they
stayed with the mouse. Engelbart’s mouse was an analog device
covering large steel wheels and exposing a series of three buttons.
By rolling the mouse over a flat surface, the operator could move
rapidly through a document, pointing the cursor at target areas and
clicking the mouse buttons to enter commands. With a few hours
of practice, a person could learn to operate the mouse and chord
keyset. Engelbart describes interacting with the NLS system as
follows:
Besides using a standard keyboard for character entry, an
[NLS] user may optionally use a five-key, one-hand, chord
keyset. Remarkably little practice is required in order to
enter alphabetic characters, one hand-stroke per character.
With less than five hours practice, a person can begin prof-
itably working in a two-handed, concurrent mode—
operating the mouse with one hand and simultaneously en-
tering command characters and short literal strings with the
other.
Here is an example of a low-level action which reveals
some basic characteristics of high-performance execution. It
is a very simple situation, but representative of what is met
over and over and over again in doing hard knowledge
work. The worker is composing or modifying something in
one area of the screen, when his eye catches a one-character
typo in another area. For a skilled [NLS] worker, the typo
could be corrected in less time than it would take someone
to point it out to him—with three quick strokes of the keyset
hand during a casual flick of the mouse hand, and an abso-
lute minimum of visual and mental attention taken from the
18
The display consoles were equipped with typewriter-style key-
boards, the five-finger chord keyset for one-character input, and a
mouse for cursor control (see Fig. 5). Screens were generated on
small CRTs in a machine room and transmitted through a closed-
circuit television system to individual workstations. When char-
acter-based displays became available, Engelbart’s team devel-
oped external boxes for attaching mice and keysets.
Fig. 5. Captured frame from the 1968 demonstration. Split screen
with Engelbart demonstrating how to use the chord keyset in his left
hand and the mouse in his right hand.
The Demonstration
Engelbart’s project received its first time-sharing system in 1968,
an SDS940 (see Fig. 6). They designed their own responsive dis-
play system to work with the computer. Engelbart described it as
follows:
The computer and display drivers time-shared their atten-
tion among multiple, 5-inch CRTs, which happened to be
the most economical size for a given percentage of screen
resolution. In front of each CRT we added a commercial,
high-quality video camera, mounted with a light shroud
over the camera lens and CRT screen. The resulting video
signal, amplified and piped out to our laboratory, drove the
video monitors that were our workstation displays. Two
display generators, each driving up to eight CRTs, imple-
mented with vacuum-tube technology, were both bulky and
very costly. It took one and a half people to keep those
things running all of the time. The stroke-generated charac-
ters and vector graphics allowed us to have flexible, mixed
text and graphic document presentations. The display gen-
erators were driven from a direct-memory-access channel
that provided very fast (i.e., one refresh-cycle time) creation
of a new display image.
19
IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 23
Fig. 6. The basic organization of the SDS940 system used in 1968.
Subroutines that did the application work were written with the
MOL940 language developed by Engelbart’s team. The control proc-
esses were specified in a control metalanguage, then compiled with a
control-metalanguage translator into the control processor, which
interacted with the user. A tree-meta-translator enabled the team to
do compiler compiling. They described a new compiler in a meta-
compiler language, then compiled it with the tree-meta-translator into
the new running compiler.
27
Now, Engelbart and the 16 members of his team were ready to
demonstrate the system they had invented and introduce their
concept of interactive computing (see Fig. 7). They decided to
risk everything by demonstrating their innovative new way of
operating computers to a major gathering of the computer indus-
try known as the ACM/IEEE-CS Fall Joint Computer Conference
in San Francisco in December 1968. Engelbart had to use the
majority of his grant money to obtain special video projectors and
microwave antennas to transmit between the lab and the confer-
ence center. Additionally, the team built modems for the demon-
stration. Engelbart described how the demonstration was pre-
sented as follows:
We set up to give an online presentation using a video pro-
jector pointing at a 20-foot screen. Brooks Hall is a large
auditorium, and that video projector could put up our dis-
play images so you could read them easily from up in the
balcony. The video projector we rented (built by a Swiss
company, Eidophor) used a high-intensity projection lamp
whose light was modulated by a thin film of oil, which in
turn was modulated by the video signal. On the right side of
the stage, I sat at our Herman Miller [an office furniture
company’s custom-designed] console. We set up a folding
screen as a backdrop behind me. I saw the same image on
my workstation screen there as was projected for the audi-
ence to see.
We built special electronics that picked up the control inputs
from my mouse, keyset, and keyboard and piped them down
to SRI over a telephone hookup. We leased two microwave
lines up from our laboratory in SRI, roughly 30 miles. It
took two additional antennas on the roof at SRI, four more
on a truck up on Skyline Boulevard, and two on the roof of
the conference center. . . . Back in our lab, we dismantled a
number of the display units in our display system, so that
we could use the cameras in San Francisco and SRI. We
borrowed a few tripods and got some extra people to be
camera people. . . .
On my console on the stage, there was a camera mounted
that caught my face. Another camera, mounted overhead,
looked down on the workstation controls. In the back of the
room, Bill English controlled use of these two video signals
as well as the two video signals coming up from SRI that
could bring either camera or computer video. Bill could se-
lect any of these four video images with optional mixing
and frame splitting. We had an intercom that allowed him to
direct the action of the people in our lab at SRI who were
generating computer images or handling the cameras send-
ing the video up from SRI.
20
Fig. 7. Engelbart demonstrating his system at the 1968 ACM/IEEE-
CS Fall Joint Computer Conference.
After spending most of his grant money on the presentation,
Engelbart’s project would have been in financial trouble if the
demonstration did not work. Fortunately, it did work, and the
people in the audience got very excited about developing interac-
tive computer systems. Engelbart’s demonstration showed the
type of information space Bush had described in his 1945 article
and the method of interactive computing that Licklider wrote
In the demonstration, he accessed files from the computers
memory and displayed them on a large screen in the auditorium.
Then Engelbart collapsed the written textual descriptions into a
series of one-line headings (a basic outline). A click on the mouse
button would then expand the headings into the larger text. An-
other command displayed a graphic representation of how the
various documents were linked in the system (see Fig. 8). Finally,
through another set of commands, Engelbart connected with a
colleague located at his lab and displayed on the screen a video
image of his coworker. They could see the document, see each
other, and discuss the project through audio contact. In this dem-
onstration, Engelbart showed hypertext links (invented with Jeff
Rulifson, see Fig. 9), computer-supported cooperative work,
video conferencing, computer graphics, and sophisticated word
processing.
To further explain his ideas to the computer establishment,
Engelbart coauthored a conference paper with English. The paper
was written to be an example of the rules of hierarchy and text
structuring used in the augmentation system.
Underlying Concepts for Contemporary Computing
24 IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997
Every paragraph . . . started with an alphanumeric label in
the manner of the classic outlining system learned by most
high school students. Engelbart and English reported that,
when made dynamic by a computer processor, their design
facilitated a number of varied and well ordered perspectives
on information. For example, users could check logic and
completeness by requesting the screen to display
“statements” having the same outline level of importance;
they could quickly locate specific supporting material by
selecting the appropriate alphanumeric heading; they could
expand and tie together topics by creating a subheading at
the most appropriate point in a given body of thought.
21
Fig. 8. During the 1968 demonstration, Engelbart explained how his
interactive system could link information. Engelbart coinvented hy-
pertext links with the system’s chief architect, Rulifson.
Fig. 9. Teleconferencing during the 1968 demonstration between the
SRI lab in Menlo Park and Brooks Hall in San Francisco. A view of
text retrieval with the image of Paxton in the corner of the screen.
Impressive as Engelbart’s demonstration was, the commercial
computer establishment did not think his vision of interactive
computing was practical. They did not think it was realistic, be-
cause the demonstration represented a paradigm shift from batch
processing punched cards to direct interaction. Despite the lack of
support from the computer establishment, the NLS project contin-
ued to develop at ARC with funding from ARPA.
On the Arpanet
In 1969, ARC became the second site connected to the Arpanet
system, the beginning of today’s Internet. Joining the Arpanet
achieved a subgoal of Engelbart’s NLS project: to support ac-
tive collaboration among groups of workers. In the spirit of
collaboration, Engelbart and his team volunteered to develop
and operate the Network Information Center for the original
Arpanet research community. Sharing information among many
computers in different locations was not a new idea to the ARC
team. They had been working on a smaller, localized network
for years. With the creation of the Arpanet, Engelbart’s NLS
system now became accessible to ARC researchers working
across geographical distances and became accessible to other
researchers outside of the ARC project. In 1970, the NLS soft-
ware was reprogrammed for the PDP-10 TENEX. Engelbart and
Lehtman describe NLS programming tools as follows:
ated: Tree Meta, a compiler-compiler, was used to bootstrap
us onto different machines ([S]DS 940, PDP-10, PDP-11,
and DEC 20) and to create the other compilers and assem-
blers we used. L10 was a block-structured language with
pattern-matching and string-construction facilities. The
same pattern-matching syntax was used by less sophisti-
cated users to generate filters in the core workshop. The
Command Meta Language (CML) was used to create user
interfaces that were independent of terminal type (display or
typewriter) and individual user preferences. CML grammars
were interpreted. Contextual entries into syntactic and se-
mantic help systems were generated from the CML gram-
mars. The Output Processor interpreted a comprehensive
document-formatting language.
22
In addition to software tools, the architecture and general char-
acter of the NLS system was oriented toward supporting collabo-
rative dialogue among distributed participants. Key features of the
system included the online Journal, electronic mail, and a Shared-
Screen Telephone conference system. The journal system, origi-
nally conceived about 1966, was developed to chronicle events by
means of a series of unchangeable entries. These entries would be
preserved for later integration into more organized treatments.
Additionally, Engelbart wanted the entries to serve the same pur-
pose as a professional journal. Engelbart compared his online
journal to a professional print-based journal:
The NLS Journal serves its users in a manner similar to pro-
fessional journals, with these significant quantitative differ-
ences: fast, flexible computer aids serve an author in creat-
ing a dialog item, and help multiple authors to collaborate
toward that end; publication time is very much shorter; sig-
nificant “articles” may be as short as one sentence; cross
reference citations may easily be much more specific (i.e.,
pointing directly to a specific passage); catalogs and indexes
can be accessed and searched online as well as in hard copy;
and full-text retrieval with short delays is the basic operat-
ing mode. The end effect of these changes is a form of re-
corded dialogue whose impact and value [have] a dramatic
qualitative difference over the traditional, hard-copy journal
system.
23
Five years after joining the Arpanet, the NLS Journal system
had 30,000 accession numbers. Approximately 20,000 were used
for Journal items and 10,000 for catalogued hard-copy items.
Both whole documents and think pieces appeared as items in the
Journal. In addition to sequencing items as a record of a
IEEE Annals of the History of Computing, Vol. 19, No. 3, 1997 25
“conference session,” any item could become a part of another
“session” or be used as a citation link. Moreover, authors had the
option of restricting access of an item to a specified distribution
list, and the item would not appear on public indexes.
In addition to the Journal, the NLS system was equipped
with a Shared-Screen Telephone option. This option developed
from a special “connection” feature in the NLS system that al-
lowed mutually agreeing users of Display NLS to connect
through the computer system by sharing the same NLS control
and display portrayal.
In this mode of teleconferencing, two or more people, posi-
tioned at separated display consoles, can link their displays
so that all see the same image, and at option any can exer-
cise control. When simultaneously talking on the telephone
the resulting dialogue can be uniquely effective—
corresponding to an in-person conference around a collec-
tive assemblage of their scratch pads, working records, and
individual support facilities. To add extra power to such ses-
sions, special conferencing-aid tools and conventions can
24
Both the Journal System and the Shared-Screen Telephone op-
tion demonstrated how computers could be used to support
“collaborative dialogue.” From Engelbart’s perspective, the most
important feature of networking was bringing together human
resources—knowledge, skill, creativity, intelligence, and drive. In
the early 1970s, he envisioned a future in which collaborative
human communication was built on a foundation of computer
networks.
Aftermath
The Arpanet technology made it possible for outside research
groups to connect to a computer network and use the NLS system
with the bootstrapping methodology. In the early 1970s, the wider
ARPA-funded community of computer researchers was beginning
to join the bootstrapping process. Ironically, just when Engelbart
decided that team augmentation or workgroup computing would
be their goal, his own team began to experience growing pres-
sures from the bootstrapping process. These pressures were tech-
nological, psychological, and social.
The technological pressures Engelbart’s team experienced
came from the building of a computer system that tried to keep
pace with the latest hardware and software advances. Conse-
quently, the system changed every six to eight months and had to
be constantly upgraded or redesigned. The social pressures were
the result of people on a regular basis having to learn new roles,
change old attitudes, and adopt different methods of working
together and with the technology. Thus, the augmentation project
became very difficult for the development team, who, in addition
to developing the technology, were also the experimental subjects
for the bootstrapping process. Psychologists and sociologists were
brought in to serve as observers and facilitators; and they became
as important to the team as the hardware and software developers.
The psychologists observed, first, that people are resistant to
change in the workplace. Second, people who initially opposed
learning a system become equally unwilling to give it up after
they have learned it. Initial resistance to technological change is
partially grounded in a general fear of the unknown. This resis-
tance to change is an essential part of introducing new technolo-
gies into established organizational settings. The psychological
and organizational tensions created by resistance to change were
apparent in Engelbart’s team. Team members were required to
observe themselves in order to create methodologies and proce-
dures that advanced with the developing computer technologies.
As a result of these pressures, a number of team members left
ARC to work at Xerox PARC.
In the early 1970s, he envisioned a
future in which collaborative human
communication was built on a
foundation of computer networks.
Although members of Engelbart’s team moved on to other re-
search projects, NLS continued to be developed. It developed at
SRI over a 13-year span and was then sold when the ARPA
funding ended. At the point when the ARPA funding stopped,
Engelbart had more or less invented half the concepts of modern
computing. Without funding, the staff went from a high of ap-
proximately 35 people down to a dozen, and finally Engelbart was
left alone with the software. In 1978, the commercial rights to the
NLS system (renamed Augment) were transferred to the Tym-
share Corporation (owned by McDonnell-Douglas), and the sys-
tem continued to be developed for another 12 years in a commer-
cial form. At Tymshare, solid enhancements were made to the
Conference Subsystem, the Mail System, and the Journal System.
Augment was then marketed as an integrated office automation
system and used by the U.S. Air Force. However, the Augment
software ran on big computers, and business waned when per-
sonal computers proliferated in the 1980s. In 1988, Engelbart
recalled:
while it was exciting to see bits and pieces of the original
NLS . . . appear commercially over the years, many of the
system’s conceptual core [ideas] have only recently been
recognized: outline editors (for easy manipulation of ideas);
hypertextual linking capabilities fully integrated into the
system; a system of recorded dialogue that transcends most
mail systems; user programmability and customizability of
the system; and most important, tools for augmenting not
just individual knowledge workers but also teams of people
both coresident and distributed over the world interacting
through a networked environment.
25
Engelbart-originated ideas have played a major role in the de-
velopment of the computer industry. But Engelbart is not rich or
famous, despite the fact that he originally developed many of the
concepts and techniques embedded in personal computers and the
World Wide Web. Today, Engelbart’s vision of using computers to
enable people to better handle complexity and urgency continues
to evolve. Currently he has expanded his original conceptual
framework and is exploring how to create high-performance hu-
man organizations through the use of computer tools such as
groupware. Engelbart states:
My own vision about pursuing high-performance organiza-
tions has matured over the years into a quite comprehensive,
multi-faceted, strategic framework. It may seem a bit radical