viking@iuvax.UUCP (Jon W. Backstrom) (01/29/88)
Since there has been a request for more information, I offer an article
that I originally posted three months ago on 'that other board'. This
paper details the history and development of hypertext as originally
envisioned by Vannevar Bush and defined in the context of a computer
implementation by Ted Nelson. Reactions are welcome.
Warning! LONG MESSAGE...but hopefully informative.
---------
Harnessing the Power of Information Technology:
The Genesis of Hypertext
Copyright (C) 1987 by Jon W. Backstrom
All rights reserved.
INTRODUCTION
The term 'hypertext' was first coined by Ted Nelson, a computer
visionary. In the 1960's, Nelson was responsible for designing
computer systems that would allow users to input and manipulate
text information (early word processors). Along with systems
designed to manipulate text, Nelson began the design of a system
that would allow documents to be linked together with pointers
forming meaningful associations between various topics.
Nelson's early work was inspired by the work of Vannevar Bush,
science advisor to Franklin D. Roosevelt. Bush is generally
regarded as the Father of Information Science and served as
Director of the Office of Scientific Research and Development.
In 1945, he authored an article calling for a new relationship
between man and the sum of his knowledge:
Bush, Vannevar "As We May Think", The Atlantic Monthly,
July 1945, pgs. 101-108.
DESCRIPTION
Nearing the end of World War II, Vannevar Bush wondered which
direction science would lead mankind. The scientists who were
mobilized for the war effort were finally free to pursue their
own interests, rather than work on destructive weapons. A great
deal of research was conducted during the war and many documents
were created.
Bush argued that sifting through and utilizing the record of
writings produced by scientific research was a difficult process.
The problem was not that too much data was being produced, but
the ability to make use of it was hampered by a lack of tools and
references which would lead you to any desired information.
Information Storage Technology
The development of new technologies and fact that complex devices
could then be produced cheaply, led Bush to speculate that there
were many mechanical aids which could effect a transformation in
scientific records. In his words, "The world has arrived at an
age of cheap complex devices of great reliability; and something
is bound to come of it."
A record must be extended, stored, and consulted, if it is to be
useful to science. The technology of the day allowed the record
to be stored through photographic means. Microfilm was being
introduced and developed as a means of storing and compressing
the scientific record. The compression and distribution of
information was very important to Bush. With the technology of
the day, Bush speculated that it would be possible to record the
entire Encyclopedia Britannica on a single sheet of film, as big
as a normal sheet of paper. Copies of this material could then
be reproduced, providing inexpensive access to information.
Creating the record is a process of collecting observations and
data, extracting parallel material from the existing record, and
inserting the new material into the general body of the common
record. While there is no substitute for mature thought, Bush
predicted great advances in the ability of people to store their
ideas. He described mechanical typewriters that would respond to
speech, scanners that would read and encode typed text, and the
advent of machines that would perform repetitive calculations and
other instructed operations at great speed (computers).
Bush stated that the process of repetitive thought, as opposed to
creative thought, would be greatly aided through the use of these
powerful mechanical aids. These aides would help the scientist
add his ideas to the record of knowledge and relieve him from
having to do mundane tasks. He would then be free to think about
those things that require creative thought.
Referencing Stored Information
After ideas have been manipulated and inserted into the record,
there is still the problem of consulting the information. The
prime action of use is selection, but there are problems in
locating information that relates to one's interest. The time
involved in searching through millions of ideas, looking for
appropriate information, may cause the scientist to not keep up
with the current scene.
There are two types of searches: simple selection and selection
by class and subclass. Simple selection involves searching all
elements of the record and matching specific characteristics.
Class and subclass selection involves searching for sequential
characteristics that continually narrow the possibilities. A
good example of this is the way a telephone exchange operates.
With each successive digit that is dialed, the number of possible
connections is narrowed and the specific phone line is selected.
The problem in finding information is caused by the artificiality
of systems of indexing, according to Bush. "When data of any
sort are placed in storage, they are filed alphabetically or
numerically, and information is found (when it is) by tracing it
down to from subclass to subclass." Bush argues that the human
mind doesn't work this way.
The mind operates by association. When one idea is in mind, the
next idea is suggested by an association of thoughts that was
forged through previous experience. In the mind, associations
that are not used often tend to fade over time, ideas are not
fully permanent, and memory is transitory. Despite this, the
speed and intricacy of associations is awe-inspiring.
Bush proposes that selection by association can be mechanized.
One can't hope to match the mind for speed or intricacy, but the
associations made would be permanent and clear. In exploring
this idea, he discusses a hypothetical mechanized private file
and library: the memex.
A memex would be a device in which an individual stores all his
books, records, and communications. It is mechanized so that it
may be consulted with exceeding speed and flexibility. The memex
is described as a desk containing stored material on microfilm,
projection screens for convenient reading, a keyboard and scanner
for the input of information, and a working surface.
Bush envisioned that material of all sorts would be available on
microfilm: books, pictures, current periodicals, newspapers, etc.
In addition, the keyboard and scanner would convert the owner's
ideas and materials onto microfilm for inclusion with the stored
library of information.
If a person wanted to consult a certain book, he would enter a
code and the book would be presented on the display screen. The
book could then be read or skimmed by the user. At any point,
marginal notes or comments may be added to the book. Also, the
index is easily accessible and other books may be consulted at
the same time on adjacent projection screens.
Bush states that the user can then build an association between
the two books being displayed on adjacent screens. By pressing a
button, an association would be made and indexed. Anytime either
reference was accessed in the future, the other could be accessed
at the press of a button. Trails between multiple documents
could be built and the net effect would be as if you edited a new
book, using parts of existing documents.
It would then be possible to produce a new microfilm document,
composed of information and the trails created by previous
associations. This microfilm could then be given to another
memex owner and used in his machine, giving him the benefit of
your notes, associations, and references.
Bush predicted that new forms of encyclopedias would appear,
ready-made with a mesh of associative trails running through
them, ready to be dropped into a memex and amplified. The user
could then access the information, adding notes and new trails
associating it with his own present library of information.
DISCUSSION
The power of Bush's ideas is amazing. Microfilm technology has
not advanced to the point where it can be used as described in the
memex, but the idea of associatively indexing information may yet
be realized through the use of computer technology. Bush's ideas
inspired the work of Ted Nelson, who began the design of a system
he called 'hypertext'.
Hypertext refers to the idea of creating and linking documents
with associative trails. The concept works much the same way as
Bush's memex, except electronic storage of information is used
instead of microfilm technology. The implementation of hypertext
has been held up by the high cost of mass storage, however.
We are finally at an age where Bush's ideas may be realized. The
cost of mass storage systems is falling quickly, computers are
being developed with sophisticated graphic displays, and the use
of telecommunications allows us to easily copy information from
place to place. All of these factors will play a role in the
success or failure of hypertext.
Optical storage systems can provide a great deal of information
for an individual computer user. With a storage capability of
540 million characters, each optical disc in a CD-ROM system can
hold the equivalent of three sets of encyclopedias. The ability
to read text and graphics data from these devices and display it
to the user is currently being demonstrated. Libraries are the
major users of this technology, allowing quick searches through
large volumes of abstracts and citations.
True hypertext cannot be implemented with this technology alone,
however, because data cannot be written back onto the optical
disk surface. In order to establish true associative links, the
user must be able to build his own trails and index other types
of documents within the system.
Current advances in magneto/optical disc technology promise to
provide a solution. These devices would allow about 500 million
characters to be stored on a single disc, with the ability to
read, write, and delete information. The marriage of CD-ROM
systems with magneto/optical devices would allow the user to make
an initial copy of the original text and then build his own
network of trails and associations, adding additional notes,
pictures, and other information to the system over time.
The importance of the computer display becomes evident here.
Bush envisioned a system with multiple projection screens that
would allow the user to view and then link documents together.
The ability to display graphics and pictures was inherent to the
microfilm media of the memex. Computer displays, however, are
not necessarily suited to the task of displaying all types of
data. In recent years, advancements in display technology are
allowing large screen displays to be built with the ability to
display both text and graphics. Graphics can be either simple
line drawings or pictures with the resolution of a television
signal; this is an important development.
The ability to display multiple documents on a screen and have
naive users understand what is happening is also important. The
development of sophisticated user environments that are intuitive
and easy to use will help make hypertext technology accessible to
a wider audience. These environments will allow multiple windows
to be opened and documents to be viewed without confusing users
or demanding a high degree of technical sophistication. This is
important if these systems are ever to become pervasive.
Bush envisioned that users of the memex would correspond with
each other and trade information by mailing microfilm that would
then be inserted into another memex. It is becoming possible to
transfer large volumes of electronic data quickly and easily.
Recent developments in high speed telecommunications equipment
are allowing computers to transfer data at increasingly higher
speeds and the impending marriage of telephones and computers
will have a big impact on the utility of hypertext systems.
It may soon be possible to correspond with another person who
owns a hypertext system through electronic messages and other
data sent directly to his computer over telephone lines that are
simultaneously used for voice communication. This would make it
possible to talk with someone while your respective computers are
busy sending data to each other.
The ability of people to quickly share information may be the
ultimate legacy of hypertext. Even if two users share the same
reference materials, the associations made by one individual can
be shared with another who may not have had the same insights.
The ability to send pictures and other graphical data will also
enhance the communication between people in society.
It is important to stress the fact that the associations built by
an individual are highly personal. Nelson envisioned a hypertext
system where mankind's entire body of knowledge would be stored
and exploited. Links and associations would be made by all users
of the system and new information would be contributed, extending
man's body of knowledge. One problem I see with this idea is
that your associations may not necessarily be mine. If all of
this data has several threads of association running through it,
a person may reach a point where he can no longer navigate the
existing trail of thoughts. Too much generalization in the
creation of associations may hinder the ability to research
topics and remain clear about one's goals.
The best hypertext system may be your own: a personal computer
with access to large volumes of data, collected over time. You
may want to contact centralized systems in order to collect new
information and add it to your personal hypertext. New data and
associations could be added, but only when you desire them.
CONCLUSION
Vannevar Bush was an incredible visionary. His ideas concerning
associative indexing of information may find realization through
the use of computer technology. Computer media can provide many
of the useful characteristics envisioned in the microfilm memex:
storing and retrieving data, viewing multiple documents at the
same time, and transferring that information to other users.
Ted Nelson's vision of hypertext is based on Bush's work. Nelson
envisions a centralized system where all of man's knowledge is
electronically stored and retrieved. Individual systems, enhanced
with communications ability, may also emerge. Personal hypertext,
providing both local storage of information and access to large
centralized volumes of data, may change the way we think and live.
REFERENCES
Bush, Vannevar "As We May Think", The Atlantic Monthly, July
1945, pgs. 101-108.
Byers, T. J. "Built by Association", PC World, April 1987,
pgs. 244-251.
Nelson, Ted Computer Lib/Dream Machines, Personal Publication,
1974.
Nelson, Ted Literary Machines, Personal Publication, 1981.
Onosko Timothy "Let There Be Light", Creative Computing, September
1985, pgs. 43-49.
--
Jon Backstrom "Yah sure...we gonna have fun, you bet!"
Computer Science Department
Indiana University ARPA: viking@iuvax.cs.indiana.edu
Bloomington, IN 47405 UUCP: {pyramid,ihnp4,pur-ee,rutgers}!iuvax!viking