A Short History of Hypertext
Tim Berners-Lee didn't invent hypertext, but he did
invent the World Wide Web. In 1989, the mild-mannered
Berners-Lee was working at a high-energy physics lab
in Switzerland. He had already written a few database
programs to store information via random links, but
those were only for his personal use. Then the idea
came to him-Click!-to create a hypertext structure
that would span the globe via the Internet, accessible
to anyone with a mouse....
Out of Order: Hypertext's Past, Present, and
Future
The Web is only nine years old, but the basic
concept behind it-hypertext-is not. Hypertext can be
defined loosely as an interconnected set of words and
images that can be "browsed," instead of
simply read in a particular, fixed order. Hypertext
can be traced back to examples from the middle ages
and to novels starting in the mid-eighteenth century.
The World Wide Web is the most fully developed form of
hypertext that is widely accessible. As a living,
breathing, hyper-textual organism, it continues to
develop in a typically messy and chaotic fashion.
Technology rarely ends up where people expect it
will. For every article, such as Vannevar Bush's
pivotal 1945 piece "As We May Think" for The
Atlantic Monthly, which foresaw today's desktop
computers, there are hundreds of articles describing a
future filled with food pills, jet packs, and
vacations on the Moon. Bush proposed that we automate
access to knowledge, before we drowned in it. In an
astounding feat of extrapolation, he predicted
microfilm (which would allow a million books to fit on
one end of a desk), digital photography, speech
recognition systems, and personal computers that could
access hypertext databases. He called his hypertext
links "associative trails," but the concept
was essentially the same as our Web. Bush's article
inspired many in the computer field, including Douglas
Engelbart. Starting in 1963 at Stanford Research
Institute, Engelbart proposed an online hypertext
system that linked and cross-referenced all the
documents in a workspace shared by users at different
physical locations.
Project Xanadu, begun in 1960 by Ted Nelson, was
even more ambitious. Its end goal was nothing less
than a "docuverse," a document universe
where all writing is linked and referenced, and where
nothing can be deleted. Once something was published,
it would exist in the document universe forever,
accreting annotations, links, and revisions. Nelson's
lofty goals have yet to be achieved with current
technology. He is viewed by some as a trailblazing
genius-and by others as a pie-in-the-sky crackpot with
a penchant for neologisms.
The desktop publishing and personal computing boom
of the early 1980s let the cat out of the bag.
Personal hypertext database programs and languages
became available for both Macintoshes and
IBM-compatibles. Electronic hypertext novels began to
appear; by choosing which links to follow, the reader
determined the order of the episodes in the overall
narrative. Hypertext as a literary form was met with
skepticism, but a few authors have gained recognition
by exploring the form beyond its "choose your own
adventure" beginnings. Hypertext literature is
now taken more seriously, and many university English
departments offer courses on it.
The inventor of the Web had a purely practical
objective: alleviate the information overload he
suffered, along with others in the field of
high-energy physics. In the early 1980s, London-born
physicist/programmer Tim Berners-Lee was working at
the CERN European Particle Physics Laboratory near
Geneva, Switzerland, when he wrote a personal
information manager he called "Enquire." It
could handle random associations by linking database
files. A new version of this program, one that could
run on any computer and be used by multiple users on a
large network (the Internet), was the obvious next
step. It would incorporate codes within the text that
would permit both text formatting and hypertext links.
In 1989, he floated a proposal for such a system.
He was met with considerable skepticism. After all,
maybe he was just another Ted Nelson or Vannevar
Bush....
"There were two years of solid pushing, of
persuading everyone against solid pressure,"
recalls Berners-Lee. "They thought it was too
complicated, that people would get lost in hypertext,
that it [interpreting the formatting tags for display]
was too slow." But he eventually won some
converts, and work began in the latter part of 1990
using students as staff. Berners-Lee wrote the first
Web browser and Web server programs on his NeXT
machine (Mosaic, developed by the folks who later
formed Netscape, would not exist until 1993).
Immediately, reality diverged from his original
vision, which assumed that users could add links to
the material they were viewing (as Bush envisioned in
1945).
For Berners-Lee, the promise of his "web"
was that users would be shielded from the inner
workings of hypertext, and that all users were created
equal-everyone could act as content providers or
editors. "You could make links very easily, in a
collaborative environment, and you never saw format
codes," Berners-Lee recalls of his vision.
"It aimed to be universal, like paper. There
would be a continuum of hypertext from personal sites
to the president's site and everything in between, a
place where people could build a common knowledge
space," he says. But this proved impractical, so
the system had to be broken into separate
"server" software, which stored and allowed
access to Web files on Internet computers, and
"browser" software to display the files on
user's desktops. More and more people have set up
their own home pages, but the basic distinction
between the Web server and the Web browser is still in
place. Links still only work in one direction-if I'm a
fan of rock musician Warren Zevon, I can add a link to
his official page on my home page. This does not mean
that Zevon's page will feature a link to mine,
however. The basic power hierarchy of the real world
is maintained, which is counter to Berners-Lee's
original democratic vision.
By the end of 1990, the Web consisted of one server
at the CERN lab in Switzerland, and one file, the CERN
phone book. Today, the number of files is approaching
30 million, and there are more than 200,000 servers.
Berners-Lee now is married, has two small children,
and has little time to surf the Web. "I guess I
am proud of it," he says of his creation.
"But now I worry about the negative aspects. The
Web is the great equalizer since everyone has access
to the same information, but you do have to have
access to it. If information providers begin relying
on it, other sources, such as 800 numbers, may be
turned off.
"And I expected to see a more heterogeneous
mix on the Web. Instead of carefully edited papers, I
expected to see groups and families using hypertext to
keep track of what they were doing from day to
day," Berners-Lee adds. As the father of the Web,
he now seems both perplexed by how his baby has grown,
and worried about its future development. It is
doubtful that he could have imagined the Pamela
Anderson Lee Fan Club home page when he was carving
out the initial specs for the Web.
Today, he is the head of the World Wide Web
Consortium in Cambridge, Massachusetts. The
organization acts as a sort of steering committee for
the Web, proposing and maintaining standards in the
face of increased pressure from companies seeking to
dominate the Web with proprietary standards.
His current wish list for the future of the Web
includes: interactivity, so a user can add content to
another's files (his original vision); "Seals of
approval" to vouch for accuracy, or suitability
for children; more automation, with "software
agents" that would act as souped-up, personalized
versions of today's Web search engines.
However, the future rarely turns out the way it is
planned. Today's buzzword is tomorrow's has-been. But
hints of where hypertext might yet go -- among future
writers, educators, programmers -- abound. After all,
in a world where a database program written by a
modest scientist in Switzerland evolves into the World
Wide Web, anything can happen.