Broad Band

by Claire L. Evans

  In a way, it’s fundamental. The Talmud is a hypertext, with layers of annotations arranged in concentric rectangles around a theological heart. Any text referencing another is considered a form of hypertext: sequels, which begin where the last page of the previous book left off; footnotes; endnotes; marginalia; and parenthetical asides. Sprawling, self-referential novels like Ulysses or Finnegan’s Wake are like flattened hypertexts, and scholars love to cite “The Garden of Forking Paths,” a short story by the Argentine writer Jorge Luis Borges, as the height of precomputer hypertext. “This web of time,” Borges wrote, “the strands of which approach one another, bifurcate, intersect or ignore each other through the centuries—embraces every possibility.” He may have loved the World Wide Web.

  The Web as we know it isn’t modeled on Borges, Joyce, or the Talmud. The most famous hypertext pioneers are men—Doug Engelbart, Jake Feinler’s mentor at Stanford, incorporated hypertext into his oNLine System, and Ted Nelson, a Bay Area counterculture hero, coined the word and has championed utopian hypertext ideas for decades—but the Web appeared on the scene only after hypertext principles and conventions had been explored for nearly a decade by brilliant female researchers and computer scientists. They were the architects of the hypertext systems that time forgot, systems with names like Intermedia, Microcosm, Aquanet, NoteCards, and VIKI, the earliest ontological frameworks of the information age. Hypertext is, in many ways, the practice of transforming pure data into knowledge. And like programming a generation before, it was where the women were.

  MICROCOSM

  To understand hypertext, I’ve turned to one of the brightest computer scientists in the world. Dame Wendy Hall is a garrulous, strawberry blonde Brit with a disarmingly warm manner and a busy calendar. We’re talking over Skype, nine hours apart. Wendy, who was appointed Dame Commander of the Order of the British Empire—the female equivalent of being knighted—in 2009 for her contributions to computer science, is in a hotel room in London, dressed for dinner. I’m in my pajamas, drinking coffee, surrounded by index cards, in my office in Los Angeles. For reasons I don’t yet understand, she has chosen this moment to catch me up on medieval European history.

  The Battle of Hastings, to be precise. “It’s something we learn about in history,” Wendy tells me, unsure if news of William the Conqueror’s eleventh-century triumph ever made it stateside. William the Conqueror earned his nickname by invading England from its southern coast and defeating the last Anglo-Saxon king, Wendy explains, as I indulge this diversion. A few years later, he decided to take stock of his spoils, the entire Saxon kingdom. He ordered an audit of everything he owned. “Every cow, every sheep, every person, every house, every village, everything,” she says. “They went ’round, by hand, counting everything up.”

  The result was an unusual book, now invaluable to historians, detailing the minutiae of the Saxon world, the only survey of its kind. Because the judgments made by the Norman assessors who compiled it were supposed to be definitive, native English people called it the Domesday Book, Middle English for “Doomsday Book.” As the British cleric Richard FitzNeal wrote nearly a century later, decisions made in the Domesday Book, “like those of the Last Judgement, are unalterable.” The Domesday Book is what led Wendy Hall, in a circuitous way, to her career creating hypertext systems long before the dawn of the Web.

  In 1986, as Wendy was beginning her teaching career, the BBC—“that’s the British Broadcasting Corporation,” she reminds me, gently—celebrated the nine hundredth anniversary of the Domesday Book by updating it for the modern world, issuing a new British census on a pair of multimedia video LaserDiscs, then the height of technological sophistication. They called them, of course, the Domesday Discs. More than a million people contributed to the project, which became a massive volunteer time capsule encoded in bits and light. “Every school in the country was asked to send in three photos of their area,” Wendy explains. Schoolchildren wrote accounts of their day-to-day life; Britons sent in photos of office parks, pubs, and windmills. One child, from the village of Spennymoor, contributed this spot-on prediction of the future:

  Robot limbs will be used when natural limbs are lost.

  Computers will take over much of the diagnosis now made by doctors.

  Food will be made tastier by artificial means.

  Children will learn chiefly by computers.

  This crowdsourced survey made up the first of the two Domesday Discs; the second was filled with interactive material about British heritage, government, and royalty, including census data and some early virtual reality–like tours of notable sites. When Wendy saw the Discs for the first time, they blew her away. It wasn’t the information that impressed her as much as the way it was displayed. “The ideas were stunning,” she tells me. The Domesday Discs were interactive, using interconnected links that could be navigated with a cursor, much as we’re accustomed to doing on the Web today. For Wendy, moving with ease from first-person reflections on British life to census data and 3-D photo tours was a rich, rewarding, and immersive experience. She’d never seen a computer do anything like it before.

  Of course, she’d never been much interested in computers. Although her alma mater, the University of Southampton, was one of the first schools in the United Kingdom to teach computer science, Wendy was a student of pure mathematics. According to her doctoral adviser, Wendy was in those years a “shy and retiring student,” working in an area of topology “so obscure that to this day I can’t understand the title of the thesis.” She learned some programming in a first-year course but found it tedious and impersonal. “I was happy in my world of mathematics, and really didn’t see, then, that computers would ever really offer me anything,” she told a radio interviewer in 2013. But once she saw the Domesday Discs, she overlooked her distaste. Suddenly, she understood what kinds of experiences computers could make possible, and as personal computers began to appear in the United Kingdom, “I began to see the future,” she says.

  Not everyone saw it as clearly. When Wendy returned to the University of Southampton after a stint teaching mathematics to trainee teachers, she accepted a lecturing position in computer science, but her enthusiasm for multimedia was out of step with the established views of the department. “One professor told me in public once that if I carried on doing this multimedia work, there was no future for me at Southampton or in computer science,” she remembers, “because I wasn’t writing compilers, or new programming languages, or doing operating systems.” Many of her colleagues didn’t consider interactive multimedia to be real computer science—it was seen as something fluffy, less serious, far closer to the humanities than to classical programming.

  But Wendy couldn’t shake the glimpse of the future she had seen: a future where images, texts, and ideas were connected through intuitive screen-based links, and computer screens were approachable to all. In 1989, she left Southampton and took a job at the University of Michigan, where she immersed herself in American tech culture, went to conferences, and finally learned that clickable multimedia on computers was indeed a serious discipline, and that it had a name: the Americans called it “hypertext,” or “hypermedia.” Although her interests seemed out there to most of her British colleagues, she was right at the cutting edge in the United States. She returned to Southampton with a blinding vision for a new hypertext system. In order to explain it, she takes me back in time—again.

  This jump is only across decades, rather than centuries. Wendy tells me about the Earl of Mountbatten, a second cousin of Elizabeth II. Mountbatten is something of an avatar for twentieth-century Britain. As last Viceroy of India, he oversaw the country’s transition into a modern republic. He captained a naval destroyer during the Second World War, and was appointed by Churchill to Supreme Allied Commander of the Southeast Asian theater, where he oversaw a bloody Burmese campaign under monsoon rains. He met Stalin. He met Emperor Hirohito. However, his prominence in British colonial history m
ade him a target. In the summer of 1979, long after his retirement, as Mountbatten was lobster potting from a wooden boat off the coast of County Sligo, he was assassinated along with his family by the Irish Republican Army. They bombed the boat to pieces, blowing his legs almost clear off.

  The Mountbattens lived in Romsey, a market village so old the medieval Domesday Book made note of its three water mills. According to the schoolchildren who surveyed the village for the Domesday Disc project, among the most salient features of life in Romsey in 1986 were a preponderance of punks (“Their spare time, of which they have a lot, is spent hanging around with friends, sometimes playing space-invader machines, with money from dole”), a beloved fish shop, a Waitrose supermarket, and Broadlands, the Palladian estate where the Earl of Mountbatten entertained royal visitors. Broadlands sits on the River Test, which only ten miles south flows past Southampton to join the briny waters of the English Channel. As it happens, the University of Southampton is known for its archives—which is why, after his violent assassination, the remains of the Earl of Mountbatten’s considerable life ended up in its library.

  At Southampton, the Mountbatten archive joined millions of manuscripts, but it was distinctly modern in comparison to most of the library’s holdings. Because Mountbatten’s public life transected all the touchstones of twentieth-century media, the library inherited some fifty thousand photographs, speeches recorded on 78 rpm records, and a large collection of film and video. There was no linear sequence to the material, save chronological order, and no hope of fitting it neatly into a database. Ten years later, Wendy Hall returned from her hypertext sabbatical in America.

  Not long after she’d settled back in, she heard a knock at her office door. Word of her interests had drifted from the computer science department to the library. “The archivist came to see me,” Wendy remembers, “and he said, ‘Couldn’t we do something wonderful? I’ve got this multimedia archive, it’s got pictures, it’s got film and it’s got sound. Couldn’t we put it on a computer and link it all together?’ And that was the beginning.”

  The Mountbatten archive was the perfect test case for a hypertext project: a vast, interrelated collection of documents spanning many different media, subject to as many readings as there could be perspectives on the last century of British history. Wendy put together a team, and by Christmas 1989, they had a running demo for a system called Microcosm. It was a remarkable design: just as the World Wide Web would a few years later, Microcosm demonstrated a new, intuitive way of navigating the massive amounts of multimedia information computer memory made accessible. Using multimedia navigation and intelligent links, it made information dynamic, alive, and adapted to the user. In fact, it wasn’t like the Web at all. It was better.

  Microcosm’s core innovation was the way it treated links. Where the Web focuses on connecting documents across a network, Wendy was more interested in the nature of those connections, how discrete ideas linked together, and why—what we would today call “metadata.” Rather than embedding links in documents, as the Web embeds links in its pages, Microcosm kept links separated, in a database meant to be regularly updated and maintained. This “linkbase” communicated with documents without leaving a mark on any underlying document, making a link in Microcosm a kind of flexible information overlay, rather than a structural change to the material.

  Wendy Hall demonstrating Microcosm in her research lab in the Department of Computer Science at the University of Southampton.

  To use one of Wendy’s examples, say I’m browsing the Mountbatten archive using her system, Microcosm, circa 1989. I’m interested in Mountbatten’s career in India, a two-year period during which he oversaw the country’s transition from colonial rule to independent statehood. This history has its recurring characters: his field marshal, the leader of the Indian National Congress, Jawaharlal Nehru, and of course, Mahatma Gandhi, whose name is everywhere in the source material. Say also that within the Microcosm linkbase, an instance of the name “Mahatma Gandhi” has been linked to some multimedia information—a video, perhaps, of a Gandhi speech. Because of the nature of Microcosm links, that connection isn’t isolated to a single, underlined, hyperlink-blue instance of those words. Rather, it’s connected to the idea of Gandhi, following the man wherever his name may turn up, across every document in the system. Further, if I were to bring a new document into Microcosm, the system would automatically identify any words corresponding to links in the linkbase and update it accordingly. Imagine the analog on the Web we know today: for every name, for every idea, for every linkable thing, a single repository of supplementary material, updated by everyone in the world, filtered based on parameters determined by the user.

  Links in Microcosm could be tailored to the user’s knowledge level and could point to several places in the linkbase at once. Microcosm was even able to dig up new links on the fly by running simple text searches on all the material in the system—a prescient design that anticipated the importance of search in navigating information. This “generic” linking, in concert with linkbases, created a system that could adapt to its users while presenting them with more opportunities to learn. “Links in themselves are a valuable store of knowledge,” Wendy explained. “If this knowledge is bound too tightly to the documents, then it cannot be applied to new data.” Which is to say: where one instance of a connection might be interesting, multiple instances, expressed laterally, look more like truth. By making space for this generic knowledge, Wendy’s system placed value on the association between documents, rather than on the documents themselves. To hypertext’s small but active community of scholars, this is what the field was all about.

  In the years between 1984 and 1991, a flurry of hypertext systems like Microcosm emerged from universities and from research labs at technology companies like Apple, IBM, Xerox, Symbolics, and Sun Microsystems. Each suggested different linking conventions, spatial associations, and levels of micro- and meta-precision over contained corpora of information. If this sounds dry, remember that managing, navigating, and optimizing information is a central pursuit of modern life—we do it fifty times a day before breakfast—and that each one of these systems had the potential to become as important to us as the Web is today.

  The young discipline of hypertext was heavily populated with women. Nearly every major team building hypertext systems had women in senior positions, if not at the helm. At Brown University, several women, including Nicole Yankelovich and Karen Catlin, worked on the development of Intermedia, a visionary hypertext system that connected five distinct applications into one “scholar’s workstation,” and invented, in the process, the “anchor link.” Intermedia inspired Apple, which had partially funded the project, to integrate hypermedia concepts into its operating systems. Amy Pearl, from Sun Microsystems, developed Sun’s Link Service, an open hypertext system; Janet Walker at Symbolics single-handedly created the Symbolics Document Examiner, the first system to incorporate bookmarks, an idea that eventually made its way into modern Web browsers.

  For women interested in the nature and future of computers, hypertext was far more collegial than other areas of computer science, which was at that time seeing a rapid decline in female participation at both the academic and professional level. The reasons for this aren’t cut-and-dried, but they reflect some of the same tendencies at play over previous generations. While “the whole foundation of hypertext is collaborative,” suggests Intermedia’s Nicole Yankelovich, and “collaborative work appeals to women,” hypertext was also, like programming before it, an entirely new field, a clean slate upon which women could mark their place. Further, hypertext was open to scholars from outside computer science departments, who emerged from such wide-ranging disciplines as interface design and sociology. What these people shared was a humanistic, user-driven approach. To them, the final product wasn’t always software: it was the effect software had on people.

  But I don’t really get that until I start talkin
g to Cathy Marshall.

  NOTECARDS

  Cathy is a hypertext researcher who spent most of her professional career at Xerox PARC, a Palo Alto think tank founded by the printer company in 1970 to help invent the paperless office of the future. “I have to ask you a little about your process,” she says, in our first interview.

  Interviewing hypertext researchers entails its share of going under the microscope: many of them have never shaken their professional interest in how people organize their thoughts. The second time Cathy and I talk, I find myself describing to her my office pin board of index cards and serial killer–like yarn threads. She shares her own approach. “If I write something and if it doesn’t work, I’ll throw out the whole thing and start again,” she tells me. “I don’t think you lose what you’ve written. It’s still in your head. Over time, what you’re doing is changing what’s in your mind—what’s on paper is just incidental.” She makes this comment offhand, but the insight knocks me out. That’s what software is, I realize: a system for changing your mind.

  Cathy grew up in Los Angeles and was one of the first women to attend CalTech, after it went coed in 1970. She was only sixteen, small—under five feet on a good day—and mildly averse to math and science; the summer she matriculated, she was more interested in The Rocky Horror Picture Show, J.D. Salinger, and macramé than differential equations. When she asked for help, she remembers one professor who told her she’d be better off as a housewife—“my housewifing skills are even worse than my math skills,” she thought at the time—and another who waited until the day she missed class to tell a dirty joke. A fellow student filled her in on the punch line, which she still remembers, something about a whorehouse keeping down the “fucking overhead.” It wasn’t the joke that bothered her—it’s that the professor waited to tell it. “I felt so conspicuous and weird,” she says.