In my senior year at the University of California, Berkeley, I visited a dormitory room decorated so unconventionally that I still remember it decades later. The room belonged to fellow student Rick Prelinger, who had covered his walls with topographical maps---detailed maps of the American terrain produced by the United States Geological Survey that are popular with hikers and hunters. Displaying a "topo" map wouldn't by itself be that unusual if it were of, say, Yosemite National Park. But Prelinger's maps covered California's flat, rural Central Valley, an area better known for cotton fields than waterfalls. The maps had names like "Avenel" and "South of Kettleman City," indicating that the area's most prominent landmark was in fact off the map. One map in the region is entitled "The Dark Hole." The maps' lack of landmarks, their detailed chronicling of an almost feature-less landscape, was the very reason Prelinger liked them. As I say, he had unconventional tastes.

Rick moved to Boston and except for the exchange of a few postcards, we lost touch. Then, some years later, I came across his name in a newspaper article in association with something called the Prelinger Archives. Over the years working in New York City, Rick had amassed a collection of what he calls "ephemeral films"--- advertising, educational, industrial and amateur films, none of which were ever intended for the local movie theater, let alone immortality. The collection, which now numbers 45,000 films and more than 100,000 reels, is housed in a warehouse in the meat market district of Manhattan. The collection includes such obscure works as "Classroom Discussion of Marijuana (early 1960s), "White Wonder" (from the Morton Salt Company), "Handymen of India" (produced by the Indian Army in 1941), and a 1925 home movie: "Tokyo Japan and Trip to Nikko by Train." Whatever the Prelinger Archives is, it's the largest collection of its kind.

Perhaps the Prelinger Archives (www.prelinger.com) could only have been invented by its namesake. If Casablanca and Star Wars, Kurosawa and Truffaut, are the Yosemites and Grand Canyons of the screen, these obscure works are the South of Kettleman Cities, the Dark Holes. Rick had put together a career doing what he always had done, looking intently where others don't. Where a casual observer sees background, terra incognita, Prelinger sees content and thinks that content worth preserving.

Rick calls his collection ephemeral films, and calls himself a "media archaeologist." His collection focuses on what I'd call "peripheral information"---information that resides in the corner of your eye, or perhaps entirely outside your field of vision. The Web has become the all-pervasive example. On the Web, most people pay attention to the portal sites, the Yahoos and Amazon.coms. But the Web is much bigger than that. It is like a monstrous library, where a book might get a few specialized readers stopping by each decade, blowing off the dust, and moving on. Prelinger and a few others are saying that such peripheral information---seemingly specialized or trivial---may turn out to be vital. Because we're swimming in information, we don't really know what's important. We're too close, and there's too much of it to assess.

According to a study conducted by Peter Lyman and Hal R. Varian at UC Berkeley's School of Information Management and Systems, the world produces between 1 and 2 exabytes---1-2 billion gigabytes, of unique information per year. That's about 250 megabytes for every person on earth, accessible by anyone with access to a wired computer. Much of this deluge is created by individuals, a phenomenon the researchers call the "Democratization of Data." And much of it is now "born digital," that is, created in digital format. That's especially true for text, and increasingly so for images.

Prelinger writes that his archives document both familiar and obscure aspects of twentieth-century culture and society: life, leisure, history, industry, technology and landscape. Unlike more commercial films, ephemeral films often feature real people at work. Sometimes, they reveal hidden truths about how we as a society viewed ourselves, what we valued, what we feared. Some of the films are so clunky, they can now pass as entertainment. Other films, dispensing advice on, say, table manners and dating to American high school students in the 1950s and 60s ---are both funny and painful to watch. Funny because they are so naive, and painful because they captured some hard truths about the way we were. Some films have become unbearably offensive---in the way they deal with race and culture---in ways their makers could not have imagined.

Not surprisingly, two of the Archives' principle constituents are mainstream television and movie producers, and scholars interpreting 20th century culture. Prelinger himself ha combined both roles. He has worked as a researcher for HBO, a premium television network, on a documentary on standup comedy. And he worked with the Okinawa Historical Film Society on a project to collect footage of World War II's Battle of Okinawa---"to reconstruct the deeply-felt and, in many ways, unfinished story of the war." (The project was never completed.)

Prelinger has another Japanese connection, as well. In 1998 and 1999, he hosted a show on Sony's Vaionet digital satellite system called The Museum of Film Archeology. The 20 episodes featured films from his collection, with Rick visiting places in America that were related to the ideas embodied in the films. Segments covered everything from aviation to animals, and gave a view of this country many others, including American citizens, would have overlooked.

Moving pictures online: $7K/terabyte

Prelinger is now moving a small subset of his collection online with the creation of the Internet Moving Image Archive. The project, which brings ephemeral films to the world's most ephemeral medium, will put eventually 1001 films on the Internet in MPEG-2 and MPEG-4 formats, free of charge. (MPEG-4 yields much smaller files that produce lower quality images.) Most material is in the public domain, and Prelinger Archives owns the rest. The site is co-sponsored by the Internet Archive (www.archive.org), whose founder, Brewster Kahle, I profiled in Pacific Connection in November 1997.

Kahle says that the Internet Moving Image Archive represents a kind of library of the future---a repository not just of books, but other media, whose collection is accessible throughout the wired world. The economies of storage will continue to make such libraries possible. According to Kahle's rough math, the Library of Congress---the largest collection of hardcopy textual information ever amassed---contains about 17 million books. Round up, and that's a 20 terabyte storage requirement, an amount the Web surpassed last year. But what does that actually cost to store? Kahle figures its about $7,000 a terabyte, including the cost of a server capable of parceling it out. So the entire Library of Congress could be put online for $140,000. "That's not pocket change, but not a big deal, either," he says.

But why bother? How much information do people need to get their hands on? Kahle argues that ultimately, the Prelinger Archives and the Web pages of the Internet Archive represent a collective memory. "We need a memory; without that, there's no concept of culture."

The MPEG-2 files that comprise most of Prelinger's online collection require just half a terabyte, or $3000 to put online---a small fraction of the cost of digitizing the material in the first place. And the costs of storage are dropping by about half each year. "The disk guys are on super-Moore's law," Kahle says, referring to the doubling of processing power every 18 months. Disk capacity has grown so fast that the Internet Archive project no longer uses tape to archive the Web. Everything is on disk and online.

"We've always had to whittle back the amount of information we can search for due to the constraints of print," says Kahle. "Now copyright is the only constraint for universal access." In the U.S., copyrights are now approaching 100 years, due in large part to the relentless campaign of corporate copyright holders. The Walt Disney Company is not going to sit quietly while Mickey Mouse goes into the public domain. As virtually everything becomes copyrighted, Prelinger and others have argued for a "copyright conservancy," in which copyright holders donate some of their intellectual property, creating the intellectual equivalent of a national park. In the meantime, most films in Prelinger's Internet Moving Image Archive are already in the public domain.

While the Archives' collection may be obscure, the films have already been discovered by the Web cognoscenti. "We've gotten 'slash-dotted," Rick says, referring to the website Slashdot.org. Since then, the site has gotten thousands of requests from downloads by people Prelinger and others have called the 'bandwidth privileged"---people, including yours truly, who have access to high bandwidth accounts. (High bandwidth is almost a prerequisite for downloading massive MPEG-2 files.) And why MPEG-2? "Streaming is for sissies," Prelinger says, because you never get possession of the material. "When we let people download, they have the movie and can do whatever they want with it. It's an open question whether the ubiquity of this material will increase the demand among people who are willing to pay. I think it will, but nobody's ever given away this much material before."

Transforming these films to a digital format is not easy. For one thing, the formats vary. They include 35mm, 16mm, 8mm, Super 8mm, as well as obsolete formats like 28mm and 9.5mm. The selected films are first transferred to Betacam SP videotape on telecine machines. The transfer process entails considerable inspection and repair, as well as supervision by a skilled operator who manipulates color, contrast, speed and video controls. The videotape masters are then digitized on a 550MHz PC using a FutureTel NS320 MPEG encoder card. Custom software drives the Betacam SP playback deck and manages the encoding process. The resulting files are then uploaded to hard disk. Playback requires either an MPEG-2 decoder board or software. Of course once transfer is complete, the films reside on disk---a much less vulnerable medium. Digitization has thus made at least a few "ephemeral films" less ephemeral.

Prelinger's foray into the Internet provided a good excuse for a small class reunion, and so one day last February, we met at Javaholics, a coffee house near Rick's home in San Francisco, to talk about his work and its meaning in this information-inundated world. I was particularly interested in industrial promotion films, the kind shown by the hundreds at computer trade shows and increasingly made available on corporate Web sites. I ordered the house brew and Prelinger an orange juice---and we raised our voices to be heard over the cappuccino machine. Here are excerpts:

What's the relationship between ephemeral films and the Internet?

The Internet has brought corporate communications and commercial speech to a new level of pervasiveness because now rather than just having to go out and find an audience, the audience is already there. Beginning around the turn of the century, corporations in America started making it a priority to communicate with consumers, members of the general public as well as employees. In the late 19th and early 20th century corporations had a bad name, in part, because of labor struggles and concentration---keiretsu-style---into anti-competitive trusts, and so they struggled not just for profit but for popular legitimacy. Companies like National Cash Register, General Motors, International Harvester and Radio Corporation of America became concerned about having a good reputation. There were many ways that films got to people in those days. Primarily they were provided for free to schools, community groups, farm groups, women's groups, to workers at lunch time, and even to union groups. When TV started in the late 30s, some were shown there. A film called "The Romance of the Tin Can," didn't have any direct advertising in it, but contained some useful quasi-educational material. It's main objective was to have viewers think well of tin cans. Industrial films have now passed into the video world and are still very much with us. There are hundreds of thousands of corporate videos made every year, some for use within a company, and some for use outside a company.

Despite their numbers, these films are very similar in style.

Yes, the vernacular has shifted a little bit but in a lot of ways it's still fundamentally the same. They just use a different representation tool to make the point. Now travel companies and real estate companies will send a video to your house by direct mail, and a lot of luxury goods are marketed by video.

How have the films by technology companies changed?

Just like the old futurist films were made in the 30s, 40s and 50s, we now have visions of the future as enabled by Sun or Dell or Microsoft. The vision of the future is remarkably static. For example, kids now relate to their parents through telepresence pretty much the same way as in the old films, where they would look at their parents through a videophone. Women are cosmetically executives in the new videos, but they still also do housework, like in the old movies. Today, companies have proven we can work much more efficiently and stay closer to our jobs with all this new enabling technology, but they haven't resolved the leisure issue. Leisure doesn't count any more. In recent years corporations have found other ways of reaching people, like corporate sponsorship of golf tournaments. And so the Web is the latest stage of an effort that's almost 100 years old.

You invented the term "ephemeral films" for films that aren't suppose to last. Are Web pages the latest version of the phenomenon of ephemeral information?

Yes. That's why the Internet Archive collects the Web. Brewster Kahle has amassed 40 terabytes of Web pages, and he cites studies that the average life of a Web page is 77 days.

I still remember your collection of obscure topographical maps. What's the connection between them and your film archive?

The maps are part of a lifetime interest in landscape and geography. I'm especially interested in what I'd call supplemental spaces, spaces that border other spaces, but don't have a name or an explicit use. In the last 10-15 years my biggest interest has been in understanding how people relate to space, how they use and relate to landscape. That's why I started collecting film. The first film I collected was a pedestrian safety film made in Oakland [California] in the 1940s. It's shot in live traffic, and in the periphery of the film is an incredible documentation of what Oakland looked like in the late 40s.

Can we understand this material in the present-day, or must be viewed from a historical context?

The films have many different contexts. But over time, new contexts accrue as the world changes. Having the perspective of time and space distance gives these films a new context. Also, the act of finding something again, one becomes identified as its finder and by repropagating it, gives a film another life.

When did companies switch from film to video?

People still make film, but in the 1980s the real transformation occurred. Our collection begins with 1903 and goes through the 80s. People still make movies, mostly high-end industrial stuff.

Why have you put some of these films on the Web?

Ordinary people have never had access to primary historical materials. Most people can't walk into a film archive and see a film, or go to a commercial place without paying a screening fee. We're really focused on access. What will happen when a canonical selection of moving images is available to anybody to do whatever they want with? You open up the possibility for people to become authors in the historical sense, not just to work with material they shot themselves. Teachers and students, for example, can make their own comments on history.

The Web has become famous for hosting promising ideas that don't make money. How will your project support itself?

The only way the archives make money is by licensing footage. So the Internet project is a total gamble. It could be killing the goose that laid the golden egg---because this stuff is going to be out there for a long time. But I have a feeling that if more people could see and work with ephemeral films, the cultural demand will increase.

Your career path is very American in that you invented your own job description; you invented your own field.

I did, and it's harder in Japan, although there must be so much opportunity in Japan because there are so many needs that are not being fulfilled by established corporations. The question is will people buy something that doesn't have a brand name. But I think there is great opportunity in Japan for goods and services that are not handled by large corporations. Japan is the almost world leader in small business.

You could have made your archive a subsidiary of another institution.

The archive is so big, it's hard to find an institution that can assimilate it. It's bigger in itself than most archives in their entirety. But ultimately I think the collection needs to be placed with a public institution. I'm one person, and have a small company---three part time workers in New York. There's no way I can insure that this collection will be preserved and taken care of adequately. Somebody should spend millions of dollars a year doing cataloging and preservation.

What are you doing in San Francisco when the Archive is in New York?

I moved to San Francisco two years ago, got married and stayed. I haven't been to New York since last June. The digitizing is done in New York and I get the disks out here, where they get uploaded to the Internet Archive. We have no Internet relationship with ourselves, just telephone and fax. I'd love to be closer to the films but we couldn't move all that stuff out here. It needs to be close to the company that represents us. When they ask us for a film, for a duplicate of a segment, they need it in an hour. It's still in New York's meat district, but in a much bigger room. We have 30,000 cubic feet there. Living in San Francisco is not a business advantage per se, but it's a beautiful place, healthy, and close to nature. I look at birds, collect mushrooms. I'm working on a movie now about Americans' relationship with risk.