A couple of weeks ago a video of Vice President Dick Cheney from 1994 came up on YouTube, with Cheney saying that invading Baghdad would invite a quagmire. I investigated this on my own and discovered that, while I could find it today via the C-SPAN archives, it wasn't clear that someone in 2003 during the run-up to the war could have, as C-SPAN's advanced search service didn't exist then. The actual video only went online this past July after its first rebroadcast since 2000. Otherwise, a diligent researcher would have had to have sent $30 to C-SPAN to get a copy of the tape in order to check whether there was something interesting on it.

As I dug further into this research, I wondered how difficult it was to find newsworthy archival news footage. Where are the decent sources, and how good are online archives? Nascent services such as Truveo and Google Video have a long way to go, and old line commercial services force you to take a laborious process to get footage. Plus, academic efforts to archive news footage are still struggling to reach critical mass.

Old Cheney Videos

I decided to learn more about archival video footage online by tracking down the source and subsequent YouTube uploading of the recent Cheney video from 1994. New York-based multimedia artist Fresh was the one who found that Cheney video that's been replayed all over the media-sphere.

Fresh regularly scans TV schedules, and figured that the "Life and Career of Dick Cheney" on C-SPAN3 might be interesting. After he caught it on his DVR, ripped it to his computer, and uploaded it to YouTube, a lot of other people found it interesting, too. The video has been viewed more than 800,000 times and counting.

Fresh had uploaded some 35 videos to YouTube over the last two months, but this time he covered the C-SPAN logo with that of his own website. I asked him why he'd done that, because it seemed a bit unfair. He told me via email that he hadn't done it out of malice; it was merely a "mock station identification," much in the way that "The Daily Show" or other comedy programs would do it. Besides, he said, he readily told any news organization which asked where he had found the video. C-SPAN requested that he at least update the description on YouTube if he couldn't upload a replacement video with their logo.

I watched the clips online and took screen captures to understand just how the video was used. Most news shows referred to it blithely as a "YouTube video" (and MoveOn.org ripped their own clip as a fundraising lure). Only a couple of reporters took the trouble to identify the source directly. Abbi Tatton of CNN and Keith Olbermann on MSNBC made clear that the video was, in fact, originally broadcast on C-SPAN. (Olbermann in 2004 explained his penchant for direct sourcing in the Online Journalism Review.)

C-SPAN has been rigorous in protecting its content and name, taking on sites that posted the video of Stephen Colbert from the White House Correspondents' Dinner, and stopping Chris Dodd's use of the word D-SPAN on his campaign site. So what did C-SPAN think about Fresh's use of the Cheney video?

"We want people to actively participate in the democratic process, but those are our cameras, and that's our footage," Peggy Keegan, a spokeswoman for the network, told me. I asked whether they saw any copyright violation, and if they'd taken any action. "No, it was only a minute's worth of video out of an hour program, so it's covered by fair use. But fair use does not translate to no attribution."

Incidentally, C-SPAN did end up uploading their original version of the video on YouTube, though, with a slower frame rate, it's nearly unwatchable. The derived footage is what most people saw, and in most people's minds it became the next "YouTube video." None of the news websites even gave links to the direct URL where viewers could watch the whole video, or even perchance take a wondrous walk through the archives.

In the C-SPAN archives, I found the original Life and Career of Dick Cheney, which said that the program had been viewed 32 times (a couple of days later the number had mysteriously rolled back to 15). I watched the other videos that were rebroadcast this past July, including a symposium at Hofstra University titled Desert Shield and the Gulf War. Before I clicked on it, it was viewed zero times. I jumped through the list of speakers to get to Cheney. At the end, he made a curious point regarding democracy during wartime which I figured was noteworthy, saying that they would have taken the nation to war in Iraq in 1991 even if Congress hadn't given President Bush the authority. Here's the full clip:

The Trouble with Video Search

As for finding other valuable archival video clips, many online journalists are unaware of where to look for important news footage online. While text search and cut-and-paste of text is pretty straightforward to anyone using the Internet, the same work in video takes considerably more time. Most journalists would rather spend their time honing their argument instead of trying to locate and edit a video.

Google's video search is a weak option because it's heavily weighted by the prevalence by user-generated content and entertainment clips, sprinkled with the occasional video ripped from a news network.

"We've come to expect Google web search to be agnostic, not explicitly favoring content from any hosts," said Mary Hodder, CEO of Dabble, a startup that now focuses on video search. "But Google video search appears to be favored to their own hosted videos."

Entrepreneur/blogger Mark Cuban made a similar point in February: "Right now Google video searches itself and YouTube. That's it. If a video is anywhere else, according to Google video search, it doesn't exist. If video continues to become such an intrinsic media type on the Net, how can Google continue to be a leader in search if they don't search other sites?"

Of the relevant top 60 search results for Cheney 1994 on Google video search, all but one of the clips are hosted on YouTube/Google (the other was from Crooks & Liars). None of the derivative clips that aired on network TV were sourced to its respective website; they were only linked to YouTube.

Last week Google News announced it would start to include video in Google News searches directly from news networks. This may have been a response to AOL's re-launched Truveo search engine. Truveo indexes video content from news networks, including even local news stations like WBZ-4 here in Boston (it already has 16,000 news items). It does not, however, group items by original broadcast date/time. Instead, Truveo has devoted a large section of their website to allow developers to design their own presentation for the videos.

There is, in fact, a video archive of at least the major networks, and it has been compiled since 1968, when Vanderbilt University established the Television News Archive. They have taped and catalogued the nightly news every day of the year since then. In 1995, they added CNN; in 2004 they added Fox News. In addition to the evening news, the archive has taped special news programs, from the Watergate hearings to the Presidential conventions, totaling some 30,000 broadcasts and 900,000 individual segments. Large as it is, there are still missing elements: Television magazines such as "60 Minutes," C-SPAN, and local news shows are not included.

A couple of years ago, the VTNA introduced a browsable interface with the design elegance of Craigslist. This main page shows every month since 1968; each click brings up a list of every day of that month, showing the news programs recorded with individual segments broken out (including commercials). This design also has the key quality of being crawlable by search engines, as Marshall Breeding, the library's Director for Innovative Technologies and Research explained. The archives are a week behind, but it still has 2,000 weeks. Once you find a program, you can retrieve it by visiting the university or paying for a tape to be copied and mailed.

Thanks to an NEH grant, the Archive digitized the evening news collection, an effort which was largely complete by September 2005 Unfortunately, that hasn't yet been translated into public access. The video can now be streamed on-site at Vanderbilt. They did reach an agreement with CNN to stream that network's content to other educational institutions. "We're working on agreements with other networks, but those may or may not come to pass in the near term," Breeding told me via email.

Meta-Catalogs Footage.net and MIC

You might have better luck in the near term turning to a commercial footage house. In 1994, John Tariot, a video producer working in Boston, created Footage.net, which became the premier online search service for film and video. There are 13,000 unique users making 15,000 searches a month and 400 "Zap Requests" -- a direct research request sent to 60 archival partners (each of whom pays Footage.net to be included in the directory).

The searches are still rough; there is no ability to filter by date. A search for "Dick Cheney" brings up an impregnable stack of results from the three major news archives (ABC: 2,757; NBC: 2,562; CNN ImageSource: 1,186). A search on "Cheney and quagmire" gives me a more manageable 14 results from ABC and NBC. Clicking on a result sends me to the page with either a transcript or a detailed summary of the video. If I want the clip (to view, or to use), Footage.net provides a form for me to send a request directly to the network referencing the program code.

The site has transferred ownership several times since 2001, and since April it has been owned by Domenick Propati, who had originally come on as a contractor to upgrade the infrastructure a few years back. Propati told me they'll be rolling out enhanced search services in a few weeks. Within each set of search results, the user will be able to further refine the search, such as by date. Also, the service is working with the archival houses to provide video previews online.

This inherent difficulty of storing film and video content, and the inability of any single service provide it, has bedeviled the research/archival community, largely organized around the Association of Moving Image Archivists (AMIA). The Library of Congress considered the problem in 1997, producing the Television/Video Preservation Study. One output of the report was the creation of the Moving Image Collection (MIC) project in 2002, a joint project of the LOC and AMIA. This project has been developed by Rutgers, Georgia Tech, and the University of Washington, and at present the initiative is in the process of moving servers to the Library of Congress.

Jane Johnson, MIC director, described film and video to me as the "poor stepchild" of archival work. "Tens of thousands of hours of television broadcast footage, along with the catalog records of many major television archives, are now available online, but search tools are still evolving, and direct Internet access to large collections is rare," she said.

"Moving Image" comprises both film and video, though it may in the future morph into "Multimedia" in order to encompass audio, which Johnson explained was even more of a step-child in the archival world.

The heart of the project is the MIC Union Catalog which will provide access to any participating archive. The expectation is that individual collections, whether non-profit or commercial, can map their catalog records (whether in MARC or other formats) into the MIC. In addition, the MIC will also identify rights holders so there is less difficulty in clearing the use. As Johnson explained: "Educators are able and willing to compensate rights holders in order to bring appropriate moving images into the classroom and scholarly publications."

MIC today is still underpopulated. The one news collection is from CNN, and that's only a subset. As a stopgap, the MIC does list 241 global archive collections, with a third of them tagged as News & Public Affairs in the U.S. Johnson explained that MIC has been in the process of transferring from the universities to the Library of Congress. In 2008 she expects the project to pursue a substantial outreach program to bring more collections into the fold.

I hope, of course, that the vision shared by Jane Johnson, Marshall Breeding, John Tariot and other pioneers in archiving can be realized. We'd like a single system where any user can search through news and public affairs broadcasts from recent history to get a full education about the people and events which shape our world today. It may not be immediately free for every user -- not even Google or Amazon purport to give book access for free -- but it should allow every user to view or listen to them somehow.

What do you think? What sources have you found online or off for archival news footage? What kind of search service would you like to see? Share your thoughts in the comments below.

Jon Garfunkel is a Boston-based software engineer who publishes media structures research at Civilities.net.

I was horrified to hear about the mass-killings at Virginia Tech on Monday, but didn't want to add my voice to the many who were writing the same thing about it. Luckily, Boston-based software engineer Jon Garfunkel, who publishes media structures research at Civilities.net and helped take the burden off my shoulders by writing up an excellent take on the tragedy and the way the media (new and old) have reacted to it. This is his special report for MediaShift. --Mark Glaser

Part 1: Reactive

After the shootings at Virginia Tech, much of the initial reaction on the Internet followed the familiar rhythm of disaster reporting in recent years. First, the point is made that it is now more likely for an average person to be at a tragic hot spot with a cell phone or Net connection. Next, the story is naturally picked up, either directly or independently, by a traditional media outlet. Finally, the advocates of citizen media claim victory while the actual story is still bleating. Rob Walker first noticed this in Slate in March 2001, when the denizens of MetaFilter first noticed an earthquake. New media triumphalism is now professionalized, and the major journalism studies programs all contribute to it.

David Cohn, a Columbia Journalism School student contributing to NYU Professor Jay Rosen's NewAssignment.Net, wrote that "behind [the tragedy] was actually an emerging network, the birth of yet another citizen journalism network." Dan Gillmor, Director of the Center for Citizen Media, wrote in an Op-Ed piece published by the Washington Examiner, "The scope of the media shift was clearer again on Monday. Some of the most widely viewed images came from a mobile phone camera aimed at the police response by a student, Jamal Albaughouti."

Amy Gahran at the Poynter Institute wrote, "Doubtless in coming days we'll be poring over the first-person blog entries, Twitter posts, forum discussions, Flickr photos, podcasts, moblogs, YouTube videos, and more from those unfortunate enough to be on that campus today. The most poignant content will get highlighted and examined; the harshest and most tasteless will get excoriated."

No longer. Among CNN's Most Watched Videos, Al-Baurghouti's video has become less popular than those produced by CNN's professionals. The #1 video yesterday was the one that was most tasteless -- that of the murderer's tirade/confession which NBC called his "multimedia manifesto." A dozen years ago, it took the New York Times and the Washington Post three months to consider whether to publish the Unabomber's manifesto, and only in an attempt to flush him out. NBC decided to air the shooter's videos, and adding insult to injury, feigned ignorance when its MSNBC prime-time lineup played them in a seemingly endless loop. NBC was utterly unprepared for the torrent of criticism; families of the victims refused to appear on "The Today Show" the following morning.

Once the killer's video came out, the triumphalism tune got remixed slightly. A WGBH spot introducing Chris Lydon's hip-to-the-blogosphere show Radio Open Source last night teased that it would be about "the first mass murder made for YouTube." Never mind that the killer sent his video to NBC via mail, or that the initial cell phone video was sold to CNN, whereupon it was pilfered onto YouTube (which YouTube executive Jordan Hoffner, either predicted or promised). On the program, Lydon mused about the apparent "split" between the killer's choice of the new media (using a computer to produce it), and the old (dropping it in the mail to the NBC).

What's chilling here is that the perpetrators of the Columbine massacre were much more Internet-savvy seven years ago; enough that, according to Sam Vincent Meddis in USA Today back in 1999, that "we've come to expect when the media focus turns to cyberspace, some pundits are trying to blame the Internet for the mass murder in Littleton."

Today, the triumphalists are eager to draw associations to new media where none exist. The Internet is here; it is used by most people; and in some cases for uses unimaginable years ago. This tragedy was not one of them. This whole reactive reporting line is not due to ill intentions; it's simply a need to find a hook to connect to readers. One can find a better hook; one only need to look.

Part 2: Proactive

What's needed here is to see media as something larger than news -- a reactive process -- and also as something more profound than mere conversations. With the speed and pervasiveness of media devices (such as cell phones), it's possible to communicate emergency information to the public right as the situation is playing out. Or before it happens.

This isn't an original idea of mine: It's what happened during Hurricane Katrina. The disaster was only beginning when the winds died and the levees broke. The New Orleans Times-Picayune re-jiggered their NOLA.com website to serve as a virtual rescue center, connecting people and resources across the region.

The Web 2.0 crowd, on dry land throughout America, followed suit by mobilizing for "Recovery 2.0" -- a set of technologies and practices for confronting disaster elsewhere. Jeff Jarvis assembled many of the ideas and people together in his Call to Convene post. One project under way, PeopleFinder, had been used to aggregate "missing persons" services amongst the various news and emergency organizations. Another objective, to Jarvis: "How can we use not just the web and the Internet but also SMS and voice phones and other means to gather news both broadly and very locally?"

In October 2005, 45 volunteers met in San Francisco, including former FCC Chairman Michael Powell. A Recovery 2.0 wiki was set up and hosted by SocialText, but since the start of 2006 the only contributions have been spam. The PeopleFinder website was also abandoned around the same time. The site was last recorded being up on January 16, 2006, according to the Internet Archive. Fortunately, the International Committee of the Red Cross continues to host Family Links, which is used for emergency situations around the world.

For all its good intentions, it seems like Recovery 2.0 was operating in a parallel universe from the spheres of power. The United States Congress took notice of the national mood in recognizing the failures of FEMA. The week after Jarvis's Recovery 2.0 post, Senator Jim DeMint introduced the Warning, Alert, and Responsive Network (WARN) Act in the United States Senate a week after Jarvis's post. This was not completely out of the blue; it had grown out of the work of the Partnership for Public Warning which had formed after 9/11. The PPW's national strategy document from February 2003 took note of the obvious:

Rapidly increasing numbers of Americans are carrying and using electronic devices daily for reasons other than warning. All of these devices could also deliver warnings, and adding this capability can very inexpensive. Use of GPS and other electronic location technologies allow receivers to know their location and to receive location-specific warning information from satellites, television, radio, Internet, and other sources.

The FCC similarly got mobilized. On January 9, 2006, the Hurricane Katrina Independent Panel was formed to analyze how the nation's communications systems served in time of disaster. They issued a 53-page, 37,000-word final report in June 2006, which dealt with the full scope of communications failures.

The WARN Act ultimately was passed in October 2006 as part of H.R. 4954, the Safe Accountability for Every Port Act. While Senator Ted Stevens was maligned by the Net neutrality movement for describing the Internet as a "series of tubes" in June, he had managed to steer the WARN Act through to its passage. The act was signed into law, and the relevant sections 601-606 were untouched by the President's signing statement. The act called for the formation of a Commercial Mobile Service Alert Advisory Committee (CMSAAC).

Only one voice in the blogosphere noticed: Art Botterell, who watched the evolution of the act. He was a bit disappointed about its limited focus on mobile phones as opposed to a multi-modal approach, but was otherwise encouraged by the efforts.

The best press came from a December 2005 article "Reinventing 911" in Wired magazine. Botterell was profiled along with some innovative emergency systems deployments. Gary Wolf wryly noted: "Open data standards aren't sexy. You can't sell them to the government for a pile of cash. And it's hard to pose in front of them for celebratory photographs."

But imagine the benefits a call that goes to 911 about gunshots or a fleeing suspect, and if you're in the area, you're informed about that. You can act, take cover, or inform the person next to you. If your device has GPS enabled, it could filter for only those messages within a few city blocks. In Portland, Oregon, the information is sent to the web so that citizens can actively search for live emergencies as well. Similarly, a private company called Mobile Campus has been providing such a service to universities; in the wake of the Virginia Tech shootings, they have been pushing their service for wider adoption.

What if Virginia Tech had had the use of a broadcast-to-cellphone service when the first shooting happened? Could it have saved lives? "Quite possibly," Botterell wrote me, noting that it's no magic solution; it presumes that the warning would have been prepared correctly.

Of course, the pervasiveness of false alerts must be considered as well. What would have happened in Boston in January when a "suspicious package" was first spotted at 8 a.m.? A couple of things might have turned out differently with that marketing gimmick for Turner that went awry. The commuters coming from north of the city might have just had an idea to take an alternate route. If the Boston Police had the presence of mind to take a picture of it and post it in an information alert system, the guerrilla marketers who had planted it might have been able to volunteer information immediately (The Boston Police has a blog, but their first use of it that day was late in the afternoon to announce the press conference).

Could such a service be abused, as many believe the color-coded Homeland Security "terror alerts" were? Hopefully not. The wireless phone vendors have agreed to make Amber alerts (about missing and exploited children) opt-in, though neither of the Verizon wireless salesmen who sold me phones in the last year thought to mention it to me. The CMSAAC committee looks like it's on the right path; its meeting notes and video are available publicly, and one can always check in with Art Botterell's blog if that's the preferred medium.

We owe it to ourselves -- and to the memories of the victims -- to turn our collective fascination with new media into workable solutions that may help prevent future tragedies.