Press Box

How Newspapers Tried to Invent the Web

But failed.

A moment of sympathy, please, for newspapers, whose readers and advertisers have been fleeing at a frightening rate.

It would be easy to accuse editors and publishers of being clueless about the coming Internet disruption and to insist that the industry’s proper reward for decades of haughty attitude, bad planning, and incompetence is bankruptcy.

But newspapers have really, really tried to wrap their hands around the future and preserve their franchise, an insight I owe to Pablo J. Boczkowski’s 2004 book, Digitizing the News: Innovation in Online Newspapers. The industry has understood from the advent of AM radio in the 1920s that technology would eventually be its undoing and has always behaved accordingly.

For instance, publishers aggressively pursued radio licenses in the early days of broadcasting and, later, sought and acquired TV licenses when they were dispensed. As early as 1947, Walter Annenberg’s Philadelphia Inquirer and John S. Knight’s Miami Herald experimented with fax editions of their papers. Seems visionary enough to me.

Newspapers and other media entities started experimenting with videotex technology in the 1970s, according to David Carlson’s Online Timeline. Newspapers considered themselves vulnerable to new entrants and worried aloud to anybody who would listen about falling readership. In 1979, the Knight Ridder newspaper chain established a videotex subsidiary to develop its Viewtron service, Boczkowski writes. Clunky and toylike by today’s standards (see the silly, pre-Donkey Kong-quality graphics), the early system required an expensive, dedicated terminal. Yet after conducting trials in 1980, the system held sufficient promise that Knight Ridder succeeded in selling Viewtron franchises to other newspapers. More than a dozen other dailies played with videotex during the decade, including newspapers in the Times-Mirror chain, Cowles Media, and McClatchy Newspapers, as well as at the Chicago Sun-Times and the Washington Post.

Howard Finberg of the Poynter Institute remembers that Viewtron could fetch from the “Miami Herald or the New York Times the night before the paper hit your doorstep,” access the Associated Press, look up airline schedules, access bank account info, and order a meal online. Not bad for the dark ages, eh?

Broadcasters joined the text fray, too. In Los Angeles during the early 1980s, CBS was testing the Extravision teletext service, and NBC was experimenting with its own offering, Tempo L.A., according to the New York Times.

So intense was the industry’s devotion to videotex and so rampant its paranoia that some other medium would usurp its place in the media constellation that the American Newspaper Publishers Association lobbied Congress in 1980 to prevent AT&T from launching its own “electronic yellow pages.” Washington Post Co. CEO Katharine Graham, then chair of the ANPA, and other publishers met with Sen. Robert Packwood, R-Ore., to discuss the legislation that would free AT&T to start its service.

As the Wall Street Journal would later report, Packwood said to the publishers, “What you’re really worried about is an electronic Yellow Pages that will destroy your advertising base, isn’t it?”

Graham’s response: “You’re damn right it is.”

Videotex failed to catch on commercially, with Knight Ridder burning through $50 million before closing Viewtron in 1986. The industry’s next favorite newsprint alternative was audiotext, and while Boczkowski writes that the format generated modest profits, it never enjoyed the wild enthusiasm that videotex did. As the decade progressed, the Chicago Tribune, Minneapolis Star Tribune, Hartford Courant, and the New York Times revisited the idea of fax newspapers. Some of the fax editions found a niche but not much more.

According to Boczkowski, newspapers didn’t rush into videotex because they were visionaries in a hurry to invent the future but because they were “reactive, defensive, and pragmatic” about their mature, lucrative business. Having observed the videotex experiments in England and elsewhere, they feared that if they didn’t adopt the technology or at least test it, somebody else would and displace them. Once they determined that nobody could make money from videotex and the technology posed no threat to the newsprint model, they were happy to shutter their ventures.

By the mid-1980s, the industry’s biggest worry was that the PC, which had eased its way into homes and the workplace like an algae bloom, would somehow supplant them. Boczkowski acknowledges that newspapers’ early online strategies were as much about blocking new competitors as beating a path to the future. That said, by the early to mid-1990s, the New York Times, the San Jose Mercury News, the Chicago Tribune, the Washington Post, the Los Angeles Times, and many others were producing electronic-edition business, striking deals with the burgeoning proprietary online systems, such as CompuServe, America Online, Prodigy, and Interchange, or throwing content up on bulletin board systems.

Publishers adored the proprietary online services because they locked down the user experience to the newspaper’s benefit. A Washington Post spokesman quoted in Boczkowski’s book applauds the way Interchange “preserves the company’s direct business relationship with Post readers.”

The publishers were pretty sure that proprietary online services were the next wave, but if you remember having used one, you know how badly they sucked. Let’s say you subscribed to AOL to read the New York Times but wanted to read a story in the Washington Post. You couldn’t get to the Post from AOL because the Post was published exclusively on Interchange. What you had to do was disconnect your screeching modem from AOL, purchase an Interchange subscription, log onto Interchange, and then navigate to the Post. A return visit to the Times required the reverse of that drill.

The extreme suckage of proprietary online services stemmed from the fact that they were “non-generative” technologies, to borrow a phrase from Jonathan Zittrain’s excellent The Future of the Internet—And How To Stop It. Nongenerative technologies can’t be tinkered with or otherwise improved by outsiders. The iPhone is a good example of a nongenerative device: Its software updates “actively seek out and erase” unauthorized modifications, to paraphrase Zittrain.

Generative technologies such as the PC, on the other hand, invite improvement by outsiders, making them more and more useful to users as time passes—and often more useful in ways that the original designers never would have imagined. When you connect a generative technology to a nongenerative one, you usually end up crippling the generative one.

Indeed, the proprietary online services—the AOLs and CompuServes—hobbled the PC, turning the versatile and powerful machine into a dumb terminal. It’s a tribute to newspapers and their keen sense of the future that they quickly determined that the online services would never attract the masses they desired. No sooner had newspapers taken up residence on the proprietary online services than they were packing up their pixels and starting their en masse migration to the World Wide Web, which was as generative as the online services were nongenerative.

Newspapers were anything but late arrivers to the Web party, according to Carlson’s Online Timeline and other sources. Among the earliest pure Web newspapers in the United States were the two dueling dailies started in San Francisco during the autumn 1994 press strike—one by union members and one by management. (As a point of reference, the high-tech sharpies at Wired spun off the site in October 1994.) The San Jose Mercury News broke from AOL and started on the Web in February 1995. USA Today launched a Web edition in August 1995. Later that year, the Boston Globe started its, and the Los Angeles Times announced plans to leave Prodigy. The New York Times and Washington Postgot webby in 1996. After that, few newspapers held back. Boczkowski writes that more than 750 North American dailies were publishing on the Web in April 1998, and by July 1999 only two of the 100 largest dailies were not.

Newspapers deserve bragging rights for having homesteaded the Web long before most government agencies and major corporations knew what a URL was. Given the industry’s early tenancy, deep pockets, and history of paranoid experimentation with new communication forms, one would expect to find plenty in the way of innovations and spinoffs.

But that’s not the case, and I think I know why: From the beginning, newspapers sought to invent the Web in their own image by repurposing the copy, values, and temperament found in their ink-and-paper editions. Despite being early arrivals, despite having spent millions on manpower and hardware, despite all the animations, links, videos, databases, and other software tricks found on their sites, every newspaper Web site is instantly identifiable as a newspaper Web site. By succeeding, they failed to invent the Web.

Addendum 1, Jan. 6: My colleague Adrian Monck  adds this sharp take about the newspaper industry’s techno dance. He writes:

As I’ve  posted before  (and others have  pointed out), there were plenty of executives who did make smart strategic decisions about the challenges facing the industry. Robert Marbut, then CEO of newspaper group Harte-Hanks, was absolutely clear about the threat and opportunities offered by new technology back in the mid-1970s:
The fact that the same technology will be used by media other than daily newspapers will mean that others could enter the marketplace for meeting information needs and encroach on the franchise of an established newspaper … new technology will make it possible for the consumer to get his needs met in a variety of ways in the future, again setting the stage for continued fragmentation of media which could lead to further encroachment of the newspaper’s share of market.
So in the 1990s Harte-Hanks dispensed with its newspaper, TV and radio interests.

Addendum 2, Jan. 6: My colleague Scott Rosenberg, who helped edit the Free Press during the San Francisco newspaper strike and later left the San Francisco Examiner to help start Salon, offers an additional answer to my question about why newspapers’ early Web experiments didn’t end up offering “plenty in the way of innovations and spinoffs”:

In the ‘90s, if you were at a newspaper and learned about the Web, you were likely to grow frustrated and disillusioned with how slowly the paper’s management was waking up to how the new medium actually worked. They got on the Web, and then just sat there. So if you had any restless or entrepreneurial gene in your body, you would sooner or later give up on your arthritic bosses and go do something interesting online yourself or with some startup. The newspaper industry suffered a steady exodus of the very people who it should have been relying on to navigate the new waters.

Addendum 3, Jan. 6: Wayne Citrin writes that I should have made mention of the work done at Knight Ridder’s Information Design Laboratory in the mid-1990s. IDL was working on a portable, battery-powered electronic tablet that its designers hoped would start replacing newspapers by 2001. Knight Ridder shut the lab in the summer of 1995.

Addendum 4, Jan. 6: Chuck Moozakis, editor in chief of Newspapers & Technology, sends me this terrific history of electronic newspaper distribution.


What goes around comes around. Back in the early 1990s, I was offered a job at Digital Ink, the precursor to I turned the job down because I had no idea what job was being offered to me. Had I won a Marshall scholarship for graduate study at Oxford University, I would have declined that, too. What fabulous honors have you declined? Send mail to (E-mail may be quoted by name in “The Fray,” Slate’s readers’ forum; in a future article; or elsewhere unless the writer stipulates otherwise. Permanent disclosure: Slate is owned by the Washington Post Co.)

Track my errors: This hand-built RSS feed will ring every time Slate runs a “Press Box” correction. For e-mail notification of errors in this specific column, type the word decline in the subject head of an e-mail message, and send it to