stating the obvious archives | about

Plug and Play Journalism
Aug 25, 1997 :: Steven Champeon

A chief complaint about the Web and other electronic information systems thus far has been their lack of interoperability. If I want to display the headlines from the News Channel of my choosing, but on my page, it has been up to me to write a filter which will extract the relevant bits from the torrent while discarding masses of presentation-oriented data.

Despite years of (ultimately fruitless) work by Ted Nelson's Xanadu team, fifty years of programming culture, and the ever-present desire for a universal information storehouse, we're far away from a satisfactory, open solution.

The few existing storehouses each focus on one well-defined area of human endeavor. Lexis/Nexis, The Library of Congress, Physicians Online, and others have massive stockpiles of useful information. Are they in the same document format? Do they use the same access/retrieval protocols? Are they indexed using a standard scheme? No.

The encyclopedia, calendar, almanac, and other efforts to provide exhaustive, single-source information about a variety of topics share one very important characteristic -- they presume the unchanging nature of categories of information. This firmament becomes stone in the style of presentation. Show someone a calendar from another culture, or an almanac from another time, and they are generally recognizable. Edward Tufte has written extensively about the presentation of information, using examples as diverse as a Javanese train schedule and a collection of mix-and-match children's outfits. As information is encoded in its presentation mechanism, however, its interoperability is lost to other potential users. This may be the key to understanding why previous efforts have been so focused on intermixing the presentation and content - it is an easy way to maintain intellectual property rights. If nobody can extract the relevant bits from your stream without tremendous effort, they may as well just let you provide the information. For an appropriate fee, of course.

With the recent announcement by the New York Times and others (backed by a Federal Court in New York) that they will not provide further compensation to freelancers for articles re-published in electronic formats, the stage is set for the next step in the history of information cleansing -- what I call "object-oriented journalism." The conceit of all journalists is that their byline matters more to the reception of a story than the masthead under which it runs. However, as the media giants like Time Warner/CNN, Disney/ABC and Microsoft/NBC take more control over the content they contract, the brand of the individual journalist will be overshadowed by the brand of the conglomerate. In the end, journalists will become little more than content programmers.

Which brings me to what I believe will be the inevitable culmination of the journalistic world: Object Oriented Journalism. OOJ is based on one of the more commonly-abused buzzwords in the software industry, "Object Oriented Programming" or "OOP." OOP is elegantly simple: rather than build large-scale Rube Goldberg systems based on branching causality, build large-scale systems based on independent modules with well-defined interfaces. Each object is considered to be a noun, capable of communication and self-reportage. Ultimately, the dream of OOP's proponents is that we will be able to buy widgets off-the-shelf, drop them into our system, and run, like you can with sunglasses or cola, or bicycle wheels. Even more importantly, we will be able to extend existing systems without rewriting them from scratch. We will look back on the way software is made today with the same fond nostalgia that we do with sausage production during the thirties. Or journalism in the nineties.

In Object Oriented Journalism, reporters will no longer need to comb the streets looking for eyewitnesses or reliable sources -- they will be able to buy them from large purveyors of journalistic objects. These objects will contain encapsulated factoids which may then be hyperinked together into brief stories, assemblages which lead the reader into ever-widening spheres of information. Need to cover a story about a murder in Muscongus, Maine? Grab the modules referring to Muscongus, Maine, the parties involved, the relevant histories of the police departments involved in the manhunt, the methods used to prove a murderer guilty or extract him or her from the clutches of foreign governments, the history of forensics, the effects of gunshot wounds on the human anatomy, the violent crime statistics for the area, nation and world, and so on. Voila.

The story exists, as do we all, less as a collection of objects and more as a pattern of relations between those objects. If the story programmer (or artist) responsible for creating the pattern (or story, or article) finds that another media giant can provide more in-depth information for the appropriate area of coverage, s/he can switch factoid providers at a whim. The danger here lies in the ever-present spectre of manipulation, bias, and revisionist history, but what else is new?

The real danger, however, has more to do with the manner in which complex systems hide the origins and nature of their components. Most Visual Basic programmers would be hard-pressed to explain the way in which the systems they used were designed, much less the complex nature of the underlying machine code. Journalists will need to learn how to do the same kind of digging they now do with contacts and sources, but now with regard to the implicit biases and lacunae present in their Backgrounder Toolkit[tm]. They will need to find the balance between creating a quick pattern which reports/represents the "facts" accurately and fairly, and delving into the depths of the objects that pattern subsumes into itself.

In a sense, we are presented with a crossroads, in that we have the opportunity to build systems which bring the self-awareness of journalism to new heights, reducing the overall importance of the byline for the whole story, but are faced with the dangers of revisionism and bias on a massive scale, hidden by the very systems which will make it possible.

Steven Champeon, possessing a skillset euphemistically described by headhunters as "eclectic", occasionally leaves his soapbox for other, more regularly updated, sites. He is NP-complete, invisible in mirrors, and laughs maniacally in his sleep.



Other pieces about online publishing: