Knowledge at a Cost

"Information wants to be free" is more than a hacker motto. In many discussions of information technology, it's an unwritten assumption. Once we're all hooked up to the Internet, or e-mail, or Lotus Notes, the story goes, everyone will know everything about every job. Hierarchy and specialization will vanish. We'll get all the information we need all the knowledge we need -- via the Net.

Something is wrong with this picture. True, information technology provides powerful tools that allow people in disparate jobs and locations to share what they know. But what we make of information -- the meaningful patterns we create from it -- still depends very much on our nuanced knowledge of particulars.

A lot of information is "sticky," to use MIT management professor Eric von Hippel's term. It is "costly to acquire, transfer, and use in a new locus." The cost lies not in duplicating files but in passing on real knowledge.

In some cases, this is because the knowledge is "tacit" and hard, if not impossible, to articulate. I can copy an edited article in an instant and send it halfway around the world by fax or modem. But it takes years to teach a young journalist the organization, style, story telling skills, and analysis that editing (or writing) the article required.

And information that could be articulated often isn't. Scientists frequently find it difficult to duplicate a lab apparatus because they're missing some critical piece of information. "It's very difficult to make a carbon copy," says a scientist quoted by von Hippel. "You can make a near one, but if it turns out that what's critical is the way he glued his transducers, and he forgets to tell you that the technician always puts a copy of Physical Review on top of them for weight, well, it could make all the difference."

Even if "stickiness" vanished, time would still be limited. Despite the hype, it still makes sense to specialize -- and to select from the gush of "free" information. Indeed, some technological advances make it less important to know about other specialties. A software programmer in 1995 needs to know much less about the specific capacities of chips than a programmer in 1985, because today's programmer is less likely to push the hardware to its limits.

This leads to what University of California, Irvine management professor Steven Postrel (a.k.a. my husband) calls the "black box principle": Treat very capable specialties as a black box and concentrate on learning about only those that have trouble meeting your requirements. Progress occurs when a specialty advances so that other areas can take its abilities for granted.

The "information is free" school of thought also tends to assume that information magically appears in computer networks. But deciding what to record and how means negotiating among people with different needs and knowledge bases. The marketing department of a steel company, for instance, wants products classified by how they're used, while manufacturing wants them classified by how they're made.

A health insurer studied by Elihu Gerson and Susan Leigh Star of the Tremont Research Institute had a five-person unit devoted to resolving conflicts over how to code medical procedures. Doctors, customers, and various internal departments disagreed about what criteria should determine the codes and how narrow the categories should be. Yet everyone had to use the same system. The unit's job was essentially diplomacy: getting the different groups to agree on coding guidelines.

"Traditionally, finding solutions to information systems problems has been framed as…arriving at the 'correct answer' via algorithmic procedures," write Gerson and Star. But in this case, there is no single correct answer. "Rather, there are multiple, possibly inconsistent, competing answers, none of which has a unique claim to validity."

Nobel laureate F.A. Hayek, one of the earliest economic theorists of the role of knowledge, would have recognized the problem. Organizations overcome such difficulties through compromise and negotiation, or they break down. But society as a whole can avoid the "knowledge problem" only at the cost of stamping out individuality.

In his classic 1945 essay, "The Use of Knowledge in Society," Hayek criticized economists who believed central planning would be simple. He wrote that "the knowledge…we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess." Those "bits" include not merely facts, but complex individual goals and preferences information that, by its very nature, can't be centralized.

In our enthusiasm for the power of information technology, we would be foolish to forget its limitations. Information may be cheap, but knowledge still isn't free.

A slightly shorter version of this article originally appeared in the June 5, 1995 issue of Forbes ASAP.