As Internet use has exploded over the past five years, it has begun to work fundamental changes on the U.S. economy. Information is rocketing around the globe faster and in greater volume than ever before. Productivity continues to rise. Technology is evolving at an astonishing clip.

Although investor enthusiasm has been dampened in recent months by the market troubles afflicting Internet enterprises, nobody doubts that a new economic force has been unleashed on the world and that we have seen only a fraction of its potential impact. As a result, the Internet has sparked a lot of radical rhetoric by information-age prophets. The Internet is different, they say. The old economic rules do not apply.

Is the Internet economy really revolutionary? Are the economic rules that govern the information-age businesses somehow different than the rules other industries play by?

Economists say the Internet and the electronic commerce it generates -- the interconnected web of computers, software, telecommunications links and content -- does behave by a set of rules that sometimes are radically different from those applying to traditional brick-and-mortar industries.

Consumers already have witnessed remarkable changes in the way business is done and companies behave. Online stores such as Amazon.com have changed shopping habits, offering books and other products with a mouse click. Travel agents now compete with online travel booking sites. The Web auctioneer eBay has created a subculture for auctions of just about anything.

The Internet might ultimately topple established ways of doing business. Because digitized content is easily copied and transmitted over the Internet, businesses that deal in intellectual property, such as the publishing and recording industries, face the threat of plummeting costs. The music-sharing software Napster, for example, is a virtual network made up of personal computers capable of circulating music for free. Napster represents a threat to the structure of the music business, which depends on getting $15 for a CD and legal sanctions for people caught copying music.

The Internet's ability to make distance irrelevant threatens the telephone industry. Internet telephoning, the ability to make a long- distance call via an Internet connection, still is a small industry, slowed by technical limitations. But industry experts believe it could, within a few years, change the shape of today's telephone market, where charges now are based on distance between calling points.

The Internet began as a government-sponsored defense computer network and has since grown into a freewheeling and unsupervised arena. As it has become a growing economic force in the United States and globally, with large corporations battling for control of key resources and market share, the government reluctantly has stepped back into the picture.

Rep. Billy Tauzin, R-Chackbay, chairman of the House telecommunications subcommittee, is a leading critic of government regulation of the Internet and has clashed with those who want the government to write some rules of the road.

What makes the Internet economy special, and its behavior sometimes counterintuitive, is that it is a network.

Networks link large numbers of people in common relationships, and they behave in sometimes unpredictable ways. For businesses, markets dominated by networks tend to rapid growth and bigness. They favor a few dominant players with the largest webs of relationships.

The Internet is a network of networks that links billions of people around the world. It includes physical networks, such as the "Internet backbone" of fiber-optic cables and routing computers that transmit information. It also includes "virtual networks," communities of people organized around a common interest or using the same software.

In their early stages, networks often grow quickly and virtual networks grow even faster. Internet economy is growing and changing so quickly that rules written today might be rendered irrelevant by the next technological innovation or large-scale economic shift. Meanwhile, some of the old rules quickly are becoming outdated.

"Let's just face it: Standard regulatory institutions are not set up for this situation," said Shane Greenstein, an economist at Northwestern University. "They are set up for a much more mature technology that changes at a much slower pace. It challenges them when things are changing rapidly under their feet and they have to figure out how to intervene -- or whether to intervene."

Even if officials were able to keep pace, they still might have trouble predicting the behavior of new industries. Networks exhibit several strange behaviors that explain why big, fast-growing companies tend to dominate in some arenas and why some companies and consumer groups are calling for government intervention.

In most businesses, the number of customers or products sold does not alone determine success or failure. Bigger might not be better. It might cost too much to make a product or service for large numbers of customers; if a company doubles its output of widgets, it might not be able to sell them at a high enough cost to make money.

Network economics, however, favors bigness. The most valuable networks have the largest number of customers. "In a network, the higher number of units you sell, the higher the sales of a good, the more valuable a good becomes," said Nicholas Economides , an economics professor at New York University.

There is an obvious reason for this phenomenon, called "network effects." A telephone network with one user is worthless. One with two users has some value. One with a million users is vastly more valuable, because any user can call those 999,999 other customers.

Network effects favor companies that command large customer bases. A familiar example is Microsoft, whose customers form a virtual network because they share common operating systems and software. In general, the more computers that use a particular operating system and software, the more people can share files and communicate. As Microsoft became the dominant operating system globally, with millions of customers, competitors such as Apple fell by the wayside.

The downside of this, of course, is that network markets might unfairly limit consumer choice, something Judge Thomas Penfield Jackson said had happened in his decision sanctioning Microsoft.

"The network industries present pretty complicated issues," Economides said. "You have a natural equilibrium in which one winner takes over most of the market. That presents special challenges for people doing antitrust or public policy."

Economists have analyzed some of the processes networks undergo as they get bigger and allow companies to gain market share and the upper hand. They are hotly debated concepts among economists because they call into question the notion that market forces invariably create the best range of choices for consumers. In the case of networks, markets may lead to monopolies and limited choices.

-- Positive feedback. Classical economics predicts that markets tend toward an equilibrium in which several companies compete with each other, none dominating. If one develops a small advantage, the others can compensate for it. Competition makes for better products and more choices.

But in the world of network economics, small advantages can quickly turn into big advantages, and an ideal balance between competitors might never come about. If companies get more valuable as their customer base grows, it can become a self-reinforcing cycle. A company grows, becomes more valuable, which sparks another round of growth. The cycle is called positive feedback because it is analogous to what happens when a small noise cycles through an unbalanced sound system, emerging as an ear-splitting electronic shriek.

-- Path dependence. Network economics emphasizes the importance of the historical "path" a company takes. If a company gains a small advantage early in the history of a fledgling market, it might end up turning that into market domination down the line. That means what happens early in the lifetime of a new market is very important. Even small, chance occurrences might tip the balance toward one company or another.

One oft-cited example of path dependence is the videocassette market. In the 1970s, two types of videocassettes were competing: Beta and VHS. Beta had an advantage of compactness, but VHS got a leg up in the marketplace and eventually became the industry standard. Some economists dispute this example, however, noting that Beta started out with more customers and that VHS eventually overcame the advantage with features, such as longer tape times, that appealed to consumers.

-- Lock-in. When a company comes to dominate with a particular networked product or service, it might be almost impossible to dislodge, even by companies offering superior products. It is locked in. The logic of lock-in is straightforward: If all the customers on a network use one operating system or type of Internet connection, for example, the advantages of sticking with it may outweigh the costs of switching to another one.

The classic example of this is the standard QWERTY keyboard, which has been around since the typewriters of the 19th century. It has no inherent advantage over other keyboard arrangements, but virtually all keyboards have it, people are trained to use it, and the costs of switching to a more efficient alternative would be high. Here, too, some economists dispute the notion that there are significantly better alternatives and say that if there were, market forces would favor them.

As officials examine the rollout of high-speed Internet services by competitors in the cable, phone, wireless and satellite industries, everyone is worried that the dynamics of networks could lead to monopolies or dominance by a few large players, perhaps re- creating a situation similar to a century ago, when AT&T consolidated its telephone monopoly.

Tauzin, for example, raises the specter of a monopoly by cable giants as a rationale for his bill granting regulatory relief to the big Bell telephone companies.

But there is no consensus among economists about the ultimate shape of the broadband marketplace. It depends partly on the federal government. The path-dependence idea means that modest actions taken today might have enormous ramifications later on, for good or ill, increasing both the potential risks and rewards for any action.

It is possible that if a single industry develops a superior and cheap standard way to deliver high-speed connections, it will quickly take over the marketplace. But it also is possible that the diversity of technologies, the range of customers and the chaotic democracy of the Internet might mean that none among the four may come to dominate.

"The big difference with AT&T is it had a concentration of patents and engineers on the frontier," Greenstein, the economist, said. "They had technological assets nobody could replicate. That's partly what led to their dominant position in the industry. If you look at the Internet today, basic technical capabilities are widely dispersed. It's not hard to do a lot of this stuff on hardware and the software side, and it's not that expensive to assemble a bunch of people who can do it."