The future of governance was recently on display in Yokohama, Japan. It was not a World's Fair, a U.N. conference or an international exposition. Rather, it was the latest meeting of ICANN, the Internet Corporation for Assigned Names and Numbers.

ICANN is not governmental in the usual sense. It does not oversee a geographic jurisdiction. It does not have an army or even a police force. What it does possess, however, is authority. At present, its authority appears somewhat limited: ICANN has been delegated responsibility for the management of the Internet's address book. That is, ICANN is making the rules that determine who gets the rights to Web site names and manages the technical facilities that makes, say, Amazon.com's (AMZN) Web site appear when you type "http://www.amazon.com" into your browser window.

But what exactly is ICANN and why should it be making rules regarding anything? Some people mistakenly think that ICANN is a U.S. government agency. This confusion is borne of a contemporary propensity to make government agencies seem more efficient by calling them "corporations" (for example, the Corporation for National Service). But ICANN is not part of the U.S. government -- or any other government for that matter.

No, ICANN is a private, nonprofit corporation. It was created by the late Jon Postel, former leader of the Internet Society and one of the architects of the Internet, after the Clinton administration announced that it would transfer responsibility for management of domain-name registration to a private organization. Up to that point the domain-name registry had been administered by Network Solutions (NSOL) , a private company, under contract with the National Science Foundation (a U.S. government agency, despite its misleading name) and later, the U.S. Commerce Department.

ICANN's critics charge that ICANN is an instrument of Internet interest groups that secretly colluded with the Clinton administration. The appearance of an objective selection process, they say, was a sham. Setting aside this controversy, delegating regulation of the Internet to ICANN (or any other nongovernmental entity) suggests an approach to the complexities of governance in this era of globalization that is likely to become common and thus deserves examination.

Trade, communications, crime -- almost every form of human activity -- now routinely crosses borders. As a result, the mechanisms we have relied upon to regulate everything from accounting standards to telephony now appear creaky and outdated. How can the U.S. Food and Drug Administration meaningfully protect Americans from dubious drugs if manufacturers around the world have direct access to U.S. consumers? How can local law enforcement officials punish operators of rogue gambling operations based offshore? How can American financial regulators verify the truthfulness of the claims made by issuers of stocks and bonds on the other side of the world?

The science fiction solution is world government. But this approach seems far-fetched because national governments will not vote themselves out of existence. The answer is more likely the growth of entities like the World Trade Organization and the World Intellectual Property Organization to which governments of the world will cede authority and responsibility for international transactions and activities. In theory, such organizations can overcome many of the logistical problems that restrict national governments. Their authority ranges across borders. They propagate and enforce regulations that apply in all jurisdictions.

The creation of ICANN extends the governance experiment one step further. Unlike treaty-based organizations such as the WTO, ICANN's creators hope to sever all formal ties to the governments of the world. ICANN is envisioned as a model for the quasi-government of the Internet because it is ostensibly responsible only to "the Internet community." This is immensely appealing -- on paper. Like many who vilify traditional government agencies, ICANN's proponents argue that by being independent of "the bureaucracy," ICANN will be lean, efficient and free of the political wrangling that characterizes traditional government. Moreover, ICANN is promised to be immune to the ideological disputes that make governance in the nonvirtual world so difficult.

But the early experiences of ICANN indicate that such governance structures introduce vexing challenges of their own.

For instance, representativeness. ICANN aspires to be a democratic government for the Internet. Without borders, however, identifying those with a legitimate right to participate in ICANN's decision-making is difficult and contentious. There is a vocal group of critics who argue that ICANN's board of directors and management are not representative of the Internet user population and, as result, biased toward corporations with an overriding interest in protection of their commercial property rights.

The last ICANN meeting, in Cairo, was embroiled in disagreement regarding the election of additional board members. The board bowed to objections that an indirect election system would give the existing board greater authority; the election procedures are being revised but will allow anyone with an e-mail address to vote for five at-large members of the board (from a list put forward by a nominating committee). Will this eliminate objections to the composition of ICANN? Not likely.

Another difficult issue involves challenges to authority. Governments have a unique tool to compel subjects to respect their authority: force. ICANN has no analogous monopoly on power, physical or otherwise. Indeed, there are ways to circumvent or ignore ICANN. For example, even as ICANN has established and implemented procedures for arbitration of domain-name disputes, opportunities to litigate disagreements abound at the state and federal level. It's difficult to maintain governmental authority if parties dissatisfied with the outcome can turn to another venue.

Accountability is also tricky. To whom, exactly, is ICANN accountable? The question is difficult to answer. At present, the Commerce Department's National Telecommunications and Information Administration exercises at least a supervisory role. In the future, however, it is anticipated that ICANN will grow wings and set out on its own. ICANN should not look to Uncle Sam for guidance as it oversees an increasingly international Internet. America's dominance of the early days of Internet governance was a natural consequence of its seminal role in the Internet's creation. But what, if any, body will assume the role now played by the U.S. government? Opponents can now turn to Congress to air objections regarding the administration of ICANN. Indeed, Congress ordered the recently released General Accounting Office report on ICANN. Who will conduct investigations in the future?

These concerns are expressed now in theoretical terms. But "what if" can quickly become "what now?" ICANN raised a few eyebrows by granting the Palestinian authority its own top-level domain name, the same status accorded nation-states. The symbolic recognition provides a small reminder that organizations to which significant authority is delegated have a funny habit of using it in ways that few imagine. ICANN may someday make decisions that affect the ability of the U.S. government to protect property rights or police Internet transactions. The consequences could be more than symbolic.

It is difficult to picture ICANN, an obscure entity with an odd acronym, as anything more than a footnote to the rise of the Internet. But a year ago it would also have been hard to imagine violent protests disrupting the WTO meetings in Seattle. The world is changing. Old boundaries are being eroded by commerce, transportation and communications. And government is adapting, taking new forms to accommodate the new reality. In this sense, ICANN is an important harbinger of the controversies to come.