Democratic Access and the Regulation of Domain Names

By Andy Oram

(I presented a shortened version of the following article at the conference titled “The Public Voice in Electronic Commerce” (Paris, October 11th, 1999). The conference took place at the Organization for Economic Co-operation and Development and was organized by Imaginons un Réseau Internet Solidaire (IRIS) and the Global Internet Liberty Campaign, among others. A few footnotes clarify issues that were raised after the talk.)

Three aspects of domain names will interest members of this audience.

The first topic fits the theme of this conference, universal access. Remember that people use the Internet not just to look or listen; they also have something to say. Perhaps, for instance, there is a small Web site operated by critical, leftist trade unionists. When a politically active person goes online, he may want to find such sites, but how will he know where to go?

As an illustration of the difficulty generated for small sites, I decided to do a simple Web search for something relatively little known, “Genet.” The first page showed links to 15 sites.

(Show first page returned by search engine.)

Let’s see, we start with a site on cystic fibrosis, then Kenyan national parks, cats…nothing I wanted. Let’s look at the next page of 15 sites.

(Show second page returned by search engine.)

Modern music, Jack Kerouac—getting closer, but still nothing I want. Let’s go to the third page.

(Show third page returned by search engine.)

Finally! A site about Jean Genet.

My point is that alternative information is not easy to find; there is too much noise on the Internet. If Jean Genet were here now, he’d know what to do with these interlopers! The first thing he would do is engage a good intellectual property lawyer.

So the subject of thieves, hoodlums, and prostitutes brings me to the second topic in my talk, intellectual property law. You see, major companies such as Hollywood film studios don’t want their customers to have the same problems they have finding Jean Genet. These companies have found a simple tool to get their sites to the public: the domain name system. If I want to find the site for the “Phantom Menace,” for instance, I simply need to type “phantommenace” into my browser. This site is located at the actual domain name phantommenace.com.¹

Let’s consider for a moment the domain name system.

Let’s say you are looking at a Web page at www.gilc.org. Everybody in the world has a site named www by now, but there are no conflicts between them because each www is under a different branch of this tree. The same is true of gilc. Such is the advantage of a hierarchical system; the person under .com doesn’t have to worry about the person under .org.

How does this affect phantommenace.com? When a company wants a domain name for a new product, they want it fast, and they don’t want to discover that some individual somewhere in the world has taken the name already. In fact, the companies don’t even want someone else to have a name that is similar; it might confuse people. Thus, the Volkswagen company has sued an Internet provider that uses vw.net. The company that makes Ajax soap sued the owner of ajax.org. The trademark interests want to change the domain name system from this:

to this:

Luckily, the courts have showed a commendable tendency to balance the rights of trademark owners against the rights of small domain-name holders. The domain-name holder can usually keep the name if he is not trying to confuse visitors, and is not holding the name for ransom, or “cyber-squatting.” This precedent was recently supported by the organization that has been put in charge of administering the domain name system, ICANN. ICANN has not completely exorcised the phantom of intellectual property that menaces the stability and egalitarianism of the domain-name system. For instance, neither the courts nor ICANN clearly give one the right to use a trademarked word in one’s domain name, if one’s site criticizes the trademark holder. But ICANN has chosen the right side on a couple key questions.

I hope that better search engines and eventually an improved naming system will eliminate conflicts over domain names entirely. Both Jean Genet and Lucasfilm will benefit from indexing systems that make the Internet at least as easy to search as a traditional library. Indexing does not mean content rating for the purposes of filtering out Web sites. I want something open and supple that helps people find things, not a single set of scales to facilitate censorship.

But the trademark issue will be remembered for its historical role: it led people everywhere to look at how the Internet was being regulated, or as some prefer to say, how the Internet was governed. It led to the first true governing body on the Internet, ICANN.²

ICANN was set up one year ago by a United States government agency, the Department of Commerce. The stated goals are listed in the accompanying table.

ICANN has encountered lots of criticism for the way it was created, the way the initial board members make their decisions, and the decisions themselves. On the other hand, ICANN is supported by many long-standing leaders on the Internet. They feel that ICANN is critical in an era of stupendous Internet growth and that it will facilitate competition. I cannot address these complex issues in an eight-minute talk, so I will elevate the discussion to higher questions of regulation and governance.

Internet activists can find a disagreement over anything. Some see ICANN as the imposition of a supergovernment not subject to control by the public. Other people see ICANN as a step away from government and part of the commercial, private development of the Internet.

ICANN is truly an anomaly. It is a sort of government, but with no citizens and so far no clear accountability. It is a sort of business, but it would predicate a disaster if it exploited names and numbers as assets. And it is a sort of voluntary association, except that you can’t abstain from it if you want an Internet presence.³

Anything set up by the United States government will be suspected by those outside the United States. For that reason, ICANN has undertaken prodigious measures to engage the international community. But the effort has been directed more toward large institutions than toward ordinary people. We are very far from having a general membership for ICANN. But among the ICANN committees is a governmental advisory committee. They also negotiate with intergovernmental organizations like the International Telecommunications Union (who has a representative here today) and the World Intellectual Property Organization. Such organizations tend to represent the needs of their corporate members. That may be a difficult claim to make while I am a guest of the OECD, but it needs input from groups concerned about aspects of life beyond business development. The World Intellectual Property Organization is not such a group; it presented a plan for resolving domain-name disputes that heavily favors trademark holders. ICANN, commendably, has listened to protests and been reluctant to accept the plan.

Critics of ICANN, especially those from North America, usually delve deeper than its current policies. They ask whether any regulation is needed at all. Now, I know that political activists outside the United States are less contemptuous of regulation than most Americans. Here are some areas of Internet policy where the two sides tend to diverge.

The Internet has been free from regulation during most of its history. I’m distinguishing between regulation of the carrier, such as a telephone line or a television cable or part of the wireless spectrum, and regulation of the content traveling over these media. That content includes all Internet traffic, including the exchange of domain names.⁴

The carrier often requires regulation because it is a limited resource or a historic monopoly. But the Internet itself developed almost outside of regulation.⁵ Groups had to agree on networking protocols, but nobody forced anyone to use them. There are many alternatives to Internet protocols, some promoted by the International Telecommunications Union, but most people just prefer the Internet. And no one forced people to use the Web. The invention of the Mosaic Web browser made it fun. So there was no need for a parent to say, “Now children, be sure to eat your candy.”

What makes the domain-name system the first place where a serious interest in regulation has appeared? The original sin, as you can see from our earlier transparency, is centralization. Note that all the branches come to a single point at the top, called the root. Coordination in some form has to occur. Otherwise, you won’t be sure you can get from one end to the other; a name in one place might not be recognized in another place.

But must coordination occur through regulation? Some fascinating debates are in progress about this. Technically, the software that resolves domain names could allow multiple roots with different data. That technology requires more testing, but if we wanted, we could use it to eliminate the technical need for centralization. Then, by separating intellectual property concerns from the sphere of domain names, we could eliminate the legal or political need for centralization.

The original sin of centralization has led to many other tragedies. We have seen the tragedy of Cain and Abel, as one set of domain-name holders tries to slay another. We have seen the Tower of Babel, as ICANN erects an enormous and costly administrative structure. And we may even see The Flood, because ICANN is in danger of bankruptcy—and not for the first time!

So I will end by posing a very serious question to the audience, one that I hope will generate discussion during this panel.

Are there Internet policy issues that call for decisions either by government or by some other form of collective popular will?

When I started my research, I assumed the OECD would answer with an unequivocal yes. After reading many papers about business self-regulation, I’m not so sure. Some issues for discussion were listed on a previous transparency; other issues include copyright enforcement and accessibility for the blind, the deaf, and those who have trouble using a mouse with their hands. The worst outcome would be to have no public debate, but instead let software designers incorporate assumptions that restrict user choices into products and protocols. This could happen, for example, as a result of government pressure to build in filtering of content. In any case, the history of domain names tells us that independent public interest groups must monitor all regulation, private as well as public.

Back to Cyber Rights home page.