Ladies and Gentlemen,
One year ago, almost to the day, when I first logged on to the IAHC mailing list, I did not know what was meant by an "Internet Year".
Now I do.
I still don't know how many Internet years there are in a calendar year, but it is clear that the concept is asymmetric. Most people think that there are more Internet years in the past twelve months than there are in the next twelve months. They are probably wrong. Life on the Internet would appear to me - from this brief encounter - to be accelerating.
It is not my intention today to give you a blow-by-blow account of the past twelve months. Suffice to say that the international debate about the Domain Name System has brought to light - indeed prominence - an extraordinarily wide range of policy issues, technical and administrative problems and focused attention on organisations and individuals the significance of whose roles and functions has suddenly been recognised, appreciated - and sometimes criticised.
Many people have participated in this process, including many of you in this room. Your efforts have been widely appreciated. I think that today both the private sector and the public authorities do know where to turn in each European country on Internet matters. This is a simple thing, but it is very important, because the Internet really does matter. As a result of the activity during the past year, several milestones have been passed:
Furthermore, the imaginative and creative future use of the World Wide Web is now fully integrated into a significant part of the European information technology research effort. If anyone doubts this, I can only recommend that you stay here tomorrow and go down town to the Esprit conference and exhibition which is currently under way at the Palais des Congrès.
While the Internet Community in Europe has been experiencing a rapid metamorphosis, the international community has adapted in a most positive way to these new developments and I salute them and thank them for it. If Europeans at all levels were uncomfortable with under representation and insufficient participation in the international process a year ago, that is rapidly being corrected and we recognise these changes as a major collective act of good faith on the part of the Internet community as a whole.
A few examples:
But underlying these developments there is an economic and technological reality. The Internet in Europe is growing enormously quickly. Few large economic or social phenomena have ever manifested the growth rates that we currently find in the Internet. The size of the Internet is reportedly doubling every year. The United States accounts for the bulk of the Internet today, but a recent report prepared for the European Commission2 suggests that the Internet in Europe is currently growing three times more quickly than in the United States. That is quite an impressive catch up rate! Internet traffic is clearly growing much more quickly than that because the World Wide Web accounts for ever increasing proportion of overall usage.
Barring unforeseen and unnecessary obstacles, there is no clear reason why these extraordinary rates of growth should slow down, globally. Indeed, much more than the half has not yet been told.
There is no lack of other reports and analyses about this phenomenon, and in this short address one can hardly review them all, but I think we should recognise the efforts of the OECD in drawing attention to the economic links between the growth of the Internet and the structure of telecommunications tariffs3 . The most impressive fact to emerge from the recent ITU report on Telecommunications and the Internet4 was that while the number of Internet hosts in India increased in 1996 by only 300 percent while the number of hosts increased in China by 800 percent.
* * *
The Domain Name System, which we are here to discuss is frankly rather arcane, and its recognised shortcomings have led to several suggestions that we should get rid of it in due course as soon as on-line directories and look-up tables allow users to announce who they are and to find their correspondents. That happy day is, however, some time in the future, and meanwhile we clearly have to fix the DNS. In the process, we have - perhaps inadvertently - triggered two other debates, one psychological and the other political.
The psychological issue is about the confluence of the Internet with our concept of ourselves. The political issues is about how the Internet will be governed in the future.
The main reason why the naming and addressing systems for the Internet are important today is that they may become a pre-figuration of the addressing system for the global communications systems of the future. From Schervan Schreiber to Negroponte, the prophets of the information revolution have not been lacking during the past two decades, but for many of us it has been slow in coming. That page has finally been turned. We are no longer talking about a selectively closed communications system. We are not even talking about a branch of telecommunications. This is about a potentially universal and fully inter-operable communications medium. The issue as to how every individual and organisation may ultimately have access to it, is upon us on an imminently massive scale.
What's in a name? 'A name is a word or phrase by which a person, thing, or class of things is known, called or spoken to, or of'5 . No etymology is offered. The word goes back as far as the lexicographers can trace it.
What for that matter is in an address? Your address points straight in your direction. It is from the Latin dirigere. Furthermore, your dress - what you are wearing - comes form the same Latin word. In this sense your address points not only to where you are, but to who you are, how you can be identified: your "dress".
These are deep and ancient concepts in our civilisation. We should respect them. Above all, we should be very cautious of policies for names and addresses which would appear to restrict individuals access to and use of their names and addresses.
The second issue is political.
The domain name issue has attracted the attention of many people because it is seen to be a microcosm - indeed an archetype of the questions which will have to be answered in Internet Governance more generally.
The governance of the Internet is a broader concept than the DNS. The constellation of organisations and individuals who have managed the growth of the Internet in recent years - and to whom we all owe so much for their manifest success - will need to be consolidated and strengthened. Private self-governance has to be legitimised and supported. In practice - and in law - this can only be done by governments and international organisations with the necessary mandates and powers. The agenda for self-regulation of the Internet is broader than the DNS. It includes the capacity and reliability of the Net itself, pricing policies, content policies, standardisation, the organisation of directories, among others. At present, the physical growth and parallel commercialisation of the Internet is racing ahead of the capacity of the existing organisational structure to deal with all these issues at once. The Policy Oversight Committee has set itself a modest and urgent task - to get the next phase of changes in the DNS right. That is more than enough with the available resources.
But it is quite clear that many observers - both supporters and critics of the POC - have in the back of their minds that this experiment in industry self-governance is a test for how other issues may be addressed in the near future. Whence what may well appear to be a disproportionate degree of political attention is currently undoubtedly addressed to this rather rebarbative branch of communications policy - the Internet DNS!
But the POC is not taking all this on board. The private self-regulating bodies must operate on the basis of a clear mandate from the public authorities. Even in the best of all possible worlds, self-governance has to leave room for political oversight and dispute resolution. This authority will have to be exercised at the international level on the basis of an agreed minimum set of principles specifying:
This last point is important. Some of you will recall that one year ago I was calling for a more representative European participation in the IAHC process. I have to say that that objective has been largely achieved, although not at all in the way that I anticipated at the time.
But the globalisation of the Internet does not stop here. Within very short measure, the self-governing structures of the Internet will have to accommodate the burgeoning interest of many other parts of the world as well. We must all recognise that the Internet must continue to be able to anticipate and adapt to change in this area of participation as well as in other better charted areas of Internet policy.
From this point of view, whilst I support the general objective of enlarging further the functional industrial participation in Internet governance, it is necessary for those industry groups who are currently seeking a seat at the relevant tables to recognise that this objective will also have to be reconciled with the global characteristics of the Internet, and that industry and users from other parts of the world will be seeking a presence and a voice.
The objectives of the POC and the CORE are quite simply to create an open, competitive, transparent and scaleable naming system which respects existing trademark rights. The technicalities are far from simple. A naming system which relies on words and mnemonics is not intrinsically as scaleable as a numbering system. Although it may be technically possible to create large numbers of TLDs, there may be practical difficulties, including how to ensure that all such TLDs respected internationally agreed policies for content, privacy, intellectual property and tax.
Another possibility would be to pile more and more names into the .COM domain. Few people are happy with that option either. The availability of new names in one and the same domain will sooner or later become scarce, if not exhausted and I think that the NSI monopoly in the .COM domain should not be prolonged beyond their existing contract. I believe that the United States Administration agrees, although we still await their announced transition plan.
But adequate, global, long term scalability within .FIRM and .SHOP will require the creation of agreed generic Second Level Domains. There cannot be only one ".SHOE.SHOP", world wide! I even doubt that the shoe-wear industry would be happy with dealing with only one ISP which happened to have the right to .SHOE.SHOP globally. There has to be a better solution which should also contribute to alleviating the problem of trademark infringement and I think it will have to be found in some form of sectoral or professional sub-classification of the new TLDs into well defined generic SLDs. One objective for the POC and the CORE could be to come up with specific proposals along these lines.
As we move into the much heralded era of electronic commerce, there are other issues which will have to be addressed. Existing laws will continue to apply to cyber-space. For example, "no new taxes" does not mean a "tax free zone", much less a tax haven. It will be quite essential that the DNS points unambiguously to a real person or company with a physical address doing business on the Net. There will be problems enough to resolve as to the applicable laws relating to Internet transactions without adding the possibility of transactions through virtual or off shore web-sites, strangers to all relevant jurisdictions.
There are a few other un-resolved questions which POC and CORE will have to address in the next few weeks. The prospect of speculative and defensive registration of names is raising a few eyebrows here and there, and I feel that I should let you know my views on these issues.
A domain name is allocated - licensed, if you like - for a specific purpose (one's web-site and e-mail, for example) for an agreed period of time. One does not own it. One has a right to use it. That is an important difference. Now, I can understand the interest in providing a registration service on behalf of third parties, but these should be third parties who already exist and who have asked for the service. I would not wish to extend this concept to speculative warehousing of names. By extension, I find the concept of a secondary market in names - brokerage of names - quite inappropriate. Absent a trademark, or copyright, there is no property in a name, certainly not in somebody else's name. There is nothing there to transfer, or to buy or to sell.
By defensive registrations, I mean applications for names which are already used by the same company in another TLD. The whole point of creating new TLDs is to expand the address space. For example, WILKINSON.NOM is a different name from WILKINSON.FIRM and so on and so forth. If organisations try and neutralise blocks of the name space in new TLDs on the basis that they already use the corresponding name in existing TLDs, then they risk prejudicing one of the principal objectives of the whole exercise.
While the POC and the CORE are quite understandably concentrating at the moment on a few big short term issues like the registration of new Registrars, the urgent development of shared database software and the resolution of the new TLDs in the Root, it will soon be necessary to look further ahead. The need for on-line directories for the DNS has been repeatedly advanced. Not enough has been done in the past. I would suggest that CORE will soon have the resources, the technology, the database - and the market - to launch the preparation of a complete multilingual on-line directory of all names registered in the new TLDs, and to keep it up to date. When the feasibility has been demonstrated, the principle should be extended to the existing TLDs as well.
Secondly, if the CORE is to offer a really global registration service, I think it will be necessary to develop a TLD policy which is more clearly open to non-English speaking parts of the world. The IAHC has already warned against fracturing the Internet (initially the risk appeared to be coming from parts of the United States market). I believe that in due course the POC will need to address the question as to how to avoid this risk in other parts of the world as well. We need to be prepared for it.
Thirdly I believe that POC and CORE will need a framework within which they can assure harmonious development of the Internet with the co-operation of the existing TLD registrars, particularly the national TLDs. Arguably the current stand-off with NSI will have to be resolved first before this is a practical possibility, but to start with I would recommend that CORE develop close links with RIPE, ARIN and AP-NIC so that the outline of such a framework can be allowed to develop spontaneously through mutual knowledge and understanding among the operators and managers of the DNS, world-wide.
* * *
Ladies and Gentlemen, I shall not draw any conclusions at this stage. That is for the Round Table at the end of the afternoon. On behalf of all of us I would like to thank David Maher and all the other POC members for coming today, particularly those who have been on the road for several weeks now. You may be aware that similar meetings took place in Tokyo and Beijing earlier this month.
I am also glad that so many of you have been able to participate. Many of you represent branches of the Internet-related industries in Europe which in practice come together infrequently and I hope that this occasion gives you all the opportunity not only to meet the POC members but also to need and discuss with each other.
I thank you for your attention.