Ushering in Banality

Silence, skepticism greets ICANN "process" for new TLDs

After years of furious battles over the domain name system, one would expect an announcement of new top-level domains to be trumpeted around the world like Jesus's second coming. Not so: on 16 November ICANN made just such an announcement, and the response -- even among devotees of name-and-number politics -- amounted to little more than a scattering of vaguely muffled yawns. Coverage that ventured beyond flat-footed explanations tended to be surprisingly negative. BBC Online, for example, opened its lead story on a glum note: "The net's new domain names may do little to open up the internet and the range of names that people can pick." An article in Australia's quoted a Gartner Group analyst as saying, "This is not created to make an average company's life easier -- absolutely not... I think it makes life worse."

From a distance, it's hard to account for just how blasé the reception was. Some details (for example, that the new TLDs are still under negotiation, hence tentative), though salient, are far too subtle: zealous attention to procedural issues has never been a strong point in coverage of ICANN (see, for example, this screed published in the otherwise-cautious New Republic). Other minor details (say, that the TLDs won't be available for several months) seem more credible but inadequate nevertheless. Even the central issue the BBC story led off with, that the new TLDs are restricted to the point of irrelevance, isn't entirely convincing. Clearly, a few hundred airline domains in .aero or a few thousand museums under .museum won't even register as a statistically significant blip in domain markets. But far less clear is how media sources famous for mangling net-related stories suddenly seemed to develop a savvy intuition that the fruits of ICANN's labors were mostly vaporous.

The explanation likely lies elsewhere: in the riche beachside burb of Marina Del Rey, California, on 16 November, where a phalanx of journalists witnessed the "historic" spectacle of ICANN's board awkwardly chiseling away at a mass of proposals until the result looked like an expanded namespace -- whatever that looks like. The board certainly had no clear idea, which (in theory, at least) was exactly as it should be for a "bottom-up," "consensus-based" organization. But -- and this was inexcusable -- nor did they have a coordinated plan or procedure for deciding on the winners; and so their discussions lurched and reeled from sophomoric ramblings to vacuous platitudes to petty preferences and back again with disorienting rapidity. And that, as the dot-commers say, at the end of the day, was the main impression left on many of the meeting's numbed observers -- journalists included. There really wasn't that much to write about; and what little to be written about risked straying beyond triumphal announcements into the arcana (forbidden by most news editors) of ICANN's inner machinations.

Happily, a few journalists pegged the proceeding for the farce they were. MSNBC's Brock Meeks, who was present, derided them as "capricious at best" and likened them to a séance summoning the spirit of Walt Disney. NTKnow, watching from a safe distance via the webcast, scoffed:

".NAME"? I'm sorry, but isn't that the kind of top level domain proposal you'd get if you grabbed a seven-year old off the streets, and told him to come up with a suggestion or you'd give him a black eye? Oddly, that seemed to be exactly the procedure chosen by ICANN for their final consultation on Wednesday in LA.... [T]he truth...was as arbitrary and bizarre as anyone could have hoped....

Sadly, though, the media silence on the subject of ICANN's defective proceedings will likely end up sanitizing the historical record and thereby reinforcing the organization's stature. And given ICANN's demonstrated inability to limit its actions -- whether to the technical coordination mandate granted by the US Department of Commerce, or even just to its own corporate bylaws -- that outcome is troubling.

Take, for example, the fact that ICANN's own bylaws (IV.1.iii.3b) require that any actions "substantially affect[ing] the operation of the Internet or third parties" -- like acknowledging new TLDs -- "require a majority of sitting Board members." Only twelve of ICANN's nineteen boardmembers were directly involved in the decision on new TLDs. Three (Eugenio Triana, Geraldine Capdeboscq, and and George Conrades) simply didn't show up. And four (Robert Blokzijl, Amadeu Abril i Abril, Philip Davidson, and Greg Crew) had recused themselves from the matter -- though Blokzijl and Abril intervened in the preliminary discussions nonetheless. How then did ICANN approve Afilias's proposal for a ".info" TLD with only eight votes? Well, with an ad hoc procedural declaration that the vote did not represent the "consensus" of the board but, instead, was a "straw poll" of those present.

Even those eight yays were an achievement, given that it took three rounds of voting to get them. The reason for that repetition borders on the surreal: the board was unsure what they were voting on.

For pragmatic reasons, ICANN had asked applicants to provide alternate TLD "strings" in their proposals, and they did; and for different (but no less pragmatic) reasons, during discussions the board used a shorthand decription, a single TLD, to refer to each proposal -- the first string, alphabetically. Thus, Afilias's preferred TLD was ".web", with ".info" and ".site" merely listed as alternates; but through most of the discussion the board discussed ".info". However, after the straw polls that decided which proposals would be awarded, boardsquatter Hans Kraaijenbrink moved that the board should specify which TLD to assign, rather than leaving that decision for ICANN's staff to negotiate.

His motion led to a disjointed discussion in which, variously, ".web" became ".info" because some boardmembers thought that's what it was, whereas another proposal (the consolidated Sarnoff/Neuvel application for ".iii"/".per"), which had coasted through the previous day's deliberations, was shot down by ICANN's staff and outside counsel Joe Sims through a process of free association deriving from the string question: ".iii" was "unpronounceable" (according to ICANN CEO Mike Roberts); ".per" might conflict with the unused ISO-assigned three-letter country code for Peru; and the application, according to Sims, was substantially different from what had been submitted (an assessment that had been explicitly rejected during the board's deliberations on the previous day); and (again Sims) it would be unwise (read: potential grounds for litigation) for the board to reject some applications on the basis of a string while accepting other applications without specifying the string. Thus, a proposal that on the first day had garnered substantial support collapsed on the second when the board became trapped in its own tangled lack of process.

Confused? Don't blame the messenger. That account accurately reflects the higgledy-piggledy path traced by the board as it tried to decide what exactly it was deciding on. Was it a specific TLD, a seemingly viable proposal, or a strong applicant? No clear consensus, with inconsistent results.

Was it the presumptive need for a new TLD? The board itself had no quantitative basis for making any such judgment, yet had effectively committed itself through the earlier phases of the application process to a "sunrise" period during which intellectual property claimants would get first crack at domains under new TLDs, thereby disenfranchising the vast majority of would-be applicants who do not have such claims. (The closest the board came to considering this matter squarely was in recitatuions of market-survey data made in applications about which string would be most appealing; in the case of ".info"/".web", they rejected this consideration.)

Or was it the reliability of an applicant's financial backing? In the case of the Diebold Corporation's ".cash"/".global"/".secure"proposal, the refusal to release financial info had been a decisive factor in its rejection; in other cases, famously skittish venture capitalist backers were given the benefit of the doubt. Was it the possibility that an innovative project such as SRI's ".geo" really "needed" a new TLD? They weren't sure; unlike ".pro", evidently, ".geo" could be implemented under a second-level domain such as "".

Or was it fear that approving a telephony-related TLD might provoke massive regulatory agencies such as the International Telecommunications Union (ITU)? Yes. Was it the viability of a registry regulating content within a new TLD such as ".kids"? Probably. Or was it the translinguistic adaptability of a TLD? In the case of ".name" no (".nom" was an alternative), but in the case of ".aero" yes. Or was that just a matter of aesthetics? Definitely -- except for when it shouldn't be. Should the board recommend a TLD different from what had been proposed? Sometimes yes, sometimes no, sometime no -- but here's one anyway.

Or was it doubt about how "democratic" nature of a TLD whose charter required accreditation by some international organization? In the case of the International Confederation of Trade Union's proposed ".union" TLD, the answer -- initiated in a bizarre historical objection raised by the "recused" Abril i Abril -- was yes (".union" was rejected); but in the cases of the ultra-elitist ".museum", sponsored by the International Council of Museums and the J. Paul Getty Trust, and the DotNom-Consortium's bourgeois ".pro" for licensed professionals, questions about democracy never arose (and both were approved). Or was it an overriding concern the geographical distribution of the registries that would support the winning applications? Yes (and many attendees were astonished to learn that, in ICANN's reckoning, the UK and Switzerland aren't in "Europe"). And so on and so forth -- ad nauseam.

Many of these aspects could, within limits, be a legitimate basis for considering a proposed TLD. For example, as the ITU rightly (and defensively) pointed out in its letter to ICANN, a telephony-related TLD may be perceived as directly infringing on national telecommunications infrastructures, which are considered by most nations to be vital to security and sovereignty. But the stunning complexity of these questions, both in their own right and in their interrelation, comes as no surprise -- except, it seems, to ICANN itself. It's in that regard that ICANN's "process" for selecting new TLDs -- notably, its attempt to sort out these questions in a matter of hours -- was first and foremost a sophomoric and amateurish farce. Hence a remark made by Christopher Chiu, a representative of the ACLU, quoted in the Industry Standard:

"They tried to become an arbiter, and all they did was make arbitrary decisions."

It is far from clear that any organization, however adept or well-staffed, could ever succeed in such an ambitious enterprise -- in fact, there's good reason to think it's impossible. And yet, by dint of the fact that ICANN is the Internet Corporation for Assigned Names and Numbers, the central premise of its very existence places it in a structural position of remarkable power to shape how countless human institutions will converge: what nominal realms they will inhabit, what organizations will govern those realms, and the policies by which they will do so. One needn't attribute the threat that poses to malevolence, because incompetence, as we saw, can be just as effective.

ICANN's proceedings are available in RealMedia format in the archives of Harvard Law School's Berkman Center for Internet and Society - here