Since Dotgay’s application for a Community Priority Evaluation on .gay failed last week, there’s been some unrest among its supporters and in the media.
An effort to get #ICANNisBroken trending hasn’t exactly set the Twittersphere alight, but there have been a handful of news stories that attack the CPE decision for failing to represent the gay community.
I think the criticisms are misplaced.
The Economist Intelligence Unit, which conducts the CPEs for ICANN, got it right in this case, in my view.
Dotgay scored 10 points out of 16 on the CPE. It needed 14 to pass. A pass would have given it exclusive rights to .gay, forcing the three other applicants to withdraw.
It could have scored 14 had it managed to get 4 points out of the available 4 on the “Nexus” criteria — the strength of the relationship between the string “gay” and the community Dotgay said it wanted to represent.
The EIU scored Dotgay zero.
The reason was that Dotgay, in its application, defined its community like this:
The Gay Community includes individuals who identify themselves as male or female homosexuals, bisexual, transgender, queer, intersex, ally and many other terminology – in a variety of languages – that has been used at various points to refer most simply to those individuals who do not participate in mainstream cultural practices pertaining to gender identity, expression and adult consensual sexual relationships. The Gay Community has also been referred to using the acronym LGBT, and sometimes the more inclusive LGBTQIA. The most common and globally understood term – used both by members of the Gay Community and in the world at large – is however “Gay”.
Dotgay lost the 4 points it needed to pass the CPE almost entirely because of this paragraph.
It overstretched, and it doing so it failed to play by the ICANN rules.
Remember, the EIU was asked to determine the “nexus”, or correlation, between the gTLD string — the word “gay” — and the people Dotgay said it was trying to represent.
By trying to be as inclusive as possible to as many different sexuality/gender identities as it could, Dotgay lost sight of the fact that the gTLD string it wants only describes a subsection of those people.
LGBTQIA isn’t even a particularly well-understood acronym. The “A” could mean “Ally”, as Dotgay said in its application, or “Asexual”, as it is often (but not in Dotgay’s application) interpreted.
Dotgay tried to define “intersex” people — humans whose genitalia or other sexual characteristics do not conform to the standard male/female norms — as “gay”. Are they “gay”?
It tried to define “allies” of gay rights as “gay”. Are we?
The majority of the straight people reading this post, myself included, would characterize themselves as an “ally” of the gay community — we’re supporters of equal rights — but we would not call ourselves “gay”.
The fact that we wouldn’t is an important part of the EIU’s logic, the reason it found Dotgay had overstretched in its community definition, and as hard as I try I can’t figure out why that logic is faulty.
“Membership in the Gay Community is not restricted by any geographical boundaries and is united by a common interest in human rights,” the Dotgay application reads.
Is that not an implicit admission that the “gay community” defined in the application actually includes the majority of the populations of most right-thinking democracies?
Regardless of the EIU’s logic, there has been a moderate amount of outrage online about its decision.
The article getting the most link love appears to be this one at Slate, written by “LGBTQ activist” Marc Naimark.
First, note the acronym Naimark uses in his bio at the bottom of the piece. There are two letters missing when compared to the Dotgay application — “I” and “A”, for “Intersex” and “Ally” or “Asexual”.
Would Dotgay have won its CPE if it had limited itself to the same five letters? Maybe, maybe not. GLAAD defines “gay” as only those people attracted to the same gender. Transsexuals may not count. I and A almost certainly don’t.
Is this just nit-picking?
Not really. The point of the CPE, taken as a whole, was to allow genuine communities to avoid expensive auctions whilst preventing gaming by unscrupulous registries that would seek to claim a valuable string without a genuine community behind them.
In a previous new gTLD round, ICM Registry defined its .xxx “community” as essentially ‘anyone who wants to be a member of the community’. The .mobi registry defined its community as basically ‘anyone with a mobile phone’.
These were both attempts, in my view, to game the rules ICANN had put in place for that particularly new gTLD application round. They were both successful.
What Dotgay tried to do with its .gay application was to define its community as basically everyone. I don’t think I would call it gaming, but I might call it a failure to follow the community rules closely enough.
So is it a bad thing that Dotgay’s CPE got rejected?
I don’t think so.
Remember, Dotgay has not been ruled out of the process. It can still compete at auction with the other applicants.
If that’s too rich for it, there may even be an opportunity for the company to combine in another way with a rival applicant, rather like DotGreen did with Afilias for .green.
In Salon, Naimark wrote that a non-Community .gay — one manged by Top Level Design, Minds + Machines or Rightside, the other three applicants — will likely be awash with pornography or homophobia:
Now, instead, .gay is up for auction, with dotgay LLC facing off against three much larger rivals whose sole aim is to make as much money as possible from .gay names. That means no oversight over who gets a name or what it’s used for. Gay bashers will be able to buy .gay domains. More significantly, the largest market is likely to be among porn sites. Any legitimate use of the name by individuals, businesses, and organizations associated with the LGBTQ community will likely be drowned in a sea of sex: On the Internet, everyone will be .gay for pay.
On the face of it, that seems like a compelling argument. Wouldn’t it be nicer if .gay was devoted to worthy causes rather than gay porn? I would probably agree with that argument.
But none of the four applicants for .gay — not even Dotgay — have any prohibitions or restrictions on porn in their applications.
There’s no reason on the new gTLD program record to believe Dotgay won’t sell .gay domains to porn sites too.
Where Dotgay does have a moral advantage against its competitors is in its explicit prohibitions against homophobic speech in the domain names it sells.
One of the policies it proposes in its application is that domain names should not be “words or phrases that incite or promote discrimination or violent behavior, including anti-gay hate speech.”
The other three applicants don’t have anything nearly as specific in their applications.
But by applying as a formal, big-C “Community” applicant, Dotgay also had to promise to restrict its gTLD to a limited number of people who were members of its self-defined “community”.
This is where I struggle.
Dotgay proposes to restrict .gay to people who obtain a special code via a number of as-yet unspecified, approved “Authentication Providers” — organizations that represent sections of the LGBTQIA community.
This process has clearly been created by the applicant, in my opinion, in order to get the required number of points in the CPE’s “Registration Policies” criteria, where you have to be restrictive to win.
What Dotgay is proposing is a system whereby in order to express your homosexuality (or membership of another LGBTQIA community, including “ally”) you need to apply to an approved gay-related organization for a special code.
That just seems wrong to me.
Whatever happened to the “self-identified” gay person? I thought that “self-identification”, in the sexual and gender identity world, was a bit of a big deal.
You need a password to “come out” online? Really?
I don’t want to be accused of “straightsplaining”, so it’s a genuine question: is it okay for a registry to need to authenticate your sexual/gender identity before you can register a .gay domain?
ICANN director Olga Madruga-Forti unexpectedly quit the board last week.
ICANN did not give an explanation for her sudden departure, which came toward the end of the ICANN 51 public meeting in Los Angeles.
The Argentinian telco lawyer’s resignation means she will miss the third and final year of her appointed three-year term.
Her decision comes almost exactly a year after Filipino entrepreneur Judith Vazquez also quit, again with no reason given, two years into her own three-year term.
This possible coincidence has led to speculation that the ICANN board has an “aggressively male culture”, whatever that means.
Both Madruga-Forti and Vazquez were selected by the Nominating Committee, which has guidelines obliging it to try to maintain a healthy gender balance on the ICANN board.
I’m not sure whether Madruga-Forti’s resignation supports or challenges my previously stated view that pro-female gender discrimination by NomCom is of questionable value.
On the one hand, NomCom has for two years in a row selected candidates — partly on the basis of their gender and geographic origins — that didn’t make it through a full term.
On the other hand, if the male-heavy gender balance on the board is to blame for these resignations, perhaps a bit of enforced balancing may help maintain a stable board in future.
It’s a tricky one.
Currently, only four of the (currently) 20-member board are female. Three have voting rights. Of those three, two were selected by NomCom. The third was elected by the At-Large.
Two of them have been on the board for less than a week, having been selected or elected for terms beginning last Thursday.
It seems likely that Madruga-Forti’s permanent replacement will turn out to be female. Just a hunch.
What do you think? Is ICANN too blokey? How important should gender balance be on the ICANN board?
The European Union request for the Greek-script ccTLD .ευ has not been thrown out, according to ICANN.
Last week DI reported that .ευ was the only one of three IDN ccTLD requests — the other two being Bulgaria’s .бг and Greece’s .ελ — to fail a test for confusing similarity on appeal.
.ευ was found to be confusingly similar to .EY and .EU, but only when in upper case.
The similarity panel’s decision would mean, I reported, that .бг and .ελ would be delegated but .ευ would not, under ICANN rules.
I wondered aloud what the Governmental Advisory Committee would think about that, given that it had lobbied for the creation of the appeals process in order to get an earlier rejection of .ευ overturned.
Shortly after publishing the article, ICANN reached out to say I was wrong and ask for a correction.
“We (ICANN) have not rejected the .ευ application,” a spokesperson said.
“Due to the unprecedented nature of the split results, the issue needs to be discussed at the senior management and Board level before a final decision is made,” he said.
The “split results” refers to the fact that there was found to be no confusing similarity with .ευ in lower case.
However, the ICANN rule I referred to says (which my emphasis):
The rule is that if the appearance of the selected string, in upper or lower case, in common fonts in small sizes at typical screen resolutions, is sufficiently close to one or more other strings, it is probable that a reasonable Internet user who is unfamiliar with the script perceives the strings to be the same or confuses one for the other.
That’s adapted almost verbatim from the original recommendations of the ccNSO. The only addition ICANN made was to add the clearly important clause “in upper or lower case” to the text.
It seemed pretty straightforward to me — confusing similarity exists regardless of case.
I pointed this out to ICANN last Wednesday and asked where I could find the rule that said the ICANN board or staff get to review a “split results” finding but have yet to receive a reply.
ICANN has reopened the contention sets for .cam and .通販 after deciding that two String Confusion Objection panels may have been wrong to reject certain applications.
Two rulings — that .cam is confusingly similar to .com and that .通販 is confusingly similar to .shop (really) — will now head to an appeals panel for a “final” determination.
The decision was made by the ICANN board’s New gTLD Program Committee this week at the ICANN 51 public meeting in Los Angeles.
The first case being reopened for scrutiny is Verisign versus Rightside, where the original SCO panel found that .cam and .com were too similar to coexist on the internet.
But a different panelist found that the two strings were not confusingly similar in objections filed by Verisign against two other applicants — Dot Agency and AC Webconnecting.
The opposing rulings meant that Rightside’s application would have been kicked out of the .cam contention set, which hardly seems fair.
This and many other “perceived inconsistencies” led to the ICANN board being pressured to come up with some kind of appeals process, which it agreed to do in February.
Verisign, unfairly in my view, was not given the opportunity to appeal the two .cam decisions that went against it, even though they were made by the same panelist for the same reasons.
The second, altogether more peculiar, case was .shop applicant Commercial Connect versus .通販 applicant Amazon.
The panelist in that case seemed to have checked his brain at the door that day, concluding that the two strings are confusingly similar simply because 通販 means “online shopping” in Japanese.
Another panelist, in a different case also involving Commercial Connect, had found that .购物 (Chinese for “shopping”) was not confusingly similar to .shop because duh.
ICANN’s NGPC has now decided that the two controversial decisions are “not being in the best interest of the New gTLD Program and the Internet community”.
Both .cam and .通販 will now be referred to a three-person panel at the International Center for Dispute Resolution, the same body that processed the original objections, for a final determination.
Campaigns in Bulgaria and Greece to get ICANN to un-reject their Cyrillic and Greek-script ccTLD requests have proven successful.
The first decisions handed down by ICANN’s new Extended Process Similarity Review Panel this week said Bulgaria’s .бг and Greece’s .ελ are not “confusingly similar” to other ccTLDs after all.
However, a third appeal by the European Union over the Greek .ευ was rejected on the grounds that the string is too confusingly similar to .EV and .EY when in upper case.
Confusing strings should not be delegated, under ICANN rules, due to the risk of exacerbating the prevalence of security risks such as phishing attacks.
Bulgaria’s initial request for .бг was turned down in 2010 after a panel found it looks too similar to Brazil’s existing ccTLD, .br.
Greece’s bid for .ελ had been blocked for looking too much like .EA, a non-existent ccTLD that could be delegated to a new country in future.
While the initial panel’s process was pretty opaque, the newly published “extended” reviews appear to have employed a fairly scientific methodology to determine similarity.
Twenty American undergraduate student volunteers were shown pairs of strings briefly on screens designed to simulate web browsing. They then had to pick out which one they’d seen.
The volunteers were also shown pairs of similar-looking Latin-script ccTLDs that already exist, in order to establish a baseline for what should be considered an acceptable level of confusability.
The Greek and Bulgarian strings were both found to be less confusing than existing pairs of Latin-script ccTLDs and were therefore given the thumbs-up. The EU string flunked in upper case.
Under ICANN’s rules, it appears that .бг and .ελ can now proceed to delegation, while .ευ has been forever rejected.
The three reports can be downloaded here.
It will be interesting to see how the ICANN Governmental Advisory Committee will react to this.
It was pressure from the GAC — driven by the European Commission and Greece — back in 2012 that forced ICANN into creating the appeals process.
At ICANN’s meeting in Prague that year, the GAC said:
The GAC is of the view that decisions may have erred on the too-conservative side, in effect applying a more stringent test of confusability between Latin and non-Latin scripts than when undertaking a side by side comparison of Latin strings.
Now the EU seems to have been told that it still can’t have its requested ccTLD, and the standard applied was exactly the same standard as applies to Latin ccTLDs.
Will the GAC accept this determination, or stomp its feet?