NS Internet column

Written 14 May 1999 for the New Statesman

There are two things that everyone knows they must say to appear knowledgeable about the MI6 affair, so it's safe to assume that they're both wrong. The first is that this list of 100 names could be read by everyone on the Internet, and the second is that once something has been published there, it will live forever, and be impossible to censor. In fact "published to the Internet" is a phrase almost completely meaningless: if something has not been indexed by one of the major search engines, then it to all intents and purposes invisible, and might as well never have been published at all. If the Vogons were to publish on the Internet their intention to demolish earth in three months' time, they would be completely safe from irritating feedback unless they were also prepared to spend hours getting the site listed by Yahoo and Hotbot as well. The original list appeared in two places where it literally true that non one in their right mind would ever look for anything: the usenet group alt.talk.royalty, and the web site of the American conspiracy theorist Lyndon LaRouche, who believes that the British Royal Family is behind the world drugs trade, and whose front page today offers a pamphlet on "The evil that is Al Gore".

It is almost certain that the British press would have missed it entirely without a D notice to inform them that there was something to look for; and even with that D notice, it took more than 48 hours before the list was found by at least two of the newspapers I talked with. After all, the shorter version of the list, naming only ten officers, had been published last autumn, as part of a claim that MI6 may have been involved in the death of Diana and is still available at various weird Swiss web sites. Civilisation did not then crumble.

The second grossly mistaken piece of conventional wisdom is the idea that when once something has been found on the net it will persist there. Things don't. Usenet postings, for obvious reasons, die away because they are essentially email;. Though there are sites like Deja News which attempt to archive much of the history of Usenet., their efforts are necessarily incomplete, and it is almost always the controversial ones which go first. There is a mechanism to allow the originator of Usenet posts to cancel them, by sending a message which follows them around the system deleting all the copies it comes across. Such cancel messages are easily forged. The Scientologists appear to have forged large numbers in their attempts to stop the publication of certain of their own documents; and it looks as if the British security services did the same thing to the alt.talk.royalty message — but only after they had alerted everyone to its existence and thus ensured there would be plenty of private copies.

To preserve Usenet messages reliably one is reliant on the efforts of private collectors. These may have axes of their own to grind. In fact they are certain to do so: what else would justify the effort and expense of archiving things. But because of the untraceable way in which most electronic documents can be altered, there is no guarantee that an archived copy on a public web site is what it purports to be. Organisations such as Microsoft and the Labour Party are quick to purge from their sites old policy documents that have fallen off message. There is a particular problem with the large MI6 list. It has no provenance at all. Even if it started off genuine — and that seems quite possible — it could perfectly well have been corrupted on its way through the LaRouche organisation. One named member of MI6 appears on the list with his full name, two postings and their date, which makes him sound very official: however, the author has appended to those details the single word "Wanker". ^{Hieronymous Bosch (7:28 Sat 15/5/1999)}

Beyond that, there is the problem of link rot. Anyone who maintains a web site knows that addresses within it change, so that outside links which point to them will stop working; any search of AltaVista or whatever will throw up about 40% dead links, which must once have been valid or else they would not have been listed at all. Again, there is a rule of thumb that the more important a document is, and the more interesting, the more likely it is that its address will change.

Paranoia is the occupational disease of spooks, and of anyone interested in their world. And the basic cause of paranoia is a tendency to overestimate your own importance in the world. So it's no wonder that paranoids are so perfectly at home on the Internet where they can be convinced that everyone is paying attention to everything they have to say. Perhaps the government will have learnt form this fiasco that sometimes the best way to hide information for ever is to publish it on the web but not tell anyone wyou've done so. ^{Hieronymous Bosch (7:28 Sat 15/5/1999)} .