Tuesday, December 14, 2004

Identity Theft

Samizdata had a lively discussion on the merits of a national ID card, which sprang from the persistent misidentification of a person with a wanted criminal of the same name. This happened despite the dissimilarity between the two in every respect except their names. Some Samizdata readers believed that such mistakes could be avoided by requiring a national identification system. But some national ID critics argued that it was the concept of equivalence between a person and his descriptors that itself constituted the danger. From Name=Person, which lay at the root of the misidentified man's troubles, is substituted ID Card=Person, a relation only as good as the institutions that issue them.

Now imagine that a criminal manages to steal/copy your ID card and starts impersonating you. The ID card and government database will of course be treated as infallible, so no doubt any arrests/crimes associated with a person carrying your ID card will be going down on your file. It must have been you who used that credit card fraudulently, they had your ID card. ID cards will be the identity thief's dream come true, and if this sort of thing can happen just because two people have similar names, just think what could happen when someone sets out to impersonate you.

The concept of a credential, which may be defined as "a document attesting to the truth of certain stated facts" occurs frequently in designing IT systems, but it also appears in politics and journalism. Consider the concept of a 'mainstream Iraqi blogger'. The brothers who author Iraq the Model have recently been visiting the United States and have been hosted by such diverse personages as Blackfive and the George W. Bush as representatives of Iraqi blogdom. Although their blog is pioneering in Iraqi terms and has existed longer than many US blogs, Professor Juan Cole of the University of Michigan does not accept the credential. (Hat tip: Glenn Reynolds and Michael Totten). Here's how he describes them.

The MR posting brings up questions about the Iraqi brothers who run the Iraq the Model site. It points out that the views of the brothers are celebrated in the right-leaning weblogging world of the US, even though opinion polling shows that their views are far out of the mainstream of Iraqi opinion. It notes that their choice of internet service provider, in Abilene, Texas, is rather suspicious, and wonders whether they are getting some extra support from certain quarters. Contrast all this to the young woman computer systems analyst in Baghdad, Riverbend, who is in her views closer to the Iraqi opinion polls, especially with regard to Sunni Arabs, but who is not being feted in Washington, DC.

The phenomenon of blog trolling, and frankly of blog agents provocateurs secretly working for a particular group or goal and deliberately attempting to spread disinformation, is likely to grow in importance. It is a technique made for the well-funded Neoconservatives, for instance, and I have my suspicions about one or two sites out there already.

Their "rather suspicious" choice "of internet service provider, in Abilene, Texas" is www.blogger.com, the same as the Belmont Club's and a zillion others, and is chiefly doubtful because Blogger charges no fees, an attribute always sure to attract the worst of mankind. Strangely, it is also the internet service provider of Riverbend, a blog which Cole professes to admire. This illustrates the problem with credentials. Iraq the Model and Riverbend are very similar in terms of attributes like nationality, presence on the web, support from friends and support fromr readers. Yet it would be hard to construct a credential from those criteria which would pass muster as a Juan Cole-approved 'mainstream Iraqi blogger' -- without including the critical requirement that they be anti-American. Cole is certainly entitled to prefer Riverbend. What is difficult is to accept an abstract construction which is ultimately contingent on Cole's personal taste. Credentials are never exactly equivalent to the object they represent, but if they are to have any meaning, they have to be regarded as arbitrarily legitimate after a finite number of correspondences.

Interestingly, Juan Cole wrote an article critical of fundamentalist tendencies in a particular religion whose identity I will withhold, but whose name you can discover by following this link. Here are its forboding characteristics as set forth by Cole himself. (Hint: It makes wide use of the Internet, has many adherents in Iran and is not Christianity). The intellectual challenge is to discover why these attributes should be condemned in this particular religion and not in another with the same 'fundamentalist' credentials.

  1. Sectarianism and exclusivism;
  2. Theocratic tendencies;
  3. Absolutism and Inerrancy of Scripture;
  4. Millennialism;
  5. Authoritarian Organization;
  6. Strict Behavioral Requirements