This is the precisely thing I don’t understand. The entire argument for the document having been faked by Gleick is actually predicated on the assumption (which you’re developing above) that Gleick is unbelievably stupid. The only other plausible assumption that the error is actually a comparatively small one in a memo that doesn’t really contain any actual information not available in the other documents. But Ockham’s razor might not be that helpful here.

The Kock link is a big deal for anyone with an intellectual stake in the dissemination of those papers. If a faker was trying to draw attention to it, then he can’t hit a fish in a barrel. He’s irredeemably dumb. The question is, is this a fair assessment of Gleick? One may not agree with him, or like his position, and of course there are plenty of successful people (academics, businessmen, politicians) out there who are simply not very clever. But a basic academic training should at least provide the basis for a better piece of fakery than this. The one thing academics can usually manage is to produce internally consistent documents. And I would argue that someone faking a document such as the memo is much more likely to over-correct than to mess up one of the key points on which we must assumed they were focussed. If Gleick did this, you’re right: he’s a dumb-ass. But it smells pretty fishy to me.

The consensus that the /style/ of the memo identifies Gleick as its author seems to me to be unreliable. We often think that this kind of analysis is straightforward, because we all know how to use language (don’t we?). We’re much happier to admit we don’t /get/ statistics. But any adequately trained student of language or literature should be able to tell you that linguistic analysis and stylistic ‘fingerprinting’ are seriously tricky matters, full of pitfalls. Analysis of written style is what I do for a living. No statistician will be surprised to here that in my line of work the road to hell is paved with false positives. Sloppy usage is an entirely unremarkable feature of official memos, internet blogs, and academic emails alike. Commas won’t do it for you. Even the phrase ‘anti-climate’ is no identifying mark (unless used consistently) when you can make it by accidental omission, or pick it up after unthinkingly from someone (even an opponent) whose words you’ve been reading recently. You need more than this to make a confident attribution, if you want to get beyond the pleasure of wishful thinking. Anomalies in the /factual/ content of such an official memo are an entirely different and much more significant matter.

Steve: you’re overlooking “means, motive and opportunity”, which Gleick had. The author of the fake memo had to be in possession of the actual documents. “Opportunity” limits the population to Gleick and the Heartland directors (and maybe a few senior Heartland staffers.) The “fiber” evidence is additional.

Obviously, a police investigation ought to examine Gleick’s computers and computers that he might have used. Police seized the UEA server almost immediately, but they don’t seem to have done anything with Gleick’s.