Scott Talkington of Demosophia examines the Sy Hersh "Stovepipe" article in the New Yorker on CIA/Neocon conflicts over intelligence evaluation.

His continued points about alpha method/beta method judgment - roughly, "presumed innocent" vs. "presumed guilty" - are interesting. I fear that his approach is somewhat theoretical, and his argument that administration Neocons were attempting to impose a "beta method" intelligence evaluation method strikes me as, itself, demonstrating the limits of an "alpha method" approach. That is, I think he's giving them too much credit, and presuming their intellectual innocence.

The second half of the article examines Hersh's wild conspiracy story about ex-CIA forgers, and I think Talkington is absolutely right that the utterly dysfunctional and borderline treasonous attitude of the CIA reported by Hersh is far more alarming than the typical, tired story of overly pessimistic Neocons looking for threats under every Eurasian rock. Is Hersh's reporting *accurate*? At this point, I feel I have to approach Hersh with a "beta method" - he's been the source of too many misleading stories in the last decade for it to be safe to presume his work "innocent".