And then I vow to go read Marshall again, and I remember why I thought he was a tool in the first place. No, your examples do *not* prove what you're arguing, you fool. Of the six examples provided of "[CIA] operative", half were ambiguous in nature. That is, unless you're reading with expectations.
1)The Spann example goes against the argument - Mike Spann wasn't a secret agent, he was a paramilitary interrogator.
2)The Haq example is deliberately ambiguous - you can't tell from that if Novak's talking about on-the-ground agents, their handlers, or the analysts somewhere up the tree who would actually *talk* to a pundit like Novak. My guess? This was a direct quote from the Langley hacks Novak cultivates.
3)The Turner example is similarly ambiguous - the operatives referred to as "Central American" seems, from the context, to represent the entire structure - agent to handler to station to section analysts - accused of relying on "unsavory local agents".
The other three are more to the point. Of those, two are from the same subject, and one refers to a review of a spy novel. In contradiction, Marshall offers three examples of "analyst" being used. Hrm. Novak is twice as likely to use "operative" as "analyst". I call "bullshit" from Marshall's own testimony.