April 24, 2006

Contemplation and Cogitation On Chatty Canaries, Cages, and Current Classified Controversies

Hatched by Dafydd

I've been thinking about "canary traps" for the last couple of days, ever since Rick Moran of Right Wing Nuthouse suggested the possibility that the "secret CIA prisons in Eastern Europe" story was just such a thing -- though he did not use that specific term (he said "sting"). It's actually a fascinating subject that lends itself to logical extrapolation.

What the heck is a "canary trap" anyway?

Suppose you believe you have a mole or leaker in your organization, and you want to find out who it is. Let's say you've narrowed it down to two suspects (the base-level canary trap).

You approach each suspect separately and give him supposed specifications on a new fighter jet the company is designing. You tell each suspect -- call them Fred and Barney -- that this information is absolutely top-secret, and they're to divulge it to nobody, inside or outside of the company. "Aviation Week would pay millions for these specs!"

In reality, however, there is no such airplane; you just made it up.

Label the specs you give to Fred "Specs-F;" Barney gets "Specs-B." Specs-F and -B each describe an advanced fighter jet, and they are substantially similar to each other... except that in Specs-F, there is a performance summary that says:

The Bat-Jet has a very high ceiling, but its real advantage is that it can fly at Mach 3.2 for brief periods of time.

But in Specs-B, the performance summary says:

The Bat-Jet is as fast as the fastest current fighter, but its real advantage is that it can operate up to altitudes of 175,000 feet above mean sea level.

Now you keep reading Aviation Week assiduously: when the article appears, if it includes the datum that the plane can fly Mach 3.2, then you know that Fred is the leaker; if it says the plane can fly up to 175,000 feet, then you know it's Barney. If both "facts" appear, then you fire both those two clods!

Let's generalize: if there are a twenty possible suspects, you prepare twenty different versions, each of which is substantially the same except for particular "enticing" details, something no journalist could resist including. If there are a hundred different suspects, you prepare a hundred versions: the mail-merge ability of word processors make this fairly easy: you need only keep the unique paragraphs in a database, along with a link to the person who received it. When the details appear, the inclusion or exclusion of various (invented) facts point the finger at the leaker -- or leakers.

In this case, we can suppose that Porter Goss may have distributed, say, ten different versions of the secret-prison story to ten top Agency employees -- those who had already been identified as likely leakers on the basis of earlier, actual classified intelligence that found its way to the antique media... with the list of suspects possibly filtered on the basis of obvious political biases of particular CIA employees.

If Goss did this, then the version Mary O. McCarthy got -- and leaked -- had some uniquely identifying details not found in any other version. When those details showed up in Dana Priest's story, Goss knew that McCarthy was the leaker... it remained only to focus the investigation on her until they could prove it.

Who would use such a weird tactic?

Canary traps have become ubiquitous in fields ranging from Hollywood screenplays to computer-software error-trapping. Studios use them to discover who has been copying scripts and selling them to rival production companies: each suspect gets a slightly different version of the script, and your own mole inside Colossal Pictures makes a copy of the version that arrived there: et voilà, you know who leaked it.

I have used canary traps myself in debugging the simplistic, clunky, and useless programs I've written for myself. In this case, I'm not trying to find a "leaker;" I'm hunting for the particular program object that is crashing the application.

I program each object in the code to print out its name before doing whatever it's supposed to do. When the program crashes, the last name on the screen is the module in which the crash occurred. (There are cleverer ways of doing this, but they require cleverer programers than I!)

Police use them, businesses use them; canary traps have shown up as plot devices in TV shows such as Mission: Impossible. You can even use them yourself, if one of your friends is gossiping about you behind your back. Because of how the practice has spread, it's tough to track the provenance of the term "canary trap" itself; many people say Tom Clancy originated it in his novel Patriot Games, but I vaguely remember seeing the term long before 1987, when that book was published. Has anybody ever asked him?

Hasn't the CIA already admitted the story was true?

John Hinderaker of Power Line is skeptical that the secret-prisons story could be a canary trap, among other reasons because some high-ranking CIA officials have "confirmed" that such prisons existed. But actually, that's not good evidence.

One point about a legal- or intelligence-related "canary trap" is that you can never admit it was a canary trap. In the present case, that fact would be known only to Porter Goss and perhaps one or two people he trusts utterly, the ones who helped him concoct it. The reason is twofold:

  • First of all, if the defense attorney found out that the intel leaked was concocted, then he would argue that it constituted "entrapment," making it that much tougher to get a conviction. Is this dirty pool? Perhaps... but not as dirty as leaking critical classified intelligence to the Post or the New York Times, for our enemies to read (I mean al-Qaeda, not Reid and Pelosi).
  • Second, the attorney would argue that since the leaked intelligence wasn't real, it couldn't have harmed national security; this might get his client off the hook. The fancy lawyer could always argue that the defendant knew it was fake all along and would never, ever have leaked real intelligence!

    Of course, if that were true, then the attorney would already know it was fake, and he would make that argument even if the prosecutor didn't tell him. It only makes a difference if the defendant really thought it was real, but his lawyer plans to lie about it.

So Goss will never willingly reveal that it was a canary trap, whether it was or wasn't.

Instead, the basics of the story (secret prisons) must be circulated department-wide as if they were true, because the suspect -- let's use "Mary McCarthy" as a name chosen at random -- will surely nose around first and try to find corroboration before leaking it to anyone.

Even if McCarthy doesn't suspect a trap, she may still suspect the ubiquitous bum intel that permeates any huge government bureaucracy like the Agency. If she hears the same story from five or six colleagues, however, she'll believe it and will be ready to pass it along to the Washington Post.

Thus, most of the story (let's say 90%) is circulated to everybody in the CIA who would be likely to know about such prisons if they really existed. Naturally, this list includes some high-level people: only the fine details would change from suspect to suspect.

And therefore, when the scandal breaks and top CIA personnel are hauled before Congress and questioned, nearly every one of them, from the bottom to almost the very top, will corroborate the basics of the story... because they actually think it's true. Even those two or three officials who know for a fact that it's false will nevertheless pretend it's true, just to maintain cover.

Note that this is not proof that the whole story was a canary trap, of course; but the mere fact that high-ranking officers believe it's true is likewise not proof that it is.

So is it is, or is it ain't?

Certainly, I have never heard any CIA agent claim actually to have been inside one of these secret prisons; nor have any guards been produced, nor administrators, secretaries, or even janitors. No buildings have been found, and you'd think a building as big and solid as a prison wouldn't be moved around very often.

None of the people who claim to have been held there as prisoners can lead investigators to a site or identify any of their captors or interrogators. If even the Europeans are saying there is no evidence -- when we know they're predisposed to accept "evidence" that would be laughed out of an American court -- that raises huge, huge suspicions. These suspicions are not insurmountable... but they're darned close.

Which Hinderaker himself notes, asking if "this is one secret the CIA has actually been able to keep, but for the leak."

At first, I was very skeptical of Rick Moran's canary-trap hypothesis; but now I'm only somewhat skeptical... and the only reason I'm still "somewhat skeptical" is that it would be such a Christmas gift if the whole story were concocted just to catch the chatty canary -- and I don't believe in Santa Clause.

But I sure I wish some reporter would just stand up and ask Director Goss flat out, "Porthos, old friend, tell us true: was the 'secret prisons in Eastern Europe' story just a canary trap?" On video; I want to see his facial reaction!

So how do you think he would react? Would he freeze just for a moment before sliding into a speedy denial? Would he show a reaction that Decker would notice in the movie Blade Runner? Or would he be so smooth, studied, practiced, that we would be left groping in the dark, as we are today?

I don't know, but I'd sure like to find out.

Hatched by Dafydd on this day, April 24, 2006, at the time of 3:42 AM

Trackback Pings

TrackBack URL for this hissing: http://biglizards.net/mt3.36/earendiltrack.cgi/696

Listed below are links to weblogs that reference Contemplation and Cogitation On Chatty Canaries, Cages, and Current Classified Controversies:

» Kappes In Their Hands from Big Lizards
The New York Times reports the imminent return of "storied operative" Stephen R. Kappes to the CIA fold. In his old office at the Central Intelligence Agency, Stephen R. Kappes once hung a World War II-era British poster that announced,... [Read More]

Tracked on May 30, 2006 4:34 PM

» Links and Minifeatures 04 24 Monday from Searchlight Crusade
RINO Sightings Recommended: Politburo Diktat, Mind of Mog, Techography, Carnival of Personal Finance Recommended: All Financial Matters (Monte Carlo hypotheticals are better than planning based upon average returns!) Carnival of The Capitalists is in a... [Read More]

Tracked on September 15, 2007 3:14 PM


The following hissed in response by: MTF

When the history of all this is written, these kinds of questions will be interesting. Right now, I'm more interested in why she hasn't been indicted, marched before a judge, and preparing to defend her actions in a trial. She is a spy, and like any other spy should face the consequences of her actions. She hasn't been arrested, and the Post says she may not be even under criminal investigation!

I also continue to hope McCarthy isn't the last person caught! Who helped her? Who led her? But mostly, I hope she doesn't get a free pass. C'mon Goss.

The above hissed in response by: MTF [TypeKey Profile Page] at April 24, 2006 4:10 AM

The following hissed in response by: cboldt

"If even the Europeans are saying there is no evidence"

The EU equivalant of our Homeland Security Chief, Gijs de Vries, is saying there is no proof. His audience, being similar to our Congress, has expressed scepticism at his conclusion.

More on Mr. de Vries, and more on the investigations can be viewed here (temporary committee on CIA rendition and incarceration) and here (alleged secret detentions).

Long story short, the UE, like the US, has left and right leaning factions. The terrorism officials are tight lipped, e.g., "there is no proof," but there are accusors (e.g., al Masri) whose stories feed the people who are predisposed to find a secret CIA operation. The jury is still out.

The above hissed in response by: cboldt [TypeKey Profile Page] at April 24, 2006 4:15 AM

The following hissed in response by: MTF

The Washington Post reported on Saturday:

"CIA officials said the career intelligence officer failed more than one polygraph test and acknowledged unauthorized contacts with reporters."

The same Post article said:

"Intelligence officials, speaking on the condition of anonymity, said the dismissed officer identified by others as McCarthy has not been charged with any crime and is not believed to be the subject of a Justice Department investigation."

Ignoring the possible harm to our war effort, to the nation's ability to conduct intelligence operations and to the simple task of keeping secrets, the Post instead pompously says this whole campaign against leaks is instead a conspiracy to silence the MSM press:

"The effort has been widely seen among members of the media, and some legal experts, as the most extensive and overt campaign against leaks in a generation, and has worsened the already-tense relationship between mainstream news organizations and the White House."

Unbelievable arrogance.

The above hissed in response by: MTF [TypeKey Profile Page] at April 24, 2006 7:04 AM

The following hissed in response by: Doc-obiwan

Yes, "canary traps" somewhat pre-date Mr. Clancy's 1987 work. My recollection is Sun Tzu, "doomed spies" -- or a variation thereof --would fit the bill.

The above hissed in response by: Doc-obiwan [TypeKey Profile Page] at April 24, 2006 1:31 PM

The following hissed in response by: Dafydd ab Hugh


"Intelligence officials, speaking on the condition of anonymity, said the dismissed officer identified by others as McCarthy has not been charged with any crime and is not believed to be the subject of a Justice Department investigation."

This is actually misleading in the extreme. Yes, it's true that McCarthy is not the subject of a Justice Department investigation... but as I understand it, that's because the CIA itself has retained control of the criminal investigation, not yet handing it over to the Justice Department.

In other words, there is a criminal investigation, and it probably encompasses Mrs. McCarthy: it's just being conducted by the Agency itself, not by Alberto Gonzales.

And of course she hasn't yet been charged; but that doesn't mean she won't be indicted by a grand jury later.

By the way, the XHTML tag for italics is <em>stuff inside</em>. I don't think <i> slips through the code trap here.

Boldface is <strong>...</strong>.


The above hissed in response by: Dafydd ab Hugh [TypeKey Profile Page] at April 24, 2006 1:58 PM

The following hissed in response by: Dafydd ab Hugh


My recollection is Sun Tzu, "doomed spies" -- or a variation thereof --would fit the bill.

I mean the term "canary trap" itself, not the technique. Can anyone find the term "canary trap" in print prior to 1987? I vaguely remember seeing it before then, but Googling is almost impossible because of the large number of hits and the inability to sort the results by date of publication.


The above hissed in response by: Dafydd ab Hugh [TypeKey Profile Page] at April 24, 2006 2:01 PM

The following hissed in response by: Mr. Davis

An implicit assumption is that McCarthy is the whole story. Given that Plame, Wilson, Scheuer and others have been playing some form of this game, the non-announcement of McCarthy's termination may have been to flush out others. I wonder if the NSA is cooperating with the CIA in its investigation.

Admissability of evidence is also going to be concerns of the CIA legal division in bringing any action.

Finally, I suspect there will be no prosecution because of the fear of what the defence could bring out in open court. I would be surprised if McCarthy loses her pension.

The above hissed in response by: Mr. Davis [TypeKey Profile Page] at April 24, 2006 2:33 PM

The following hissed in response by: bpilch

I think it is pretty clear from your own example that canary traps were around a lot longer, maybe even prehistoric times, if they were trying to catch Fred or Barney.... Yabba Dabba Doo!!!

The above hissed in response by: bpilch [TypeKey Profile Page] at April 24, 2006 2:59 PM

The following hissed in response by: Bill Faith

I can't remember hearing the phrase "canary trap" before Clancy used it but I don't think it would fit his style to invent the term; his work is too well researched. I just assumed the first time I saw him use it that it must be the term actually used in certain circles I'd never been invited into.

The above hissed in response by: Bill Faith [TypeKey Profile Page] at April 24, 2006 5:05 PM

The following hissed in response by: Bill Faith

I linked from Canary Traps Explained.

The above hissed in response by: Bill Faith [TypeKey Profile Page] at April 24, 2006 11:01 PM

Post a comment

Thanks for hissing in, . Now you can slither in with a comment, o wise. (sign out)

(If you haven't hissed a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Hang loose; don't shed your skin!)

Remember me unto the end of days?

© 2005-2009 by Dafydd ab Hugh - All Rights Reserved