A Guide to Federal Surrealism: Fiction and the American Intelligence Community

By Grayson ClaryJanuary 17, 2015

A Guide to Federal Surrealism: Fiction and the American Intelligence Community

IN A FEBRUARY 2014 ESSAY for The Chronicle of Higher Education, Eric Bennett assessed the CIA’s role in funding and promoting the Iowa Writers’ Workshop. By Bennett’s account, the Agency has long had a role in the development of American fiction. Worried that Soviet schools would instill communist ideals in foreign students, the CIA funded competing cultural institutions that were designed to lure the world’s left-leaning thinkers to American programs that “fortified democratic values at home and abroad.” Spooks shepherded Pasternak’s Doctor Zhivago onto the world scene; spooks underwrote the Iowa Writers’ Workshop and The Paris Review.

Although the relationship is no longer so clear, this unexpected link between creative writing and the Central Intelligence Agency continues. As the Agency continues to grow and publish reports on its own activities, it has developed a style all its own. Federal Surrealism is the American intelligence community’s house style: an entire genre complete with style manuals, recurring characters, and an ever-stunning series of plot twists. While the public has no access to most works in the federal tradition, some pieces are revealed in the declassification of major reports. Notable entries include the Warren Report, Style Manual & Writer’s Guide for Intelligence Publications, and The Committee Study of the Central Intelligence Agency’s Detention and Interrogation Program, commonly called “the torture report.” The Senate Select Committee on Intelligence (SSCI) released the last, which you’ll want to know is pronounced “sissy.”


The founding work of secret literature is the Report of the President’s Commission on the Assassination of President Kennedy — the Warren Report, so styled after chairman and Chief Justice of the Supreme Court of the United States Earl Warren. The Warren Report lays down certain essential characteristics of the clandestine genres. The first is hyperbolic length. Simply put, the literary output of the national security apparatus is not made for reading. This isn’t a straightforward question of classification; the brute fact of a 15-volume report denies access, classified or no. This is true of the mammoth pile of Pentagon Papers and true too of the 6,000-page Senate report on CIA interrogations, of which only the 600-page executive summary has been made public. A telling anecdote: there are exactly 11 words on one page of the Pentagon Papers that were slated for redaction but which the black pen accidentally overlooked. Are they any less secret for it? Hardly. Length obscures, dissolving guesses, facts, and half-truths into a murky stew.

You might suppose that length signals an interest in totality, but you’d suppose incorrectly — the IC doesn’t believe in totality. National security literature is anti-encyclopedic, anti-epic, anti-novelistic. This solves a literary mystery: Is there a long-form fictional prose genre other than the novel? Of course. There is the Warren Report. This is a judgment from epistemology, not conspiracy. Lee Harvey Oswald killed John Fitzgerald Kennedy, after all. But could the Warren Report be expected to communicate the same? A 15-volume report is inevitably less complete, less closed, than the seven words reading, “Lee Harvey Oswald killed John Fitzgerald Kennedy.” Deep state literature never records the secret itself. Not really. The works are all Janus-faced ritual texts, spells that by turns summon and banish the classified from the corner of your eye.

The intelligence community is deeply invested in shades of truth, untruth, known, and unknown (a reference to Donald Rumsfeld springs to mind). Spy writers use a very precise vocabulary on these points. They would like you to understand that “covert” and “clandestine” describe two entirely different varieties of action; that employees of the Central Intelligence Agency are officers, not agents; and that they deal in both secrets and mysteries. In that spirit, the Warren Report is fictional; it deals in secrets, mysteries, things not tame enough for nonfiction. On this point, I would refer you to the attitude evoked by Director John O. Brennan in his response to The Committee Study, where he writes,

The Agency takes no position on whether intelligence obtained from detainees who were subjected to enhanced interrogation techniques could have been obtained through other means or from other individuals. The answer to this question is and will forever remain unknowable.

The emphasis is mine, though who could resist adding it?


The gold standard for terminological precision — another quintessentially intel quirk — is set by Style Manual & Writer’s Guide for Intelligence Publications. This was compiled by the Central Intelligence Agency’s Directorate of Intelligence, though in a sense, per its preface, it has many fathers: “The works of Barzun, Bernstein, Copperud, Follett, Folwer, the Morrises, Strunk and White, Gregg, and other recognized arbiters of English usage.”

What’s happened, here, to the CIA’s trademark protection of sources and methods? For all we know, Barzun is now in grave danger; Copperud can fend for himself. Ancestry aside, much of Style Manual’s pedantry is innovative and telling. The second footnote, to cite one example, is beautifully hegemonic: “Note that, even in proper names, we always use the American spelling for English words spelled differently in the British Commonwealth.” The Labour Party? I think you mean the Labor Party. On this side of the Atlantic, we speak the President’s English; that much we earned in the Revolution.

Nothing is left to chance. The spy writer is advised, “In paraphrasing communist statements, put [Socialist or socialist] in quotation marks. The same applies to imperialism and imperialist (and to anti-imperialism and anti-imperialist), which are terms communists use in describing their opponents.” Not in this century, granted, but the point is well taken. There is no such thing as imperialism, only “imperialism.” There was also no Korean War or Vietnam War, though there was a Vietnam war and a Korean war. The distinction is required because these wars were “‘undeclared.’” I don’t add my own political or interpretive bias here: undeclared carries quotation marks in the original.

Most informative is Style Manual’s “Word Watchers List,” which deals with “possibly troublesome words, word types, and word problems.” Here its authors show off a heady command of language, the talent by which the security services get reality to behave itself. They are so bold as to gainsay the United States Constitution — “The Preamble […] is out of bounds grammatically when it speaks of a more perfect Union” — and they are so delicate as to advise that regime “has a disparaging connotation and should not be use when referring to democratically elected governments or, generally, to governments friendly to the United States.” Best yet: “The DI [Directorate of Intelligence] is not in the business of deciding whether something is good or bad.” More than nitpicking, this is worldview maintenance — the point where use of language becomes use of weapons.

Torture becomes “enhanced interrogation.” Word choice is world choice, the spies know.


Last but not least — it probably defines the whole enterprise — there is the intelligence officer’s habit of abusing the text, using either the footnote or the redaction. I mentioned that intelligence literature is never total; this is why. It refers incessantly to realities beyond the text, texts beyond the present text, marked usually by note or by black pen.

For example: One of the central controversies in the declassification of The Committee Study was the redaction of pseudonyms used to identify — “identify” — CIA officers (in fact, the pseudonyms were used in the first place to obscure the identities of the officers pseudo-named). The CIA and White House expressed concern that readers might connect the appearances of particular characters throughout the work, tracing given officers through incidents, through promotions, through disciplinary hearings, etc. They might, in other words, read The Committee Study and draw conclusions about its protagonists. The Senate’s investigators countered that this is the point of names and narratives per se.

This reasoning is not very welcome in secret literature. The pseudonyms were redacted.

So were the letters — individual letters, A, B, etc. — used to identify the nations that hosted CIA black sites. Never mind that they were promptly identified in The Washington Post after the report’s publication. The black pen struck; the letters were redacted. To what end? Aesthetic, I have to think. Josh Gerstein, writing in Politico, quoted Chris Anders of the American Civil Liberties Union on the point: “It looks like some kind of piece of art.” (Over the last decade, conceptual artist Jenny Holzer has been creating “Redaction Paintings” from documents declassified via the Freedom of Information Act.)

Some 7 percent of the words in The Committee Study are blacked out. On some pages, I’d eyeball the figure at 60–70 percent. My favorite segments are the email “To:” lines, which usually feature a suggestive mix of [REDACTED] and REDACTED. This has a marvelous world-building kind of effect. The clandestine world is peopled with secrets of every breed (you can have any color secret you want, to borrow from Henry Ford, as long as it’s black). So there’s always more to learn, something refused the reader. In this respect, Agency stories are never-ending stories; each secret has its roots in another, a chain of causal hand-me-downs that reaches back to Cain’s scheming against Abel.


Critics charge that the “torture report” — which in a feat of word choice accuses no one of torture — is not a neutral document. This is plainly true. The Committee Study­ is a work of literature, an entry in the longstanding tradition of the American intelligence community. Not long after Christmas, Melville House Publishing released the document as a painfully handsome book. (The worst gift-givers are probably those who give late and give government documents.) This republication, complete with neat borders and an attractive cover, probably confuses its readers. It creates the impression that literature of the deep state is like other literature, that classified writing is awfully similar to public kinds of writing. But no writer in any open literary tradition has ever imagined feeding a man hummus, pasta with sauce, nuts, and raisins through his rectum. Creativity of this kind is indigenous, uniquely, to the secret styles of the American intelligence community.

Ultimately, when the intelligence community commands language, more than tone or structure changes — the world bends a little too. Public polling now suggests that Americans don’t have such a problem with enhanced interrogation tactics after all. The Melville House edition of The Committee Study, meanwhile, sold out the very same day it went on sale — 50,000 copies, with a reprint on the way to meet excess demand.

I don’t know that I know how secret literature works. But I’m not surprised it sells.


Grayson Clary lives and works in Washington, DC.

LARB Contributor

Grayson Clary lives and works in Washington, DC.


LARB Staff Recommendations

Did you know LARB is a reader-supported nonprofit?

LARB publishes daily without a paywall as part of our mission to make rigorous, incisive, and engaging writing on every aspect of literature, culture, and the arts freely accessible to the public. Help us continue this work with your tax-deductible donation today!