In 1986, the CIA invited Tom Clancy to discuss how he made military technology intelligible to a general reader. The NSA placed him inside its Security Week programme. What survives is a fragmented but unusually rich field manual on research, judgment and communication.
CIA records establish the subject, date and audience of the Langley presentation. The NSA school catalogue records a 48-minute, 24-second programme. No public transcript of the CIA talk appears in the released file; the NSA recording survives.
DECLASSIFIEDCENTRAL INTELLIGENCE AGENCY27 FEB 1986“…presenting complex technical matters to the lay reader.”
CIA announcement, 27 February 1986
This dossier keeps documented fact, collaborator testimony and professional synthesis visibly separate.
An insurance broker with no military service publishes a novel built from open sources, conversations with former submariners and persistent technical questioning.
OSWR and the Office of Central Reference invite him to discuss methods used in Hunt and Red Storm Rising to make technical subjects accessible.
The NSA catalogue describes a talk on the extensive research behind the novels and what happened after they appeared.
The Potomac Chapter lists Clancy as speaker. No public transcript has been located.
Clancy closes a day devoted to printed communication, security material design and motivational writing.
The official history names Clancy among former speakers; the lecture text has not been located.
Clancy left no treatise on analysis. His method emerges from how he found sources, the questions he asked, his work with Larry Bond and the institutions surprised by how much an outsider could infer.
He combined professional journals, fleet guides, manuals, maps, wargames, commercial imagery and practitioner conversations. No single item had to contain the answer.
The advantage is created in fusion. Assess the credibility of each fragment and the validity of the relationship you construct among them.
The Harpoon rules explained sensors, weapons and interactions for naval officers who were not necessarily gamers. The system had already been translated into a manipulable model.
Training manuals, operator guides, incident reports and simulations may explain causality better than the most technically advanced source.
He took manuals apart and reorganised them in binders. Reading became a working system, not a collection of finished books.
An archive without a path between claim, evidence, contradiction and caveat is accumulation rather than analytic capacity.
He kept asking until he understood what a term meant operationally: sensor geometry, error, system interaction and the proper wording of an order.
Terminology encodes the distinctions a profession believes matter. Approximate definitions produce approximate models.
Specifications were available. Harder was learning how people think inside a submarine, what they notice and what they regard as normal.
Written procedure does not reveal who bypasses it, which signal a veteran notices or which error carries shame inside a unit.
When speaking to Arkady Shevchenko, he asked where meetings took place, what the room looked like and whether water or vodka was served.
Who sits where, who speaks first and what happens when the boss is wrong may reveal real hierarchy better than an organisation chart.
He began with the problem a platform had to solve and inferred the characteristics required to make the mission feasible.
Ask what must be true. What infrastructure, trade-offs and observable signatures would that capability require?
He did not stop at weapon range. He followed environment, detection, classification, decision, engagement, effect and adversary response.
Technical detail earns space when it changes decision time, options, risk, detectability or effect.
Fiction needs a coherent sequence even when it is rare. Intelligence must calibrate how often that sequence could occur.
After building the compelling scenario, count assumptions, check the base rate and search for the simpler explanation.
Larry Bond stressed that games do not prove an outcome. Repetition makes the factors controlling the contest reappear.
Vary warning, readiness, logistics, rules of engagement and competence. Look for recurring dependencies, not the final score.
In wargaming he searched for unusual uses of systems, including decoy, masking and deception roles.
The adversary is under no obligation to attack doctrinally, elegantly or in the manner your defence was optimised to meet.
Technology appeared when it changed a choice: a contact must be classified, a torpedo evaded, a decision window closes.
Just-in-time explanation preserves mechanism while reducing cognitive load. The reader understands because there is an immediate problem to solve.
He selected figures that established scale, feasibility or danger rather than dropping the data sheet into the scene.
Speed becomes warning time. Accuracy becomes weapons required. Production capacity becomes replenishment tempo.
A platform became credible through its operators, command culture, training and the tension between different mandates.
Inventory shows what exists. Doctrine, maintenance, morale, authority and tolerance for failure show what can be employed.
In dealings with the CIA, he explained what came from commercial sources and what had been deliberately altered in fiction.
Label observed, reported, corroborated, inferred, assumed and modelled elements before polished prose levels them into apparent fact.
He argued that secrecy is a national-security tool and that deterrence requires an adversary to know enough to be afraid.
The useful question is how much of the judgment can be used, by whom and at what abstraction, without exposing the source or vulnerability.
Select the sources. The explanatory model will strengthen, along with the danger of mistaking coherence for truth.
There are not enough pieces for an explanation. Begin with a source.
An operational version of the Clancy method, adapted for a briefing, assessment or open-source investigation.
State what the consumer must understand or decide, rather than only the research topic.
Actors, objectives, sensors, communications, dependencies, constraints and decision points.
Seek nominal performance, actual performance, doctrine, organisational culture and industrial capacity.
Keep observation, report, inference, assumption and model output distinct.
Generate several coherent chains and specify indicators that would strengthen or weaken each.
Introduce technical detail where it changes what an actor can see, choose or accomplish.
First truth and uncertainty; then disclosure risk and the minimum detail required.
Check only what your product demonstrably achieves. The score remains in your browser.
A memorable scenario feels more likely because it can be imagined in detail.
Accuracy about a bolt lends borrowed authority to an untested strategic judgment.
Access to one expert is confused with complete truth about an institution.
Source sensitivity is treated as a substitute for reliability.
Real fragments are connected into an elegant model that nothing attempts to falsify.
Narrative momentum pushes alternatives and gaps out of the reader's field of view.
The public file preserves the invitation, logistics and honorarium trail. It does not preserve a transcript or recording. Any list of ‘lessons delivered at Langley’ that goes beyond the announced subject must be treated as reconstruction.
Rigour begins where you refuse to turn an archival gap into a convenient memory.
Links open the original documents or recordings. Labels show how directly each supports the dossier.