Key Theories & Frameworks
The continuous process of intelligence: (1) Planning & Direction — requirements set by policymakers, (2) Collection — gathering raw data via HUMINT, SIGINT, OSINT, etc., (3) Processing & Exploitation — converting data into usable form, (4) Analysis & Production — creating finished intelligence, (5) Dissemination — delivering to consumers.
Established intelligence analysis as a discipline distinct from espionage. Kent's "Strategic Intelligence for American World Policy" defined the relationship between intelligence and policy — intelligence informs but does not make policy.
Systematic identification of indicators that suggest an adversary is preparing aggressive action. Developed after Pearl Harbor intelligence failure. Used matrices of political, military, economic, and social indicators.
Applying Bayesian probability to intelligence analysis — updating prior beliefs as new evidence arrives. Addresses cognitive biases, encourages analysts to consider alternative hypotheses, and quantifies uncertainty.
Analysts identify all plausible hypotheses, gather evidence for each, and then systematically disconfirm hypotheses. Evidence that disconfirms a hypothesis is more diagnostic than evidence that confirms it.
Examines why intelligence surprises occur. Betts: failures result from organizational pathologies, cognitive biases, and political pressures. Jervis: systemic, cognitive, and bureaucratic factors all contribute to surprise.
Core Concepts
Case Studies & Examples
U-2 spy planes photographed Soviet medium-range ballistic missiles in Cuba. EXCOMM debated responses — chose naval blockade over invasion. John McCone's intuition about missiles proved correct. Demonstrated the value of technical collection and structured decision-making.
US had broken Japanese diplomatic codes (MAGIC) but failed to anticipate the attack. Intelligence was dispersed across Army and Navy, no central analysis, and warning messages were not delivered in time. Led to creation of CIA and the intelligence community structure.
Israel's intelligence (AMAN) correctly had information about Egyptian and Syrian military preparations but concluded they would not attack — "the Concept." Egypt's deception plan (Operation Badr) successfully concealed attack intentions. Led to creation of the Agranat Commission.
Ames, a CIA officer, spied for the Soviet Union/Russia for 9 years. Compromised over 100 operations, led to execution of at least 10 assets. Caused by financial greed and alcohol. Exposed massive counterintelligence failures in CIA's internal security.
Practice Questions
A: HUMINT (human sources), SIGINT (signals interception), IMAGERY (satellite/aerial photos), MASINT (measurement and signature intelligence), OSINT (open sources), GEOINT (geospatial), and CYBERINT (cyber collection). Each has strengths and weaknesses — they complement each other.
A: (1) Cognitive biases — analysts see what they expect, (2) organizational pathologies — information doesn't flow, (3) political pressures — intelligence shaped by policy, (4) adversaries actively deceive, (5) complexity — too many variables, (6) information overload — signal lost in noise.
A: Analysts list all plausible hypotheses, gather evidence for each, then look for evidence that DISCONFIRMS each hypothesis. This reduces confirmation bias. Evidence that disconfirms a strong hypothesis is most valuable. Remaining hypotheses are ranked by elimination.
A: Strategic intelligence is long-term, informs national policy and strategy (e.g., China's military modernization over 20 years). Tactical intelligence is short-term, informs immediate military operations (e.g., enemy unit positions today). Strategic assesses capabilities and intentions; tactical assesses immediate threats.
A: (1) Politicization — analysts pressured to support policy, (2) policy bias — analysts' assumptions influenced by policy preferences, (3) cognitive closure — ambiguity aversion leads to premature certainty, (4) groupthink — consensus valued over accuracy, (5) communication gap — policymakers don't understand intelligence limitations.