The First Five Seconds: “Spy the Lie” by Philip Houston, Michael Floyd and Susan Carnicero

Most writing about deception detection is built on the premise that liars leak — that stress produces tells, and a trained observer can read them. Spy the Lie starts somewhere narrower and more defensible: ignore general demeanor and look for a cluster of indicators whose first sign begins within five seconds of a question.

The authors are three former CIA officers — Philip Houston, Michael Floyd and Susan Carnicero — writing with Don Tennant. Houston developed the interviewing method at the Agency and later carried it into the private sector; the book, published by St. Martin's Press in 2012, is that method written down for a general reader.

The core discipline is the discarding. Almost everything a person does in an interview is noise: nerves, temperament, dislike of being asked. The authors link a cluster of indicators to a specific question by requiring its first sign to begin within five seconds; the response can continue beyond that interval.

The claim to treat carefully is the accuracy one. Deception-detection research has been unkind to the field generally, and while this method is more disciplined than most, the book does not engage seriously with the literature on how poorly humans perform at this task. Use it to structure an interview, not to reach a verdict.

Summary

Spy the Lie, by Philip Houston, Michael Floyd and Susan Carnicero with Don Tennant, sets out a deception-detection method drawn from the authors’ CIA careers. The procedure is to ask a clear, short question and look for a cluster whose first indicator begins within five seconds. That response is classified against a defined list of deceptive indicators. No single indicator counts, and the unit of analysis is a cluster of two or more. The chapters on question construction carry much of the practical value, because an ambiguous question cannot produce a readable answer. Since the book says little about false positives, readers should use the method to choose the next question rather than to judge a person.

What the book actually covers

The method is procedural rather than intuitive. Ask a clear, short question. Start a mental clock. Look for the first indicator within five seconds, then consider the response that follows. Classify what you see against a defined list of deceptive indicators — among them failure or reluctance to answer, repeating an earlier answer, inappropriate questions, aggressive replies, overly specific answers, excessive politeness and qualifiers.

Crucially, the authors insist that no single indicator means anything. A cluster of two or more in response to the same question is the unit of analysis. The cluster can include later behavior in the same uninterrupted response; general demeanor is not enough.

The chapters on question construction are the most practically useful and the least glamorous. A badly built question produces an ambiguous response that no amount of observation can resolve, so most of the work is done before anyone answers anything.

The weakest material is the case studies, which are chosen to demonstrate the method working and inevitably read that way. There is little on false positives — on the honest, frightened interviewee whose cluster of indicators means only that they are frightened — and that is the failure mode most likely to cause real damage in a corporate investigation.

Why it matters for your risk posture

Read it because the interview is usually the highest-value hour in an internal investigation and is almost never trained for. Investigators are trained on documents, on process, on evidence handling. The conversation is left to instinct, and instinct is where the errors are.

Read it for the discarding discipline specifically. The most common interviewing error is over-reading — building a theory from a shifting posture or a broken gaze, then interpreting everything afterward through it. A method whose first instruction is to ignore almost everything is a corrective to exactly that.

Finally, read it against its own limits. In an employment or misconduct context, a deception assessment is not evidence and should never be recorded as though it were. Its legitimate use is to direct the next question and the next document request. Treated as a finding, it becomes a liability — a judgment about a named person, made on a contested method, in a file that may later be disclosed.

In a corporate investigation, interviews come after documents have been reviewed and evidence preserved. That order lets each question be short and built around what the record already shows. We also allow for the frightened and truthful interviewee whom the authors largely leave out. A cluster of indicators therefore leads to a further document request or a follow-up interview, and it is never written up as a conclusion about the person.

Key takeaways

• The first indicator must begin within five seconds of the question. Look for a cluster in the response that follows.
• No single indicator means anything. Two or more in response to one question is the unit of analysis.
• Most of the work is in the question. Ambiguous questions cannot produce readable answers.
• Over-reading is the standard failure. A method that discards aggressively is a corrective to instinct.
• It directs inquiry; it does not establish fact. Never record a deception assessment as a finding about a person.

About the authors

Philip Houston spent twenty-five years at the CIA, where he developed the interviewing and detection method the book sets out, and later applied it in corporate investigations. Michael Floyd and Susan Carnicero are also former CIA officers with backgrounds in interrogation, screening and behavioral assessment; all three subsequently worked in the private sector applying the same approach.

The cover credits the three of them "with Don Tennant," the collaborating writer — a distinction worth preserving, since he is frequently listed as a fourth co-author and is not one.