The FBI publishes a brochure on elicitation aimed at employees of companies with information worth stealing. It is short, plain, and considerably more useful than most books on the subject, because it lists the techniques and then explains that the target's experience is of having a pleasant conversation.
That is the definition. Elicitation is the extraction of information through conversation that the source does not perceive as questioning. Because it never crosses the threshold at which a person recognizes that they are being interviewed, none of their defenses activate. Nobody guards against a chat.
The catalogue is small and the items are recognizable once named.
Flattery and the appeal to ego. People correct experts and instruct novices, and both are disclosures. Deliberate false statements, which exploit the near-irresistible pull of correcting an error — the surest way to obtain a number is to state the wrong number confidently. Feigned ignorance, which lets the target explain. Quid pro quo, in which the collector discloses something first, activating reciprocity. Oblique referencing, approaching a subject from the side so the sensitive item arrives as an aside. Assumed knowledge, which is chapter 101's technique. Bracketing, offering a range so the target narrows it. Criticism of the target's organization, which reliably produces a defense containing detail. And the long silence, which people fill.
The mechanisms are the ones in Part 1 plus one that is specific to this chapter.
Diana Tamir and Jason Mitchell demonstrated in 2012 that self-disclosure is intrinsically rewarding: talking about oneself engages mesolimbic dopaminergic regions associated with reward, and participants in their study forgone money for the opportunity to do it. Disclosure is not merely something people are willing to do. It is something they are motivated to do, and a skilled collector's main task is to create the conditions rather than to ask.
Which is why the conversation feels good, and why the target leaves it in a better mood than they entered.
The settings are the ones where professional talk is normal and expected: conferences, trade shows, recruitment approaches, industry dinners, online professional networks, and the long friendly relationship that develops over eighteen months before anything is ever requested. Chapters in Part 8 of this guide document what this looks like when run by a state service.
The countermeasure is not suspicion of friendly people, which is unlivable and does not work anyway. It is a rule about categories rather than about interlocutors: decide in advance what you do not discuss outside your organization — architecture, staffing, schedules, vulnerabilities, who holds which access, what is not working — and hold that line regardless of how the conversation feels. A category rule can be applied when your judgment is compromised by enjoying yourself. A judgment call cannot.
And the diagnostic sign, when it appears, is asymmetry: a conversation in which you have learned nothing about them and they have learned a great deal about you, and in which you cannot remember being asked a single question.
Cross-ref: Scharff technique chapter