The scene where the analysis begins
The interrogation room is designed to remove distraction. Nothing extra: a desk, three chairs, a single light source from above. The light does more than illuminate — it organises the space, foregrounding what matters and cutting away the rest. The investigator knows: in such a room what counts is not the words, but the way they appear.
Across from him sits a person. He holds himself calmly, even composed. His hands rest on the table; his fingers are steady. His gaze is direct. Looked at on the surface, everything is in order. But the investigator does not work on the surface.
To his left sits the expert. He barely moves. His task is not to intervene but to listen. Not to the content — to the structure.
— "Tell us how it happened."
The man begins to speak immediately, without a pause. The speech flows evenly, without stumbles, without going back. He arranges events in sequence, as if he already knows what comes next.
And it is precisely at this moment that the expert shifts his torso slightly. The movement is almost imperceptible, but for the investigator it carries a meaning: the account sounds correct — too correct.
Why behaviour is misleading
Here begins the territory where the everyday picture of an interrogation stops working. The intuitive cues a layman relies on turn out to be unreliable. Nervous does not mean lying. Calm does not mean truthful.
A person can be afraid of the situation itself and get tangled in his own words while being innocent. And conversely, he can be cool-headed and confident because he has prepared his version in advance.
Practice has long shown that behaviour is a weak compass. It is too variable, too dependent on personality, on experience, on psychological resilience. Professional work has therefore moved onto a different plane. At the centre is no longer the person as an object of observation, but his account as a structure.
The hypothesis that changed the interrogation
An Interrogation, in Terms of Signal
This shift is tied to the work of the German psychologist Udo Undeutsch, who formulated a principle that radically changed the analysis of testimony.
His idea is simple: a story of a really lived event differs in its inner organisation from a fabricated one. It is not a question of content, and it is not a question of emotion. It is a question of structure.
In other words, the interrogation turns into the analysis of a text — not a literary one, but a psychological one. The investigator and the expert do not work with what the person says, but with how he builds it.
How memory sounds, not a version
When a person recalls a real event, he does not replay a finished story. He reconstructs an experience. And experience does not exist as a tidy sequence. It is fragmented, riddled with accidents, full of details that do not obey the logic of a narrative.
For this reason the speech of such a person inevitably contains features that look at first glance like "noise".
He goes back: "No, wait, it was different at first…"
He qualifies: "There was a smell… like oil, or something similar…"
He adds details that do not strengthen his position: "I didn't immediately understand what was happening…"
These elements do not embellish the account and do not make it more convincing. On the contrary, they can "spoil" it from the point of view of logic. Yet they are precisely the markers of a real recollection. Because memory does not aim at perfection. It aims at restoration.
The Method That Changed European Criminalistics
Memory is not a recording. It is a reconstruction.
How a construction sounds, not an experience
An account that a person constructs is built differently. Logic, not experience, is at work here. The person builds a model — sequential, tidy, free of redundancy.
He avoids superfluous details because they increase the risk of slipping up. He aims for clarity because clarity is easier to hold in mind.
Such an account looks convincing. It is smooth, sequential, "correct". There are no returns, no doubts, no accidental insertions. Each event follows the previous one as if in a well-written screenplay.
And precisely this is what makes it suspect.
The paradox is that plausibility in the everyday sense is a poor criterion of truth in an interrogation.
Where the boundary between truth and a model runs
Microexpressions, Pauses, the Rhythm of Questions
The expert in the interrogation room does not assess how logical the story is. He assesses how well it matches the nature of memory. And if the account is too well organised, it means its organisation is not natural.
At some point he says quietly to the investigator: "He is building, not remembering."
The distinction is fundamental.
Memory depends on experience. It is bounded by what the person actually lived through. It includes elements impossible to fully control: chance details, sensory impressions, non-sequiturs.
Construction depends on logic. It rests on a person's ability to hold a model in mind and reproduce it without breakdowns. But any model has a limit of complexity.
It is here that the weak spot appears.
How an investigator dismantles a false construction
The investigator changes direction. He does not increase the pressure — he changes the structure of the questions.
— "Tell it again, but from the end."
This technique looks simple, but it destroys linearity. A logically constructed story loses its support, because its elements were linked in one direction only.
Then come the clarifications:
— "Who was standing to your left?"
— "You said it was quiet — what sounds, exactly, did you hear?"
— "What did you do immediately after that moment?"
Where Method Ends and Coercion Begins
If the person remembers, his account begins to enrich itself. If he constructs, it begins to break.
It is precisely these breakdowns that the expert records.
The chief mistake of an interrogation
Errors begin where the investigation ignores structure.
When the goal is not to understand but to obtain a confession, distortions appear:
pressure replaces analysis
a confession substitutes for the truth
the structure of the account stays out of view
Under such conditions a false confession can look convincing, while a real recollection can seem weak.
The limit of a lie
At the end of the interrogation the investigator closes the folder. The expert makes no loud declarations. They do not discuss whether they "believe" the person.
They analyse the construction.
Because in an interrogation what is decisive is not the content of the words, but their origin.
A lie can be learned. It can be rehearsed. It can be made convincing. But it has a limit. It does not withstand a collision with memory.