Case File 210 · Investigation Psychology

Why Do People Say I Don't Remember During Questioning?

"I don't remember" is usually a true statement about the state of someone's memory, not a confession or a dodge. Forgetting is the normal fate of most of what we experience; detail fades quickly, and

By Murder Book Editorial TeamPublished by Gungrounds, the studio making Murder BookPublished 3 October 2026Updated 3 October 2026

"I don't remember" is usually a true statement about the state of someone's memory, not a confession or a dodge. Forgetting is the normal fate of most of what we experience; detail fades quickly, and ordinary events are encoded loosely in the first place. Stress at the time of an event or during questioning can make retrieval harder, time widens the gap, and saying "I don't know" is often the accurate and responsible answer rather than guessing. People sometimes also use "I don't remember" strategically, to avoid committing to a detail or to buy time. But because genuine and strategic uses look the same from the outside, the phrase cannot tell you, on its own, that a person is hiding something. It marks a gap to probe carefully, not guilt to infer.

Key takeaways

Forgetting is normal and fast; not remembering a detail is the default outcome for most experiences, not an anomaly.
Acute stress, both during an event and during questioning, can impair memory retrieval.
Saying "I don't know" can improve the accuracy of what a person does report, because they are withholding a guess rather than inventing one.
Pressuring someone past "I don't remember" risks producing confabulated or coerced detail that feels like memory but is not.
The same words can be genuine or strategic, so "I don't remember" is not evidence of lying or guilt.
A careful interviewer treats the gap as something to explore with better cues, not as an admission.

Why do people forget so much?

Most of what happens to us is never retained in any durable way. The classic demonstration is Hermann Ebbinghaus's forgetting curve, which showed memory dropping off sharply soon after learning and then levelling out. A 2015 study by Murre and Dros successfully replicated the original curve, confirming the basic shape more than a century later: a steep early loss followed by a slow decline.

Everyday events are also encoded far less completely than we assume. We attend to the gist and discard most of the specifics, because remembering every licence plate and timestamp would be useless. So when someone cannot recall what colour shirt a stranger wore or the exact order of events, they are describing the ordinary limits of memory. Why memory is not a recording explains why complete recall was never on the table.

Does stress make "I don't remember" more likely?

It can, in two different ways. Stress around the time of an event shapes what gets stored, often sharpening memory for the central, threatening detail while leaving the periphery thin. Stress at the moment of recall is the part people overlook. A 2017 meta-analysis and review by Shields and colleagues concluded that acute stress can significantly impair memory retrieval — the very process an interview depends on.

That matters because questioning is itself stressful, especially for someone who feels suspected. The pressure of the room can make it genuinely harder to retrieve details that the person would recall in calmer conditions. A blank under pressure is consistent with an honest, anxious witness as much as with anyone else. Stress and testimony looks more closely at how pressure interacts with memory.

Is "I don't know" a better answer than a guess?

Frequently, yes. Research on how people regulate their own memory reports shows that when people are free to volunteer or withhold an answer, they can substantially improve the accuracy of what they do report. Koriat and Goldsmith's work on the monitoring and control of memory describes this as a strategic trade-off: withhold the answers you are unsure of, and the answers you give become more reliable.

This reframes "I don't remember" as something closer to quality control than evasion. A witness who declines to guess is protecting the record from error. The problem is that interviewers do not always accept it. When a "don't know" is rejected and the same ground is pushed with more pressing questions, people — children especially — may start answering anyway, which is where unreliable detail creeps in.

What goes wrong when you push past "I don't remember"?

Pressing someone to produce an answer they do not have is one of the ways investigations generate false detail. If the social pressure to say something is strong enough, people can fill the gap with a plausible reconstruction, or absorb a suggestion embedded in the question itself, and then experience the result as a memory. Elizabeth Loftus's research on the misinformation effect shows how readily post-event suggestion can be incorporated into an account.

Taken to its extreme, relentless pressure contributes to false confessions: people can come to report, and sometimes to believe, things that did not happen. Why innocent people sometimes confess traces how that occurs. The safe practice is to accept the gap, change the retrieval cue, and come back to it, rather than treating "I don't remember" as a wall to batter down.

Can "I don't remember" ever be strategic?

Yes. People do sometimes use the phrase to avoid committing to a detail they would rather not confirm, to stall, or to leave themselves room to manoeuvre. That is a real possibility, and it is why the phrase is worth noticing.

But the crucial point is that the genuine and the strategic versions are indistinguishable from the words alone. There is no tell that reliably separates "I honestly cannot recall" from "I would prefer not to say." Treating the phrase as proof of concealment is the same error that runs through lie detection generally, where people average only about 54% accuracy. If a claimed memory gap matters, what resolves it is independent evidence or a better-designed question — not the assumption that the gap is a lie. See why there is no single reliable sign of deception, and the related question of why people avoid answering.

How does a memory gap work inside an interrogation game?

Murder Book: The Art of Interrogation turns these gaps into something the player has to interpret rather than punish. It is an upcoming narrative interrogation game from Gungrounds, a studio in Zagreb, Croatia, inspired by real police footage and psychological interviews. Development screenshots show a casebook that includes behaviours such as Non-Response and Stonewalling alongside Stress Response — categories, not conclusions.

When a suspect says they do not remember, the game lets you decide how to read it and how to respond: ease the pressure and try a different angle, present something concrete that might jog or test the account, or note the gap and move on, all while keeping several explanations of the case in play. A blank might be honest forgetting, the strain of the room, or avoidance, and the design keeps that uncertainty alive rather than resolving it for you. Cases run in focused 5–10-minute sessions. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.

Sources