Classification & Assessment in Psychiatry · Paper A
Memory falsification (confabulation, paramnesia)
Several distinct memory disturbances can look similar at first glance, but each has a specific cause and a specific clinical meaning. Two patients might both 'remember' an event that never happened, e.g. one believes it with complete delusionalRelating to a fixed false belief held with total conviction despite evidence against it.Read more → conviction while the other simply filled a gap unknowingly and is not fully sure. In practice, that difference in certainty and awareness is exactly what separates these phenomena from each other.
ConfabulationFilling a gap in memory with a made-up answer the person genuinely believes is true, not a deliberate lie.Read more → is the unintentional filling of memory gaps with fabricated or misplaced material, without awareness that it is happening, classic in Korsakoff's syndrome, and often inconsistent from one retelling to the next. DelusionalRelating to a fixed false belief held with total conviction despite evidence against it.Read more → memory is a delusionA fixed false belief, held with total conviction, that is not shared by the person's culture and is not changed by evidence against it.Read more → experienced as if it were a recollection, held with full delusional conviction. Retrospective falsification is the unintentional distortion of a true memory by the patient's current mood. Déjà vu and jamais vu are false feelings of familiarity or unfamiliarity, while a false memory is a recollection of an event that did not occur at all. Take a depressed patient recalling a childhood event as far more painful than it actually was at the time: that mood-coloured shift is retrospective falsification, not confabulation or a false memory. Clinically, the patient's current mood state is itself a clue to which memory phenomenon is occurring.
A fresh doctor hearing a patient confidently describe an event that never happened might assume the patient is simply lying. Memory falsification (a group of conditions where recollection goes wrong in a specific, patterned way) is not lying at all: the underlying mechanism, the patient's awareness of the error, and the clinical implication differ sharply between the different phenomena. Getting the label right changes how you interpret the history, and sometimes points straight at an underlying diagnosis, e.g. a treatable vitamin deficiency.
Memory falsification covers several distinct phenomena where a patient's recollection of the past is inaccurate: (unintentionally producing false memories to fill a gap in real memory), false memory (a full recollection of an event that never occurred), déjà vu and jamais vu (false feelings of familiarity or unfamiliarity), pseudologia fantastica (fluent, deliberate fabrication with no obvious gain), and retrospective falsification (unconsciously reshaping a real memory to fit the patient's current mood).
Déjà vu and jamais vu are usually normal, transient experiences that most people have felt at some point, and on their own they are not evidence of psychiatric or neurological disease. They only become clinically significant when frequent, prolonged, or accompanied by other features: for example, epilepsy can present with recurrent déjà vu as part of a seizure aura, which is a genuine investigation-worthy red flag rather than a normal quirk of memory.
Two questions separate these phenomena from each other: is the patient AWARE (consciously knows) that something might be wrong with their memory, and how CONFIDENT (certain) are they in the false content? is unaware and often inconsistent: the patient does not know they are filling a gap and may give a different fabricated account each time they are asked. memory, by contrast, is held with full delusional conviction: fixed, unshakeable, and consistent on repeat questioning. Retrospective falsification sits in between, distorting a real memory to match the patient's current mood without the patient noticing the distortion.
Take a patient with Korsakoff's syndrome (a memory disorder caused by thiamine, vitamin B1, deficiency, classically from chronic alcohol use). Ask what they did yesterday and they answer fluently and confidently, describing a plausible but entirely fabricated day. Ask again an hour later and the details have changed completely. This is : unintentional, inconsistent, and without any awareness that a gap is being filled. The mechanism is thought to involve damage to the mammillary bodies and disrupting the encoding of new memories, so the brain fills the resulting gaps automatically rather than leaving them blank.
A patient with a psychotic illness may report, with complete conviction, a "memory" of an event that could not have happened: for example, being fully conscious and present at their own birth. This is memory: a experienced in the form of a recollection, not a distortion of a real event. Retrospective falsification is different again: a depressed patient recalling a childhood event as far more painful than it genuinely was at the time is unconsciously reshaping a real memory to match their current low mood. The event genuinely happened; only the emotional colouring has shifted.
is not deliberate lying: the patient has no intent to deceive and genuinely believes what they are saying in the moment they say it. Pseudologia fantastica sits at the opposite end of that spectrum: fluent, elaborate, and knowingly fabricated, often to impress others or avoid consequences. False memory is a recollection of an event that did not occur at all, whereas retrospective falsification distorts the emotional tone of an event that did occur. Déjà vu is a false sense of familiarity with something new, whereas jamais vu is a false sense of unfamiliarity with something well known.
There is no single test for "memory falsification" as a category: each phenomenon is identified from the pattern of the history plus, where relevant, targeted investigation. Re-interviewing is the single most useful diagnostic manoeuvre: ask the same question twice, ideally on separate occasions, and see whether the account is inconsistent (pointing to ) or rigidly fixed (pointing to memory). Collateral history from a relative helps confirm whether the described event genuinely happened at all. Where Korsakoff's syndrome is suspected, check for the other features of the Wernicke-Korsakoff spectrum ( (eye-movement paralysis), (unsteady gait), and confusion) and send bloods including thiamine level and liver function. Frequent or stereotyped déjà vu, especially with other seizure features, warrants an EEG and consideration of epilepsy rather than being dismissed as normal.
Recognising should prompt urgent investigation for thiamine deficiency: treat with parenteral thiamine (e.g. IV Pabrinex) before giving glucose, since glucose without thiamine can precipitate or worsen : this sequencing is one of the few genuine psychiatric emergencies hiding inside a "memory" topic. Longer-term management of established Korsakoff's syndrome is largely supportive (structured routine, cognitive rehabilitation, and alcohol-relapse prevention where relevant) a pragmatic, Maudsley-style approach rather than reliance on a single licensed drug. memory occurring as part of a psychotic illness is managed the way psychosis is generally managed, with antipsychotic medication as first-line treatment and follow-up under a NICE-informed care pathway; the false memory itself is not targeted in isolation, because treating the underlying psychosis is what resolves it.
The classic trap is calling "lying": the decisive word is UNAWARE, since the patient genuinely does not know the memory is false. A second trap is mixing up déjà vu (a new situation feels familiar) with jamais vu (a familiar situation feels new) under exam time pressure. A third trap is assuming a single -memory-type experience automatically means , rather than weighing it alongside the whole clinical picture and screening for other features first.
Ask whether the patient knows something is wrong (awareness) and how sure they are of the false content (conviction): that pair of questions sorts , memory, and everything in between.