Read this topic

Behavioural Science & Sociocultural Psychiatry · Paper A

Episodic memory

The idea

Episodic memory is your personal diary — a record of specific lived moments, each stamped with a time and a place — kept in a completely different part of the brain from your encyclopedia of general facts. A patient can have the encyclopedia open and fluent while the diary has stopped being written altogether, which is exactly why fluent conversation is not proof that memory is fine.

How it works

Episodic memory is memory for specific, personally experienced events — the 'what, where, when' of a moment — retrieved with autonoetic awareness, the felt sense of mentally reliving it, not just knowing a fact.
It depends chiefly on the hippocampus and the surrounding medial temporal lobe (entorhinal, perirhinal, parahippocampal cortex), which bind the separate strands of an experience into one retrievable trace.
It is Tulving's counterpart to semantic memory (decontextualised facts and word meanings, leaning on anterolateral/polar temporal cortex) within the broader declarative (explicit) memory system.
It is dissociable from non-declarative systems that survive hippocampal damage: procedural memory (skill learning, basal ganglia/cerebellum) and implicit priming — for example, the classic amnesic patient who improves daily at a motor task with no memory of ever practising it.
It is dissociable from working memory and short-term/immediate memory (a brief online holding span, prefrontal-parietal, that does not require the hippocampus) and from prospective memory (remembering to act in the future, frontally mediated).
Frontal-lobe damage does not touch the trace itself but can impair the STRATEGY used to search for it — poor free recall that normalises with recognition or cueing is a retrieval-strategy problem, not a hippocampal one; in practice this is the single cheapest bedside test in the whole topic.
It is the newest and most fragile long-term memory system, and therefore usually the first to fail in hippocampal pathology — most notably Alzheimer's disease, where impaired delayed recall that does not improve with cueing is an early, sensitive marker. The dense anterograde amnesia following bilateral medial temporal lobe damage, the historical HM case, illustrates this same fragility directly.

How to spot it

A vignette describing fluent speech and intact general knowledge alongside an inability to recall a recent, specific event is describing episodic memory impairment, not normal cognition.
A vignette describing improving performance on a practised skill in a patient with no memory of ever practising it is showing procedural memory spared alongside episodic memory lost.
A vignette describing normal immediate digit span alongside failed recall after a delay is showing working memory intact and episodic memory impaired — a dissociation, not a contradiction.
A vignette describing normal recall on formal testing alongside real-world failure to carry out a planned action (missed medication, missed appointment) is describing a prospective memory deficit, not an episodic one.
A vignette describing poor free recall that becomes normal on a recognition task is describing a frontal retrieval-strategy deficit, not hippocampal amnesia.

Tell it apart

Episodic memory (specific lived events, autonoetic) vs semantic memory (decontextualised facts, no reliving) — both declarative, but only episodic memory carries the felt sense of the moment.
Episodic memory vs procedural memory — procedural (skill) learning is non-declarative and classically SPARED when episodic recall is devastated, unlike either half of declarative memory.
Episodic memory vs working memory/short-term memory — a normal immediate digit span does not mean intact DELAYED recall; the two dissociate because only episodic recall needs the hippocampus.
Episodic memory vs prospective memory — prospective memory is remembering to DO something later, not recalling something that already happened; a patient can have one intact and the other impaired.
Episodic memory loss (hippocampal) vs a frontal retrieval-strategy deficit — in hippocampal loss, recognition is impaired ALONGSIDE free recall because the trace itself is gone; in a frontal deficit, recognition or cueing rescues performance because the trace survived and only the search strategy failed.
Episodic memory vs autobiographical memory — autobiographical memory is the broader self-referential life narrative, built FROM episodic detail plus semantic self-knowledge, rather than a synonym for the single-event system.

Watch for

⚠️ Reading fluent conversation and intact general knowledge as proof that memory is fine — that pattern is the textbook picture of episodic memory loss with semantic memory spared, not reassurance.

⚠️ Assuming intact skill learning (procedural memory) or a behavioural change from priming (implicit memory) means episodic memory is also intact — both survive independently of the hippocampus.

⚠️ Calling poor free recall a hippocampal memory disorder without checking whether recognition or cueing rescues it — that check is exactly what separates a frontal retrieval-strategy deficit from genuine amnesia.

⚠️ Treating working memory (a normal digit span) as evidence that episodic memory must also be intact — they dissociate, and the exam relies on candidates conflating them.

⚠️ Confusing prospective memory (forgetting to DO something later) with episodic memory (forgetting something that already happened) — a patient can have a normal delayed-recall score and still fail every prospective-memory task.

⚠️ Missing that a genuinely inattentive or delirious patient can 'fail' an episodic memory test without any hippocampal disease at all — sustained attention has to be checked before the memory result is trusted.

Use it again

First askis the stem describing a FACT (semantic), a SKILL (procedural), an ONLINE HOLDING task (working memory), a FUTURE intention (prospective), or a SPECIFIC PAST EVENT (episodic)? Only the last is episodic memory.
If free recall is poor, check whether the stem also tells you what happens on RECOGNITION or with cueing before naming hippocampal amnesia — if recognition rescues it, think frontal retrieval strategy instead.
If general knowledge or fluent speech is explicitly described as normal, do not treat that as ruling out a significant memory deficit — check specifically whether recent, specific-event recall is described as impaired.
If attention or arousal is fluctuating (delirium-type picture), do not conclude episodic memory is impaired until inattention itself has been excluded as the cause of poor performance.
If both storage (recognition) AND strategy (free recall) fail together, that combination points to the hippocampus itself, not the frontal lobe.

In detail

Hook — the diary versus the encyclopedia

Picture two filing cabinets in your head. One is an encyclopedia: it holds facts — the capital of France, how a kettle works, what the word "kind" means — with no note of when or where you learned any of it. The other is a diary: it holds moments — the taste of last year's birthday cake, the argument you had on Tuesday, the walk home in the rain — each one time-stamped, and each one you can step back into and feel again. Episodic memory is the diary, not the encyclopedia, and the two live in different parts of the brain. That single split is the reason a person can lose one cabinet almost completely while the other stays intact, and it is the reason this topic keeps appearing in psychiatry exams: which cabinet has failed tells you where in the brain to look.

Plain definition

Episodic memory is your memory for specific events you personally lived through — a particular thing that happened, at a particular time, in a particular place. The psychologist Endel Tulving named it in 1972 and gave it a precise test: does recalling it feel like knowing, or like remembering? Knowing that Paris is the capital of France requires no sense of the moment you learned it — that is semantic memory (-an-tik, meaning memory for general facts and word meanings, not tied to any one occasion). Remembering your first day at medical school comes with a felt sense of being back there, hearing the noise, seeing the room — that felt sense is called autonoetic awareness (AW-toh-noh-EH-tik, meaning "self-knowing," the specific subjective feeling of mentally travelling back in time), and it is the signature of episodic memory alone.

The critical caveat — a fluent, knowledgeable patient can still have a ruined diary

The single most important thing to fix before anything else: a patient who answers general-knowledge questions perfectly, speaks fluently, and seems entirely "with it" in conversation can still have a devastated episodic memory. Families and even junior clinicians read fluent conversation as proof that "nothing is really wrong," and miss that the diary — not the encyclopedia — is the part that failed. You will see this trap dressed up in a hundred different vignettes: a patient who knows facts, follows a conversation, and yet cannot tell you what happened five minutes ago. That pattern is not reassurance. It is the textbook picture of episodic memory loss.

First principles — why the brain keeps two separate stores at all

A fact like "Paris is the capital of France" is useful precisely because it does not depend on the one occasion you learned it — it needs to generalise across every future use, so the brain strips away the specific time and place and keeps only the content. A personal event is the opposite: its VALUE is in its — which doctor said what, on which day, before which decision — so the brain has to bind the who, what, where and when into one linked package rather than dissolving them into a general rule.

Binding several separate strands of one experience into a single retrievable trace is expensive, effortful work, and it is done chiefly by the (hip-oh-KAM-pus, a curved, seahorse-shaped structure buried in the medial on each side of the brain) together with the surrounding entorhinal, perirhinal and parahippocampal cortex. Semantic memory, once it is well-established, leans more on the anterolateral and polar temporal cortex and needs the hippocampus far less. That is why the two systems can come apart: damage the hippocampus and the diary stops being written, while the encyclopedia — built up over a lifetime and no longer dependent on that binding machinery — keeps working.

Structured walk-through — watching the two systems come apart on a real test

Consider a standard delayed-recall test. A patient is read a five-item word list — apple, penny, table, ocean, violin — and asked to repeat it back immediately. Someone with entirely normal short-term retention will manage all five right away, because immediate repetition barely touches the at all. Ten minutes later, after a distracting conversation, the examiner asks for the list again with no cues. A person with genuine hippocampal, episodic memory loss recalls perhaps zero or one of the five words freely — and, crucially, does no better when shown a mixed list of ten words (the original five plus five new distractors) and asked simply to recognise which ones were on the original list. Both free recall AND recognition fail together, because the trace itself was never properly stored.

Now compare a patient with a frontal-lobe retrieval-strategy problem instead of a hippocampal one. On free recall they also do poorly — perhaps one or two of the five — because an inefficient, disorganised search strategy fails to locate the memory. But shown the same ten-word recognition list, they suddenly pick out all five correctly. The trace was there all along; what failed was the STRATEGY for finding it, not the storage itself. That one extra step — offering recognition after a failed free recall — is the single cheapest bedside manoeuvre in this whole topic, because it tells you in thirty seconds whether you are looking at a hippocampal problem or a frontal one.

Look-alikes and differentials

Looks similar	How to tell it apart
Semantic memory	Decontextualised facts, no sense of reliving; often spared early when episodic recall of the day's events has already failed.
Procedural memory	Gradual skill learning (/); classically SPARED in dense hippocampal amnesia, unlike episodic recall.
Working memory	A few-second online holding span (prefrontal-parietal); normal digit span does not mean intact delayed recall.
Prospective memory	Remembering to DO something later, not to recall something past; frontally mediated, separately impairable.
Executive/frontal retrieval	Poor FREE recall that normalises with recognition or cueing — the trace survives, only the search strategy failed.

Diagnosis — how episodic memory impairment is actually established

There is no blood test for a damaged diary. Diagnosis rests on standardised delayed-recall testing set against age-matched norms: a word-list or short-story recall after a filled delay (as in the Addenbrooke's Cognitive Examination-III and the , both in routine UK cognitive-assessment use), with the recognition step above used specifically to separate a hippocampal storage failure from a frontal retrieval-strategy one. Structural MRI adds a biological anchor: hippocampal atrophy is one of the earliest and best-validated imaging markers of Alzheimer's disease, the commonest cause of a progressive episodic memory disorder.

Before any of this is read as dementia, reversible mimics have to be excluded — delirium (fluctuating, attention-driven, and this is where checking sustained attention actually matters, since a genuinely inattentive patient will "fail" a memory test without any hippocampal disease at all) and depression-related pseudodementia, where poor effort and motivation on testing, not lost storage, drive the low score.

Management — what changes once episodic memory impairment is confirmed, and whose guidance it rests on

Management follows the cause, since episodic memory loss is a pattern, not a diagnosis in itself. In the UK, NICE guideline NG97 governs the assessment and support pathway once a progressive cognitive disorder is suspected: it directs referral for a full neuropsychological or memory-clinic assessment, structural imaging, and — once Alzheimer's disease is confirmed — consideration of an acetylcholinesterase inhibitor, alongside practical support (written reminders, structured routines, safety planning) that compensates for the diary no longer being written reliably. Where the picture instead fits depression-related pseudodementia, treating the depressive episode is the first-line step, and the memory impairment should recover substantially as mood improves — a genuinely different, reversible plan from the neurodegenerative one, which is exactly why getting the pattern right at diagnosis matters for what happens next.

Exam pearl

Whenever a stem pairs FLUENT conversation and INTACT general knowledge with an inability to recall recent, specific events, the answer is episodic memory impairment, not "nothing significant." And whenever free recall is poor but recognition or cueing rescues it, think frontal retrieval strategy, not hippocampal amnesia — the cueing step is the detail the exam is testing you on, not decoration. Before you move on: could you explain to a friend why the same patient can ace a general-knowledge quiz and still fail to remember the conversation you just had with them?

In one line

Episodic memory is your personal diary of lived moments, written by the and read back with the feeling of reliving them — which is why it is usually the first thing to fail, and why a person can lose it completely while their encyclopedia of general facts stays open on the desk.