You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
The multi-store model treated long-term memory (LTM) as a single, undifferentiated store, but research has shown that LTM is composed of several distinct types, each with different characteristics, conscious status and neural substrate. Endel Tulving (1972, 1985) proposed the most influential distinction, dividing LTM into episodic, semantic and procedural memory. This account both deepens our understanding of memory and exposes a key weakness of the MSM. For Edexcel cognitive psychology, Tulving's types of LTM are the counterpart to the working memory model: just as the WMM fractionated the MSM's unitary short-term store, Tulving fractionated its unitary long-term store.
Key Definition: Long-term memory (LTM) is the relatively permanent memory store of potentially unlimited capacity and duration. Tulving argued it is not unitary but comprises episodic, semantic and procedural memory.
This lesson addresses the Edexcel 9PS0 — Paper 1, Topic 2: Cognitive Psychology content on the types of long-term memory: episodic, semantic and procedural. You are required to know the defining features of each type (its content, whether it is time-stamped, and whether it is consciously [explicitly] or unconsciously [implicitly] retrieved) and the evidence that distinguishes them — principally the case studies of HM and Clive Wearing and supporting brain-imaging. In assessment-objective terms, you should be able to describe and distinguish the three types (AO1), apply them to examples of remembering (AO2), and evaluate the distinction through clinical and biological evidence and the debate about whether the types are truly separate (AO3).
Connects to…
flowchart TD
LTM["Long-Term Memory"]
LTM --> EXP["Explicit / Declarative<br/>(conscious recall)"]
LTM --> IMP["Implicit / Non-declarative<br/>(unconscious)"]
EXP --> EPI["Episodic<br/>personal events,<br/>time-stamped"]
EXP --> SEM["Semantic<br/>facts & meanings,<br/>not time-stamped"]
IMP --> PRO["Procedural<br/>skills & actions,<br/>hard to verbalise"]
Key Definition: Episodic memory is the ability to recall personal events and experiences from one's own life. These memories are "time-stamped" — they carry information about when and where the event happened and the emotions attached to it.
| Feature | Detail |
|---|---|
| Content | Personal events and experiences (your first day at school, a holiday, a conversation) |
| Time-stamped | Includes a record of when the event occurred and its temporal/sequential context |
| Contextual | Includes where it happened and who was present |
| Conscious recall | Explicit — requires conscious, deliberate effort; Tulving spoke of "mental time travel" back into the experience |
Tulving emphasised that episodic recall involves a distinctive autonoetic (self-knowing) awareness — a sense of re-living the event — that distinguishes it from merely knowing a fact.
Key Definition: Semantic memory is our store of general knowledge about the world — facts, concepts and the meanings of words — that is not tied to a specific time or place of learning.
| Feature | Detail |
|---|---|
| Content | Facts, concepts, meanings and general knowledge ("Paris is the capital of France") |
| Not time-stamped | You usually cannot recall when or where you learned a given fact |
| Conscious recall | Explicit — consciously retrieved, but without re-living a personal experience |
| Shared | Much semantic knowledge is held in common across many people |
Tulving argued that episodic and semantic memory, while both explicit, are functionally distinct, and that episodic memory may have "grown out of" semantic memory developmentally — children acquire factual knowledge before they reliably form autobiographical episodic memories.
Key Definition: Procedural memory is memory for how to do things — skilled, learned actions. It is implicit (retrieved without conscious awareness) and difficult to put into words.
| Feature | Detail |
|---|---|
| Content | Motor and cognitive skills (riding a bicycle, typing, playing the piano, tying shoelaces) |
| Implicit | Operates without conscious awareness; you do not consciously recall each step |
| Hard to verbalise | Difficult to explain in words — try describing exactly how you balance on a bicycle |
| Durable & automatic | Resistant to forgetting and runs automatically with little cognitive effort once over-learned |
| Feature | Episodic | Semantic | Procedural |
|---|---|---|---|
| Content | Personal events | Facts and knowledge | Skills and actions |
| Time-stamped? | Yes | No | No |
| Conscious? | Yes (explicit) | Yes (explicit) | No (implicit) |
| Easily verbalised? | Yes | Yes | No |
| Example | Your 18th birthday | Knowing 2 + 2 = 4 | Riding a bicycle |
| Main brain areas | Hippocampus & prefrontal cortex | Temporal lobe (anterior/lateral) | Cerebellum, basal ganglia, motor cortex |
The strongest evidence comes from patients with brain damage who show dissociations — one type of LTM is impaired while others remain intact.
HM (Henry Molaison): Following bilateral surgical removal of much of his hippocampus (and nearby medial temporal tissue) to treat severe epilepsy, HM developed dense anterograde amnesia. He could not form new episodic memories — he failed to recognise people he had met repeatedly after the surgery — and the acquisition of new semantic facts was also grossly impaired. Yet his procedural memory was preserved: on the mirror-drawing task he improved steadily across sessions, demonstrating skill learning, even though on each occasion he had no episodic memory of ever having done the task before. This is a textbook dissociation: implicit/procedural memory intact, explicit/episodic memory devastated.
Clive Wearing: A professional musician who contracted herpes simplex encephalitis, which destroyed his hippocampus and surrounding temporal lobes. He has almost no episodic memory — he experiences each moment as though he has just regained consciousness, repeatedly writing in his diary that he has "now" woken for the first time — and his semantic memory is patchy (he cannot recall facts about much of his own life). Yet his procedural memory is strikingly intact: he can still sight-read, play the piano and conduct a choir to a high standard, despite having no episodic memory of his musical training. Like HM, his profile shows a clean separation between a shattered explicit system and a preserved implicit one.
A further important pattern is the opposite dissociation. In semantic dementia (a form of frontotemporal degeneration affecting the anterior temporal lobes), patients progressively lose semantic knowledge — they cannot name common objects or understand familiar words — while episodic and procedural memory are comparatively preserved. Set alongside HM, this approaches a double dissociation: if semantic memory can be lost while episodic is spared, and (in other patients) episodic lost while semantic is spared, the two must depend on separable systems.
| Type | Associated brain areas |
|---|---|
| Episodic | Hippocampus and prefrontal cortex (Tulving's PET work implicated the right prefrontal cortex in episodic retrieval and the left in encoding) |
| Semantic | Temporal lobe, especially anterior and lateral regions |
| Procedural | Cerebellum, basal ganglia and motor cortex |
Tulving and colleagues' PET studies of healthy volunteers found different patterns of cortical blood flow when participants retrieved episodic versus semantic information, providing biological support that the types are genuinely distinct rather than merely convenient labels — converging with the patient evidence.
An AO2 scenario question typically asks you to identify which type of LTM is involved in a described act of remembering, and the discriminating questions are always the same: is the memory time-stamped (tied to a specific personal occasion), and is it conscious/explicit or unconscious/implicit? Consider a musician preparing for a concert. Recalling the piece of music itself — the notes, the structure, what the composer intended — draws on semantic memory: it is factual knowledge not tied to when it was learned. Recalling the particular rehearsal last Tuesday when the conductor lost their temper draws on episodic memory: it is a personal, time-stamped event recalled with a sense of re-living. Actually playing the instrument — the finger movements, the coordination, the automatic execution of scales — draws on procedural memory: it is implicit, hard to verbalise, and runs without conscious attention to each step. The same everyday activity therefore engages all three systems, and being able to allocate each element to the correct type, justified by the time-stamp and conscious-status criteria, is precisely the skill a scenario question rewards. Clive Wearing makes the point vivid: his procedural musical skill survives intact while his episodic memory of ever having learned or performed it is gone.
Beyond the three-way division, Tulving added two refinements that mark out top-band understanding.
First, he linked each type to a different kind of conscious awareness.
| Type of memory | Kind of awareness | What it feels like |
|---|---|---|
| Episodic | Autonoetic ("self-knowing") | Re-living — a sense of mentally travelling back and experiencing the event again |
| Semantic | Noetic ("knowing") | Knowing a fact is true without re-living any experience of learning it |
| Procedural | Anoetic ("non-knowing") | No conscious awareness at all — the skill simply runs |
This matters for the exam because it explains why episodic memory is special: it is not merely memory for events but memory accompanied by a distinctive first-person, time-travelling awareness that semantic and procedural memory lack. A candidate who can articulate the autonoetic/noetic distinction shows understanding well beyond a list of three types.
Second, Tulving argued for a developmental and dependency relationship between the types. Procedural memory develops earliest in infancy, semantic memory next, and episodic memory last — children acquire facts before they reliably form autobiographical, time-stamped memories — and the later systems are thought to depend on, and to have evolved out of, the earlier ones. This developmental ordering is mirrored, in reverse, by patterns of breakdown: in many amnesias and dementias the most recently developed system (episodic) is the most vulnerable, while the oldest (procedural) is the most robust — exactly the profile seen in HM and Clive Wearing. This convergence of developmental and clinical evidence on the same ordering is a powerful additional argument that the types are genuinely distinct systems rather than arbitrary divisions of one store.
It is also worth noting, for precision, that the implicit / non-declarative category to which procedural memory belongs is wider than motor skills alone. It also includes priming (prior exposure to a stimulus speeding later processing of it, without conscious recollection), classical conditioning (learned emotional and reflex associations) and habits. Amnesic patients such as HM show intact priming and conditioning despite devastated episodic memory, which extends the dissociation evidence beyond procedural skills and further supports a fundamental explicit/implicit divide in long-term memory.
The significance of Tulving's work is best appreciated in relation to the multi-store model. The MSM treated long-term memory as a single, undifferentiated store into which information flowed via rehearsal and from which it was later retrieved. Tulving's evidence makes that picture untenable on three counts.
| MSM claim | What the types-of-LTM evidence shows |
|---|---|
| LTM is one store | LTM is at least three functionally distinct systems with different properties |
| LTM is uniform in how it operates | Some LTM is conscious/explicit (episodic, semantic), some unconscious/implicit (procedural) |
| Damage to LTM should impair all long-term memory | Damage can selectively impair one type while sparing others (HM, Clive Wearing, semantic dementia) |
Just as the working memory model dismantled the MSM's claim that short-term memory is a single store, Tulving's types dismantle its claim that long-term memory is a single store. Taken together, the two critiques show that the MSM's greatest weakness was its treatment of each store as unitary — and that a more accurate model of memory must be multi-component throughout. Recognising this thread — that the WMM and the types of LTM are parallel critiques attacking the same assumption at different stages of the memory system — is exactly the kind of synoptic understanding that distinguishes the strongest answers.
A major strength is the support from case studies of amnesia. HM and Clive Wearing both show intact procedural memory alongside devastated episodic memory, a dissociation that is extremely difficult to explain if LTM is a single store — damage to one store should impair all long-term memory, not spare skills while abolishing personal recollection. This matters because such selective sparing is exactly what Tulving's model predicts: different types depend on different neural systems, so damage to one can leave another untouched. The implication is that the distinction between (at least) explicit and implicit/procedural memory is not merely a theoretical convenience but corresponds to a real division in how the brain stores information — a conclusion strengthened by the opposite dissociation seen in semantic dementia, which together approaches the double dissociation that is the strongest single argument for two independent systems.
The model gains further support because biological evidence converges with the clinical evidence. Tulving's PET work and subsequent fMRI studies show that retrieving episodic versus semantic information engages different brain regions in healthy participants, addressing the worry that the patient evidence might be an artefact of abnormal, damaged brains. This matters because it means two independent methodologies — neuropsychological case studies and functional brain imaging — point to the same conclusion. The implication is that the types are dissociable in the intact as well as the injured brain, which makes the distinction considerably more robust than if it rested on lesion cases alone. (A caveat, considered below, is that neuroimaging is correlational, and mapping a function onto a region does not by itself prove the region is necessary for it.)
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.