You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
Where the multi-store and working memory models describe memory as a system of stores, Frederic Bartlett (1932) offered a radically different picture in his book Remembering: memory is not a faithful recording that we simply replay, but an active reconstruction shaped by what we already know and expect. On this view, remembering is more like rebuilding an event from fragments than retrieving a stored copy, and the rebuilding is guided by mental frameworks Bartlett called schemas. Because reconstruction is shaped by prior knowledge, memory is systematically prone to distortion — a claim with enormous consequences for the reliability of eyewitness testimony. Bartlett's reconstructive theory is a cornerstone of the Edexcel cognitive topic and the historical root of the modern understanding that memory is unreliable in predictable, patterned ways.
Key Definition: Reconstructive memory is the idea that remembering is an active process in which we rebuild a memory using fragments of the original experience combined with our existing knowledge, expectations and beliefs (schemas), rather than replaying an exact stored copy.
This lesson addresses the Edexcel 9PS0 — Paper 1, Topic 2: Cognitive Psychology content on reconstructive memory and schema theory: Bartlett's (1932) theory that memory is an active reconstruction; the role of schemas in encoding and recall; the concepts of effort after meaning, confabulation and rationalisation; and the "War of the Ghosts" studies that supported the theory. In assessment-objective terms, you should be able to describe reconstructive memory and schema theory and outline Bartlett's method and findings (AO1), apply them to everyday memory distortions and scenario examples (AO2), and evaluate the theory through its research support, methodological weaknesses and real-world applications (AO3).
Connects to…
Before Bartlett, the dominant approach (associated with Ebbinghaus's nonsense-syllable experiments) treated memory as the passive retention of decontextualised material. Bartlett objected that this stripped remembering of the very thing that makes it human: meaning. In real life we remember meaningful stories, events and experiences, and we make sense of them using what we already know. His key insight was that this "making sense" does not stop at encoding — it continues to operate at recall, so that what we reconstruct is shaped as much by our expectations as by the original event.
Key Definition: A schema is a mental framework of organised prior knowledge about a person, object, situation or event, built up from experience, which guides how we interpret new information and fill gaps when we remember.
Schemas do useful work: they let us interpret ambiguous situations quickly, predict what is likely to happen, and reconstruct plausible detail when our memory is incomplete. But the same efficiency is the source of error. When an event does not fit our schemas, we tend — usually without awareness — to change the memory so that it does. The result is memory that is coherent and meaningful but not necessarily accurate.
Bartlett distinguished several kinds of schema that shape what we remember. Scripts are schemas for sequences of events — the familiar order of actions in a restaurant, a lesson or a doctor's appointment — and they lead us to "remember" script-consistent steps that may not actually have occurred. Role schemas organise our expectations about how people in particular social roles behave, so we may misremember a person's actions to fit our stereotype of their role. Self-schemas organise beliefs about ourselves and can bias autobiographical memory towards a flattering or consistent self-image. Because these frameworks are built up over a lifetime of experience and are largely automatic, we are rarely aware of the moment at which a schema reshapes a memory — which is exactly why schema-driven distortion feels, to the person experiencing it, like ordinary accurate remembering.
Crucially, Bartlett argued that a schema operates at two points in the memory process. At encoding, it determines what we notice and how we interpret an event, so that schema-inconsistent information may never be properly registered in the first place. At retrieval, it guides the active reconstruction, supplying expected detail to fill gaps and smoothing away anything that does not fit. This "double action" is important for the exam because it explains why distortion is so pervasive and so hard to correct: even a witness who tries hard to be accurate cannot recover detail that a schema prevented them from encoding, nor easily suppress the plausible-but-false detail that a schema generates during recall.
Key Definition: Effort after meaning is the tendency to focus on the overall meaning or "gist" of information and to actively try to make it fit our existing schemas, at the expense of the exact surface details.
Bartlett coined this phrase to capture what participants do when confronted with unfamiliar material: they strive to make it meaningful. In doing so they retain the general sense of an event while reshaping specifics to match familiar patterns. Effort after meaning explains why we remember the "point" of a story accurately but misremember its details, and why unfamiliar or culturally alien elements are so readily lost or transformed.
To test his ideas, Bartlett needed material that would not fit his English participants' existing schemas, so that any schema-driven distortion would be visible. He chose "The War of the Ghosts", a Native American folk tale that was unfamiliar in style, structure and content to his early-twentieth-century Cambridge participants. Its supernatural elements, unusual causal logic and unfamiliar cultural references (canoes, hunting seals, spirits) gave participants little schema support, forcing them to "make sense" of it in their own terms.
Bartlett used two related techniques.
| Technique | Procedure | What it reveals |
|---|---|---|
| Repeated reproduction | The same participant recalled the story on several occasions, separated by increasing intervals (minutes, then days, weeks, months or years) | How an individual's memory of the same material changes and distorts over time |
| Serial reproduction | The story was passed along a chain of people — the first person's recall was given to the second to recall, and so on | How information becomes progressively distorted as it is transmitted (a laboratory analogue of rumour) |
Across both techniques, participants' reproductions were not accurate copies. They changed in strikingly consistent ways as they were made to fit familiar schemas.
Two of these processes are named explicitly on the specification and deserve careful definition.
Key Definition: Rationalisation is the alteration of unfamiliar or illogical-seeming material so that it fits the person's existing schemas and appears more coherent, logical or acceptable to them.
Key Definition: Confabulation is the (unintentional) production of memory details that were not part of the original event — plausible-seeming "fillers" that the mind generates from schema-based expectation to complete a fragmentary memory. Crucially, the person is not lying: they genuinely believe the confabulated detail is a real memory.
Rationalisation is why the "ghost" elements were smoothed into something more comprehensible; confabulation is why participants confidently reported details the story never contained. Both flow directly from the reconstructive nature of memory: when the reconstruction meets a gap or an oddity, schemas step in to patch it, and the patched version is experienced as genuine memory.
The serial-reproduction technique is worth dwelling on because it doubles as a laboratory model of how information distorts as it spreads through a community — the psychology of rumour and gossip. As the story passed from person to person, each reproduction became the "original" for the next participant, so distortions compounded: an unfamiliar detail assimilated by the third person was then further simplified by the fourth, and so on. By the end of a chain the reproduction often bore little resemblance to Bartlett's original, having been progressively conventionalised into something that made sense within the shared cultural schemas of the group. This cumulative drift towards the familiar and the plausible is precisely the pattern later observed in real rumour transmission, and it illustrates a key claim of reconstructive theory: distortion is not random noise but a directional process that pulls memory towards the expected. The same mechanism has obvious implications for the reliability of second-hand testimony, for how accounts of an event shift as witnesses talk to one another, and for how historical or anecdotal accounts become conventionalised over repeated retelling.
A striking regularity across Bartlett's reproductions is that the theme or overall meaning of the story was comparatively well preserved even as specific details were transformed or lost. This dissociation — durable gist, fragile detail — is itself strong evidence for reconstructive theory. If memory were a faithful recording, the gist and the detail should decay together; instead, participants held on to the meaningful "shape" of the story (a journey, a conflict, a death) and reconstructed the particulars around it using their schemas. This is why, in everyday life, we can confidently recount the point of a conversation or film long after we have forgotten the exact words or scenes — and why we may unknowingly invent plausible specifics that fit the remembered gist. The gist/detail asymmetry is a useful discriminator in the exam because it shows that reconstruction is not simply "forgetting more over time" but a qualitatively different, schema-guided process.
The diagram below summarises how a schema shapes memory at both encoding and retrieval.
flowchart TD
EVENT["Original event or story"] --> ENC["Encoding<br/>(interpreted through existing schemas)"]
ENC --> STORE["Fragmentary trace stored"]
SCHEMA["Existing schemas<br/>(prior knowledge & expectations)"] --> ENC
SCHEMA --> RECON["Retrieval = active reconstruction"]
STORE --> RECON
RECON --> OUT["Recalled memory<br/>(coherent & meaningful,<br/>but distorted)"]
RECON -.->|gaps filled| CONF["Confabulation"]
RECON -.->|oddities smoothed| RAT["Rationalisation"]
RECON -.->|unfamiliar omitted| LEV["Levelling"]
Bartlett's insight was later developed into a more formal schema theory and supported by better-controlled studies. A classic demonstration is the "office" study (Brewer and Treyens, 1981), in which participants waited in an academic office and were later asked to recall its contents. They tended to "remember" schema-consistent objects that were not present (such as books) and to omit schema-inconsistent objects that were present (such as a skull), showing that recall is powerfully shaped by expectation. Studies of this kind matter because they reproduce Bartlett's core finding — schema-driven distortion — under far tighter experimental control than his original work allowed, addressing one of its main weaknesses (see Evaluation).
Schema theory also distinguishes how schemas cause error:
| Mechanism | Effect on memory | Example |
|---|---|---|
| Omission (levelling) | Schema-inconsistent detail is dropped | Forgetting the skull in an office |
| Assimilation | Detail is shifted to fit the schema | "Canoe" recalled as "boat" |
| Sharpening | Schema-consistent detail is elaborated or emphasised | Adding books that were never there |
| Rationalisation | Illogical material is made coherent | Smoothing away the supernatural |
A major strength of reconstructive memory theory is its explanatory power for the unreliability of eyewitness testimony. If memory is reconstructed through schemas rather than replayed, then witnesses should be systematically vulnerable to distortion — filling gaps with schema-consistent (but false) detail and reshaping events to fit expectations. This is exactly what later eyewitness research found: leading questions and post-event information alter witness recall, and stereotypes shape what witnesses "remember". This matters because it means Bartlett's theory is not a laboratory curiosity but the theoretical foundation for a body of research with direct consequences for the justice system, where mistaken identification has contributed to wrongful convictions. The implication is that reconstructive memory has high applied validity: it predicts and explains a real, high-stakes phenomenon, which is a strong marker of a theory's worth.
A further strength is that Bartlett studied memory for meaningful material — stories and everyday scenes — rather than the nonsense syllables favoured by Ebbinghaus. This matters because everyday remembering overwhelmingly involves meaningful information, so a theory built on meaningful material is more likely to capture how memory actually works outside the laboratory. The implication is that reconstructive memory has greater ecological validity than the store models' reliance on decontextualised word lists and trigrams — it explains the kind of remembering that matters in real life, such as recalling conversations, events and experiences, where schema-driven distortion is readily observable.
Set against these strengths, Bartlett's original studies were methodologically weak. His instructions to participants were vague — he did not always tell them to reproduce the story as accurately as possible — so some distortion may reflect participants guessing or embellishing rather than genuine memory error. The scoring of reproductions was subjective, relying on Bartlett's own interpretation of what counted as a "distortion", with no standardised measure. This matters because a lack of standardisation and objective scoring threatens the internal validity and replicability of his findings — we cannot be sure the distortions were memory failures rather than artefacts of loose procedure. The implication is that, although Bartlett's ideas have been strongly supported, his original evidence is not by itself compelling, and the theory rests more securely on the later controlled studies (such as Brewer and Treyens) that reproduced the schema effect under proper control.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.