Schema, Reconstruction and False Memory

What a schema is, and where it acts

A schema is a mental framework built from previous experience that organises knowledge and guides the processing of new information. Schema theory makes one central claim: memory is not a recording that is played back, it is a reconstruction assembled from a partial trace plus expectation. The claim matters because it locates distortion in two different places.

• At encoding — schema-consistent material is attended to and stored, schema-inconsistent material is often filtered out. Bransford and Johnson (1972) gave participants an intentionally vague passage about sorting items into groups; those told the title before reading understood and recalled far more than those told it afterwards or not at all. Comprehension itself depended on having the right schema active at the moment of input.
• At retrieval — even a well-stored memory is rebuilt at the moment of recall, and gaps are filled with what should have been there. Brewer and Treyens (1981) left participants briefly in an academic office and then tested recall: many confidently reported books, which the room did not contain, because books belong to an office schema. Anderson and Pichert (1978) had participants read a story about a house from a burglar's or a home-buyer's perspective; when the perspective was switched before a second recall, participants produced details they had failed to report the first time — the information had been stored but was not accessible under the first schema.

Reconstruction and false memory

Bartlett (1932) gave English participants the Native American folk tale The War of the Ghosts and asked for repeated reproductions. Recall grew shorter, unfamiliar elements were rationalised into familiar ones, and the story drifted towards the participants' own cultural schema. Loftus and Pickrell (1995) went further and implanted a wholly false memory: participants were given four childhood narratives supplied by a relative, one of which — being lost in a shopping mall — had never happened, and about a quarter came to report remembering it, some adding details of their own. If a schema can generate detail that feels like memory, then vividness and confidence cannot be used as evidence of accuracy.