Your memory does not work like a video recorder. When you take in new information, you fill in gaps using what you already know (constructive memory), and every time you recall something you rebuild it on the fly (reconstructive memory). The mental frameworks you use to do this, called schemas and scripts, help you understand everyday events quickly but also introduce predictable errors, such as "remembering" things that were never there.
Difficulty: Intermediate | Prerequisites: Introductory cognitive psychology (encoding, storage, retrieval basics)
This topic moves beyond the laboratory basics of encoding and retrieval and asks how memory works in real life, where information is messy, incomplete, and wrapped in prior knowledge. It sits in the "everyday cognition" portion of PSY 310 and connects directly to later material on eyewitness testimony, false memories, and flashbulb memories (covered in Part 2). You should already be comfortable with the distinction between encoding and retrieval and with the idea that memory can fail at either stage.
Constructive processes of memory
The tendency to fill in missing details at the time of encoding, using prior knowledge and inference. In simple terms, your brain writes in the gaps of the story as you take it in, so what you store was never purely "what happened."
Constructive memory
The filling-in that happens at encoding. Think of it as your brain auto-completing a sentence before the speaker finishes.
Reconstructive memory
The filling-in that happens at retrieval. Every time you recall an event, you are partly rebuilding it from fragments and expectations, not playing back a recording.
Schema (plural: schemata)
A structured mental framework for representing knowledge about a concept, event, or object. It contains "slots" (the basic dimensions that describe the thing) and "fillers" (the specific features most likely to fill each slot, ordered by how typical they are). Think of it as a template your mind lays over new information to make sense of it quickly.
Slots
The basic dimensions or categories within a schema. In a Purchase schema, slots include Buyer, Seller, Item, and Compensation.
Fillers
The specific features that typically fill a schema's slots, ranked by how common they are. For the Buyer slot, "person" is a more typical filler than "government."
Script
A particular kind of schema that describes an event with a regular, predictable order of actions. Think of it as the mental screenplay for routine events like eating at a restaurant or visiting a doctor.
Inference
Information you add to your memory that was never explicitly stated or observed. You infer that Jane wanted ice cream when she heard the truck, even though the passage never said so.
Understanding discourse requires inference
Without prior knowledge about a topic, it can be impossible to correctly understand a passage or event.
People draw inferences to understand, and then they remember those inferences as though they were part of the original.
Much of what we "remember" was never explicitly said or seen.
Classic example: Charniak (1973)
The passage reads: "Jane heard the jingling of the ice cream truck. She ran to get her piggy bank and started to shake it. Finally some money came out." None of the following is stated, yet you infer all of it:
Jane wanted ice cream.
Ice cream trucks sell ice cream.
Piggy banks hold money (coins).
You need money to buy things.
Jane did not have money readily available.
The truck will leave soon, creating urgency.
These inferences are constructed at encoding. They become part of the memory trace itself.
Constructive memory: gaps filled at encoding
When you encounter a story, event, or picture, you fill in missing information using what you already know. The Jane example above is a textbook case: you "add" her desire for ice cream to the stored memory.
Reconstructive memory: gaps filled at retrieval
Every time you recall information, you rebuild it. The result is that your current beliefs and perspective at the time of remembering can colour what you report. Separating encoding effects from retrieval effects requires careful experimental design.
Memory favours meaning over wording
Demonstration: participants read "The new baby stayed awake all night." When later asked to fill in the blank "The new baby ______ all night," many write "cried," a word that was never in the original sentence. Memory for exact wording of spoken or read material is poor. Memory for the gist, the meaning, is considerably better.
War of the Ghosts study (Bartlett)
Subjects read a war story from another culture. Weeks later, their recall showed systematic errors: they replaced unfamiliar cultural details with familiar ones (e.g. "sailboat" in place of "canoe"). This is one of the earliest demonstrations that memory is shaped by existing schemas.
What a schema contains
A schema is a way to represent knowledge. It has two components:
Slots: the basic dimensions that describe a thing or event (e.g. Buyer, Seller, Item, Compensation).
Fillers: the specific features likely to fill each slot, ordered by typicality or frequency (e.g. the Buyer slot is most often filled by "person," less often by "corporation" or "government").
The Purchase schema
A worked example that shows how schemas operate:
Buyer: Person, corporation, government
Seller: Store, person, corporation
Item: Object, property, financial instruments
Compensation: Money
Event: starts with Buyer holding Compensation and Seller holding Item; ends with Seller holding Compensation and Buyer holding Item.
Restriction: Value of item = Compensation
Schemata in practice: the Woodrow example
Story: "Woodrow went into the candy store. He picked up a Mars bar and handed the clerk a dollar. He left the store."
To understand this, you map the story onto the Purchase schema:
Buyer: Woodrow
Seller: candy store
Item: Mars bar
Compensation: $1
Starting point: Woodrow has $1, candy store has Mars bar
End point: candy store has $1, Woodrow has Mars bar
From this mapping alone, you infer a wide range of unstated facts: this is a purchase; Woodrow is the buyer; he now owns the Mars bar; the store no longer does; he left with the Mars bar but not with the dollar; a Mars bar costs no more than a dollar.
People remember the interpreted content, not the literal input
The schema-driven interpretation is what sticks. You remember the meaning the schema produced, not the exact words you read or heard.
Do schemata help or hurt memory?
Both. You must weigh the helpfulness of schemata in recalling things that happened against the errors they introduce.
Schemata are helpful for recalling "nonaccidental" properties, the features that belong because they fit the schema (e.g. a typewriter in an office).
Schemata are harmful when they produce false memories of schema-consistent items that were never present (e.g. "remembering" books in an office that had none).
Schemata do help memory for unexpected or salient items, because those items stand out against the schema's expectations.
The difficulty is that you usually cannot tell which parts of a memory are genuine recall and which are schema-driven inference.
Scripts defined
A script is a particular kind of schema describing an event with a regular, predictable order (e.g. going to a restaurant: enter, be seated, look at menu, order, eat, pay, leave).
Bower, Black & Turner (1979)
Identified common scripts and tested memory for script-based stories.
Told participants a story of a hero's journey that included irrelevant details (e.g. "it rained") and obstacles.
Found that participants inserted script-consistent actions that were never in the story.
Results: obstacles were remembered 53% of the time; irrelevancies only 32%.
Interpretation: people use the script to understand events and to flag "interesting" events (those with implications for the script). Interesting events are remembered more often.
Brewer & Treyens (1981)
Tested incidental memory for an office. Participants were asked to wait in an office before the experiment began; the waiting room was the experiment.
Participants "remembered" typical office items that were not present (books, a telephone).
They did accurately remember typical office items that were there.
Salient, unexpected items (e.g. a giant stuffed octopus on the desk) were also well remembered.
Takeaway: the office schema both helped recall (for items like a typewriter that fit the schema and were present) and hurt it (false recall of books that fit the schema but were absent).
Schemas and scripts explain why eyewitness accounts of the same event can differ dramatically: each witness fills in gaps with their own prior knowledge. This principle underpins debates in courtroom testimony and police interviewing techniques. It also explains everyday phenomena like misremembering what a friend said in conversation (you recall the meaning, not the words) or "remembering" items in a room that fit your expectations but were never there.
"Memory works like a recording." It does not. Memory is constructive at encoding and reconstructive at retrieval. There is no point at which you have a pristine copy of what happened.
"Schemas only cause errors." Schemas are genuinely useful: they help you understand ambiguous events, recall schema-consistent items, and notice unexpected details. The errors are a trade-off, not the whole story.
"If I remember it vividly, it must be accurate." Vividness and confidence are poor indicators of accuracy. This theme runs through the entire everyday memory literature.
"Reconstructive memory means I'm making things up." Reconstruction is not lying or fabrication. It is an automatic process that happens every time you retrieve a memory, blending stored traces with current knowledge and expectations.
⚠️ Be able to distinguish constructive (encoding) from reconstructive (retrieval) memory and give an example of each.
⚠️ Know the components of a schema (slots and fillers) and be prepared to apply the Purchase schema to a novel example.
⚠️ Understand the dual nature of schemas: they both aid and distort memory. The Brewer & Treyens study is a classic exam question.
⚠️ Scripts are a specific subtype of schema. Bower, Black & Turner's obstacle-vs.-irrelevancy finding (53% vs. 32%) is a commonly tested result.
⚠️ "Memory for meaning is good; memory for exact wording is bad" is a summary statement that appears in many forms on exams.
True or false: Constructive memory refers to what happens at retrieval. Answer: False. Constructive memory refers to encoding; reconstructive memory refers to retrieval.
Fill in the blank: A schema's basic dimensions are called ______, and the features that fill them are called ______. Answer: slots; fillers.
True or false: In the Brewer & Treyens study, participants correctly recalled that no books were in the office. Answer: False. Many participants falsely "remembered" books that were not there.
In Bower, Black & Turner (1979), were obstacles or irrelevancies remembered more often? Answer: Obstacles (53% vs. 32%).
Q: Explain the difference between constructive and reconstructive memory, and give one example of each.
A: Constructive memory is the filling-in that occurs at encoding. For example, when reading the Charniak passage about Jane and the ice cream truck, you infer that Jane wants ice cream even though this is never stated. Reconstructive memory is the filling-in that occurs at retrieval. For example, participants in Bartlett's War of the Ghosts study replaced "canoe" with "sailboat" when recalling the story weeks later, because their cultural schema reshaped the memory at recall.
Q: What is a schema, and how does it differ from a script?
A: A schema is a structured framework for representing knowledge, made up of slots (dimensions) and fillers (typical features). A script is a specialised type of schema that represents an event with a regular, predictable order of actions (e.g. visiting a restaurant). All scripts are schemas, but not all schemas are scripts.
Q: In the Brewer & Treyens (1981) study, what did participants falsely remember, and why?
A: Participants falsely remembered typical office items (such as books and a telephone) that were not present in the room. Their office schema supplied fillers for those slots, and they could not distinguish the schema-driven inferences from genuine memories.
Q: Why were obstacles remembered better than irrelevancies in the Bower, Black & Turner (1979) study?
A: Obstacles have implications for the script: they disrupt the expected sequence and so stand out as "interesting" events. Irrelevancies (like "it rained") do not engage the script and are processed more shallowly. Obstacles were recalled 53% of the time, irrelevancies only 32%.
Q: A friend tells you a story and later you recall details they never mentioned. Which memory process best explains this, and why?
A: Constructive memory. At encoding, you drew inferences to fill gaps in the story using your prior knowledge. Those inferences became part of the stored memory, so you now recall them as though your friend said them.
This material connects directly to eyewitness memory and the Loftus misinformation paradigm (Part 2), where the same constructive/reconstructive processes explain why post-event information can distort a witness's account. It also links to flashbulb memories, where high confidence does not guarantee accuracy, and to the false/recovered memory debate, where schema-driven reconstruction may produce memories of events that never occurred. More broadly, schemas and scripts tie back to the levels-of-processing framework: deeper, more meaningful processing (which schemas facilitate) leads to better memory for gist, even as it makes verbatim recall worse.
Constructive memory, reconstructive memory, constructive processes of memory, schema, schemata, slots and fillers, script, purchase schema, inference in memory, gist memory, verbatim memory, Bartlett War of the Ghosts, Charniak 1973, Bower Black Turner 1979, Brewer Treyens 1981, memory distortion, false recall, schema-consistent errors, everyday memory, PSY 310, cognitive psychology, memory and knowledge