Lecture 8: Memory - Storage, Retrieval and False Memories

What this lecture covers

This lecture asks how information moves from perception into short-term and long-term memory, what makes it easier or harder to retrieve, and why memory is far less reliable than it feels. It builds on the previous lecture's introduction to attention and change blindness, using that unreliability of perception as a bridge into memory itself.

You'll come away understanding the practical difference between short-term and long-term memory, why organizing and understanding material helps you remember it, and why vivid, confident memories, including eyewitness testimony and flashbulb memories of dramatic events, can still be wrong or even implanted. This sets up the following guest lecture on love and attraction.

Key ideas

• Short-term (working) memory capacity: roughly seven plus or minus two items, measured in "chunks" rather than raw symbols, so recoding information into fewer, more meaningful chunks (like known words instead of loose letters) lets you hold more.
• Long-term memory (LTM): a vast, largely unmeasured storage system holding vocabulary, faces, facts and life events, distinct from the limited-capacity short-term store.
• Depth of processing: memory improves dramatically when you process a word's meaning rather than its appearance or sound, which is why understanding beats rote rehearsal for moving things into long-term storage.
• Maintenance vs. elaborative rehearsal: simply repeating information (maintenance rehearsal) keeps it active only briefly, while connecting it to other knowledge (elaborative rehearsal) helps it stick.
• Retrieval cues and the compatibility principle: memory is easier when the retrieval context matches the encoding context, illustrated by the scuba-diving experiment and context-, state-, and mood-dependent memory effects.
• Forgetting has multiple causes: physical decay of memory traces, interference from similar new information, and changes in available retrieval cues over time.
• Anterograde vs. retrograde amnesia: retrograde amnesia loses past memories (often from a blow before consolidation), while anterograde amnesia (as in H.M. and Clive Wearing) blocks forming new explicit memories, though implicit memory, like skill learning, can still improve.
• Memory suggestibility: leading questions, suggestive interviewing, and hypnosis can implant false but sincerely believed memories, as shown in Steve Ceci's Sam Stone studies and Elizabeth Loftus's eyewitness research.

Walkthrough

Change blindness as a bridge into memory (0:00)

The lecture opens with another real-world change-blindness demonstration, reinforcing the previous class's point that people's sense of perceiving a stable, detailed world is largely an illusion; the visual system and memory exploit the fact that most things stay the same, so precise representations usually aren't needed. Bloom uses this to reintroduce the three memory types from the prior lecture (sensory, short-term/working, and long-term) and frames the day's topic as how information moves from momentary sensation into lasting storage.

Short-term memory and chunking (0:00)

Bloom contrasts long-term memory's huge, essentially unmeasured capacity with short-term memory's sharp limit. A live number-recall demonstration illustrates George Miller's finding that short-term memory holds about seven items, plus or minus two, but that the relevant unit is the "chunk," a meaningfully grouped piece of information, not a raw symbol. Examples include recoding letter strings into words, binary digits into a single number, and chess positions or football diagrams that experts remember far better than novices because their knowledge lets them chunk the material meaningfully.

Getting information into long-term memory (13:22)

The lecture distinguishes maintenance rehearsal, which merely repeats information and rarely produces durable memories, from techniques that build structure and meaning. The classic "levels of processing" experiment is described: subjects who judged whether words fit a sentence's meaning remembered them far better than those who judged only appearance or sound. Bloom also covers mnemonic devices, such as vivid imagery for names, and reads a passage that becomes far easier to remember once its topic (washing clothes) is revealed, illustrating that understanding drives retention more than repetition.

Retrieval cues and forgetting (18:27)

Bloom explains retrieval cues as associations that bring memories back, and the compatibility principle, the finding that memory is best when retrieval context matches encoding context, illustrated by a scuba-diving experiment and by context-, state-, and mood-dependent memory. Elaborative rehearsal (deliberately connecting new material to existing knowledge) and elaborative retrieval (persisting through structured questioning to recover memories people initially think are lost) are both described as ways to improve recall. He also offers a consolidation-based theory of déjà vu, where a slight delay in encoding creates the illusion that a just-experienced event happened earlier.

Why we forget, and the puzzle of childhood amnesia (25:50)

Three explanations for ordinary forgetting are laid out: physical decay of memory traces, interference from newer, similar information, and changes in retrieval cues over time. This leads into childhood amnesia, the well-documented but poorly explained difficulty adults have recalling events from before about age three, with candidate explanations including changing retrieval cues, the reorganizing effect of learning language, and neural maturation of memory-related brain regions.

Brain damage and the two kinds of amnesia (25:50)

Bloom distinguishes retrograde amnesia (loss of memories from before an injury, often due to a period before an event that never got consolidated) from anterograde amnesia (inability to form new explicit memories), the pattern seen in the well-studied patient H.M. and in Clive Wearing. He highlights the key discovery that such patients can still form new implicit memories, shown through a mirror-tracing skill task where performance improves across sessions even though the patient reports no memory of having done the task before, and references the film Memento as a vivid illustration of this dissociation.

The power of suggestion: implanted memories (37:15)

Using Steve Ceci's "Sam Stone" studies, Bloom shows that children exposed to a stereotype about a clumsy visitor and to suggestive questioning came to sincerely believe he had ripped a book and misbehaved, even though nothing of the sort happened. Elizabeth Loftus's eyewitness research is then introduced, showing that subtle wording changes, such as "the broken headlight" versus "a broken headlight," or "smashed" versus "hit," can shift what people later report remembering, including details, like broken glass, that were never present.

Hypnosis, repressed memory, and flashbulb memories (44:56)

Bloom argues that hypnosis does not unlock stored memories like a recording but instead makes people highly motivated to produce a convincing account, often fabricating detail without realizing it. He covers the contested debate over repressed memories of trauma, noting that while some psychologists believe genuine repression occurs, most consider recovered memories obtained through suggestive therapy unreliable. The lecture closes with flashbulb memories, such as recollections of September 11, arguing that their vividness and confidence do not guarantee accuracy, since retelling and hearing others' accounts can blend or distort the original memory. He then takes final audience questions on implicit memory for skills and on the doubtful evidence for true photographic memory.

Before you watch

• Recall the attention and change-blindness material from the previous lecture, since this lecture opens by extending those findings into memory.
• Be ready to do the in-lecture number-recall exercise actively (without writing down numbers as prompted) since it demonstrates the short-term memory limit firsthand.
• It helps to know the difference between explicit and implicit memory from the prior lecture's overview, since several examples here (skills, amnesia patients) depend on that distinction.

Check your understanding

1. What is a "chunk," and why does chunking let people remember more than the raw number of items would suggest?
2. Explain the levels-of-processing finding and why it implies that rote repetition is a weak strategy for long-term retention.
3. What is the compatibility principle, and how did the scuba-diving experiment demonstrate it?
4. Contrast retrograde and anterograde amnesia. Why did mirror-tracing tasks change how psychologists understood memory in anterograde amnesia patients?
5. How did Elizabeth Loftus's leading-question experiments show that memory can be altered after the fact, and why does the Sam Stone study suggest these false memories are sincerely believed rather than lies?