Core topic 03 of 13
How memory actually works, and why it edits itself
Memory does not store events and play them back. It rebuilds them each time from fragments, and the rebuild is shaped by what you know now, what you were asked, and who asked you. That is why two people who stood in the same room describe different rooms.
Frederic Bartlett showed this in 1932 by asking British students to retell a Native American folk tale over several weeks. The story steadily lost its unfamiliar elements and gained details that fit English cultural expectations. Nobody was lying. Their memories had reorganised around what made sense to them.
What holds up
The forgetting curve Replicated
Hermann Ebbinghaus taught himself nonsense syllables in the 1880s and measured how fast they disappeared. Loss is steepest in the first hours and then flattens. Jaap Murre and Joeri Dros replicated the original study in 2015 with modern methods and got a comparable curve, which is unusual longevity for a nineteenth century result.
The spacing effect Replicated
Spread the same total study time across days and you remember far more than if you pack it into one session. Cepeda and colleagues reviewed 254 studies in 2006 and found the advantage across ages, materials and delays. The optimal gap grows with how long you need to retain the material, roughly 10 to 20 percent of the target interval.
The testing effect Replicated
Trying to recall something teaches you more than reading it again. Roediger and Karpicke demonstrated this in 2006, and the effect has replicated in classrooms as well as laboratories. Students reliably predict the opposite, rating rereading as more effective while performing worse on the test.
The misinformation effect Replicated
Elizabeth Loftus and John Palmer asked people how fast cars were going when they smashed into each other, and a separate group when they hit each other. The smashed group gave higher speeds and were more likely to report broken glass that was never in the film. Wording after the event changes the memory of the event.
Flashbulb memories feel accurate and are not Replicated
Talarico and Rubin recorded students' memories of 11 September 2001 the day after, then again weeks and months later. Detail accuracy dropped like any other memory. Confidence stayed near ceiling. The vividness is real, the reliability is not.
The contested corner
Recovered repressed memories Contested
Whether traumatic memories can be pushed out of awareness and later retrieved intact is one of the sharpest disagreements in the field. Clinicians report cases. Memory researchers point to the ease of implanting detailed false memories in the laboratory and the absence of a demonstrated repression mechanism. Both sides agree that suggestive questioning during therapy can create memories, which is why interview technique matters so much in legal settings.
Working memory capacity of seven items Contested
George Miller's 1956 paper on the magical number seven is still quoted as a hard limit. Miller himself called the number a coincidence in the paper's own words. Nelson Cowan's later work put the figure closer to four chunks when rehearsal is blocked, and capacity varies with the material.
What this means if you are trying to learn something
The two techniques with the strongest evidence, spacing and retrieval practice, both feel worse than the ones that do not work. Rereading and highlighting produce a sense of fluency that your brain reads as knowing. Struggling to recall an answer produces the opposite feeling and the better outcome. John Dunlosky's 2013 review of ten study techniques put spacing and practice testing in the top tier and highlighting, rereading and summarising in the bottom.
Sleep does the consolidation work, which is covered in more detail on the sleep and dreams. Cramming until 3am is a straightforward way to lose what you just studied.
Watch this on the channelShort videos on false memory experiments, the forgetting curve and study methods that actually work.
Open Psychology Cue on YouTubeQuestions people ask about this topic
Why do I forget things so quickly?
Forgetting is fastest right after learning. Ebbinghaus measured the drop in 1885 and Murre and Dros replicated it in 2015. Most of what you hear in a lecture is gone within a day unless something brings it back.
Does spaced repetition actually work?
Yes, and it is one of the most reliable findings in psychology. Cepeda and colleagues confirmed it across 254 studies in 2006. Increasing gaps between reviews beat cramming at every age and subject tested.
Can memories be false without a person lying?
Yes. Leading questions, doctored photographs and suggestion from a trusted source can produce detailed memories of events that never happened, and the person experiences them as ordinary memories.
Are vivid memories of shocking events more accurate?
No, they are more confident. Talarico and Rubin found that accuracy for 11 September memories decayed at the normal rate while confidence stayed high.