The everyday model of memory is a recording device. You pay attention, the thing is saved, and later you either find the file or you do not.
The recording model gets one thing badly wrong. A new memory is not saved; it is provisional. For hours after you form it, and in a different sense for years afterwards, it is still being processed, reorganised, and either stabilised or discarded. That process is consolidation, and most of the practical advice about studying is really advice about consolidation whether it says so or not.
Two timescales, two different things
Synaptic consolidation happens in minutes to hours. The connections involved in a new trace are strengthened biochemically, which requires protein synthesis. Interrupt this window and the memory does not stick. This is the mechanism behind retrograde amnesia after a head injury: events from the minutes before the blow are lost, while older memories survive, because the recent ones had not finished setting.
Systems consolidation happens over weeks to years. A memory that initially depends on the hippocampus gradually becomes supported by the neocortex, so that over time it can be retrieved without the structure that first encoded it. Along the way it loses detail and gains structure, which is why old memories feel more like summaries than recordings.
The everyday consequence of the second process is worth stating plainly: your memories change every time you revisit them. They are re-encoded on retrieval, in a state that is briefly unstable again before restabilising. Recollection is reconstruction, not playback, and the confident vividness of an old memory is no guarantee of its accuracy.
Sleep is where most of it happens
The strongest and most replicated finding in this area is that sleep actively supports consolidation rather than merely providing a rest from interference.
During slow-wave sleep, the patterns of activity that occurred during learning are replayed, and hippocampal traces appear to be transferred toward long-term cortical storage. Learn something, then sleep, and you retain more of it than someone who stayed awake the same number of hours. The effect shows up in verbal learning, in motor skills, and in tasks requiring insight.
Two things follow, and both are annoyingly practical:
- An all-nighter before an exam trades consolidation for exposure, which is the worst version of that trade available. You add a few hours of studying and remove the process that would have made the previous week's studying durable.
- The nap after the lecture is not a wasted hour. It is part of the studying.
This is covered in more depth in sleep and memory.
Why spacing works, mechanically
The spacing effect is usually presented as an empirical fact with no explanation attached. Consolidation supplies part of the explanation.
Study something twice in one sitting and the second pass happens while the first is still fresh and unconsolidated. There is little to strengthen, because nothing has settled. Study it again two days later and you are retrieving a trace that has been through at least one consolidation cycle, which both tests it and triggers another round.
Massed practice, which is to say cramming, produces a fast rise and a fast collapse. Distributed practice produces a slower rise and a much flatter forgetting curve. Same total minutes, different arrangement, very different outcome a month later.
What this means for note-taking
If a memory keeps being processed after the session ends, then the useful question is not "did I write everything down" but "did I give the trace something to consolidate."
Capture during, process after. Verbatim transcription creates almost nothing to consolidate, because you never decided what anything meant. A shorter note made of your own decisions gives the process better raw material. See why writing helps memory.
Do the rewrite within a day. Rewriting rough notes into clean ones the next morning lands in the window where the trace is still malleable and a retrieval attempt strengthens it. Two weeks later it is an archaeology exercise. How to review your notes sets out the rhythm.
Do not confuse the note with the memory. The file is safe the moment you save it. The memory is not, and it will not consolidate on its own if you never revisit it. That is the honest limit of external memory: it stores information beautifully and does nothing for retention.
Leave gaps on purpose. Time between passes is not lost time. It is when the work happens, which is the counterintuitive core of desirable difficulties.
The caveats worth knowing
Consolidation is a well-established phenomenon, but the popular version overshoots in two places.
The neat story of "hippocampus hands memories to cortex and steps away" is a simplification; for vivid autobiographical memories the hippocampus appears to stay involved indefinitely, and the details are still argued over.
And reconsolidation, the idea that retrieving a memory makes it temporarily editable, is real in the laboratory but is often stretched in popular writing into claims about rewriting your past at will. Treat the mechanism as established and the self-help applications as unproven.
The short version
Learning does not end when you close the book. It ends, or fails, over the following night and the following weeks.
Which means the levers are not only what you do while studying: sleep enough for the trace to set, come back to the material after a gap rather than piling on repetitions, and revisit by retrieving rather than rereading. More in how memory works, spaced repetition, and memory and focus.