Encoding and Retrieval

Encoding is the first-pass work of taking material in and putting it into a form that can later be rebuilt from memory. Retrieval is using that material from memory before checking the source: recalling it, explaining it, solving with it, or teaching it.

A better first pass means fewer later retrievals from memory. The number of later retrievals never reaches zero. Forgetting continues over time, so later retrieval has to stay scheduled.

Encoding and retrieval are kept separate because taking material in well and getting it back out well are different actions. Intake and access still change each other. The two are not competing techniques. The useful question is how study time is split between them, and the answer changes as skill changes.

Core takeaways

• A better first pass means fewer later retrievals. Later retrievals never drop to zero, because forgetting continues.
• Retrieval fails for two reasons. Either the first pass was thin, or the first pass was adequate and access to the memory has decayed with time.
• Retrieval changes the memory it draws on. Each later retrieval is also a new encoding event.
• While encoding skill is weak, most study time goes to retrieval. Retrieval shows a result within days; encoding skill takes much longer to build.
• Start interleaved retrieval first. Add a new encoding method in a later week, so two method families are not being learned at once.
• The default schedule is one day, one week, and one month. Any kept schedule works better than none, and the gap grows with how long the memory has to last.

Why retrieval never drops to zero

Retrieval can fail for two reasons. The first is intake: the first pass was thin, so there is little in memory to rebuild. The second is access: the first pass was adequate, but the memory can no longer be reached.

Material taken in as a network of relations decays more slowly on the intake side than material taken in as a list of separate facts. So later retrievals are needed less often. Access decays with time no matter how good the intake was. This is why a first pass, however good, never removes the need for retrieval.

Retrieval has a second effect beyond checking the first pass. Recalling material from memory changes the stored memory. Each later retrieval is also a new encoding event.

Where this week’s study time goes

While encoding skill is still weak, most of the week’s study time goes to retrieval. Encoding skill takes a long time to build, and a learner cannot easily judge their own encoding skill from the inside. Retrieval shows a result within days. Even a thin first pass, tested rather than restudied, scores better on a delayed test within that window.

The first habit to build is interleaved retrieval, applied to whatever encoding the learner already has. A new encoding method belongs to a later week. Starting both in the same week means learning two new method families at once, and neither gets learned to competence.

The interleaved retrieval habit also puts regular spacing in place. Mixing topics while retrieving forces Higher Order Learning, which later makes a new encoding method easier to pick up. That is a reason to start with interleaved retrieval. There is no finding that interleaving teaches encoding skill.

The early cost, and why it falls

The early phase takes many sessions, exposes many gaps, and uses a lot of time. That is the expected state while the first pass is still weak. Each pass restores what has decayed and re-encodes the material, so the next pass covers more material in less time. A month of material does not take a month to revisit. The workload is expected to fall.

Noticing that the first pass was thin is conscious incompetence: the drop in confidence that comes when the learner can see what they missed. This differs from the overconfidence finding, which describes people who cannot yet see what they missed. Noticing the gap is the start of the repair, and it is not itself a failure.

The retrieval schedule

The default schedule is a retrieval pass at one day, one week, and one month. This is a workload default. Keeping any schedule works better than keeping none. The useful gap still grows with how long the memory has to last. While encoding is weak, the schedule is what keeps the material in memory. Better encoding later allows a looser schedule.

A same-day pass supports encoding and does not count as the first real test. The standing system is Spaced Interleaved Retrieval. The formats are listed in the Interleaving Table. After a few days of interleaved retrieval the gaps should be visible, even if encoding still feels confused.

What changes as encoding improves

Later retrievals take less time as the first pass improves. They do not stop. This week’s work is the retrieval habit, on the encoding the learner already has. The early cost is the cost of starting. Better intake later allows a looser schedule. The memory still needs retrieval.

How to practice this

1. Start interleaved retrieval this week on material you have already taken in. Recall, explain, solve, or teach it before you check the source. Notice that gaps become visible within a few days, even while encoding still feels confused.
2. Mix topics inside each retrieval session. Notice that this forces Higher Order Learning. Do not expect it to teach encoding skill.
3. Do not add a new encoding method in the same week. Notice whether you are trying to learn two method families at once. Neither gets learned well when both start together.
4. Schedule retrieval passes at one day, one week, and one month. A same-day pass supports encoding and does not count as the first real test. Notice that a kept schedule holds the material better than no schedule.
5. When a retrieval pass shows your first pass was thin, keep going. Notice the drop in confidence: you can now see what you missed. Seeing the miss is the start of the repair.
6. Keep running the passes across the month. Notice that each pass covers more material in less time than the one before.

Related pages

• Spaced Interleaved Retrieval: the standing retrieval system this page is the trade-off for. It covers spacing, interleaving, and method by knowledge type.
• Interleaving Table: the method menu. This page does not teach formats.
• Revision: the first retrieval loop.
• Higher Order Learning: what interleaved retrieval is supposed to force even before encoding skill arrives.
• Schema: a sibling page on encoding quality, about what survives pruning.

1 more related pages

• Bear Hunter System: the encoding method that eventually reduces retrieval load.

Open questions

• How far the early-phase time cost actually falls as encoding skill rises, for a learner who never adds a formal encoding method.

Sources

• Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior. Deeper, more relational intake slows forgetting.
• Craik, F. I. M., & Tulving, E. (1975). Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology: General.
• Ebbinghaus, H. (1885/1913). Memory: A Contribution to Experimental Psychology. Forgetting is real; retrieval never goes to zero.
• Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science. A test beats restudy on a delayed test within days.
• Rowland, C. A. (2014). The effect of testing versus restudy on retention: A meta-analytic review of the testing effect. Psychological Bulletin. Retrieval practice needs an encoded trace; it does not replace first-pass encoding.
• Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review. Encoding and retrieval influence each other.
• Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science. Holding test delay fixed and varying only the gap moved final recall along an inverted U. The 1/7/30 rhythm is a workload default; gap size still scales with how long the memory is needed.
• Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it. Journal of Personality and Social Psychology. Overconfidence at the bottom, in people who cannot yet see the miss, which differs from the dip that arrives when the miss becomes visible.