Against spaced repetitions

This is not a typo. I am actually arguing against spaced repetitions. For me, it is not the perfect solution for long-term memorization, but a hard choice for hard situations. What makes me go against the golden standard of long-term memory, and why?

When spaced repetitions shine

Spaced repetitions are great to remember rare words, dates, and numbers. We encounter rare words rarely, and if we use them in a sentence our audience is likely to frown. So we use the most artificial memory device I know, to bring those words up again and again until we know them intimately well.

There are small tricks about using the tools correctly. Like choosing the right software and presentation, formulating questions, and reducing exposure for the things that we know. Each language guru has a slightly different set of tools and rules. Most of these tools actually work quite well for their purpose: to learn rare words.

Ignoring the logic

The first issue that is fundamentally solvable: adding logic to the spaced repetition. Design patterns in programming, engineering standards, and medical protocols, legal precedents, and math formulas… There is a lot of things we may choose to encode in cue cards.

When we encode “meaningless” data, like rare words, we may choose etymology or ontology method, but may also encode the words as-is. However, when we encode highly logical information, we need to encode the logical connection itself. This logic is not a part of the data we try to memorize, it is a part of the derivation.

Very few memory experts actually choose to encode the logic and derivation in the cue cards. This goes against the mental palaces and major methods used for mnemonics. It looks too much like what good teachers used to do a hundred years ago. Guess what, many of our ancestors were actually well-educated and wise.

Write an essay for future search

Actively using and combining various data sources is very important. Yet you probably start to hate me. There is a good reason. School essays are very bad. Students are graded for their ability to generate some arbitrary template, and not for the effort to connect various data sources.

Why do the students wrap copy-pastes from many sorts into a format that makes plagiarism detection software fail? The kids often hate this so much, that the parents write texts for them. Why do debate clubs focus on rhetoric devices rather than research?

School essays are really bad for actually learning something. Write an essay in any form you want, combining the elements you learned. The idea is practicing the way we formulate our markers into text. The more logic (not data pieces) you can rephrase and connect between the sources from memory, the better.

Rely on Google

Remembering attributes [or keywords] is more important than visual or sound associations.

If you remember the name Lasalle (arbitrary choice of general)  as “a big room”, you may try to write Lesaloon instead. Then you will not be able to find the guy in the search engine… “Napoleonic hussar general” will actually work in Google

For example, I told my kids about Robert Liston who: Amputated the leg in under 2​1⁄2 minutes (the patient died afterward in the ward from hospital gangrene; they usually did in those pre-Listerian days). He amputated in addition to the fingers of his young assistant (who died afterward in the ward from hospital gangrene). He also slashed through the coattails of a distinguished surgical spectator, who was so terrified that the knife had pierced his vitals he feigned from fright. I always forget the name of the doctor, so I Google “operation scot surgeon 19th-century” to get the name.

You do not have to memorize everything, just remember enough to find the data on google in less than 1 minute.

An enormous waste of time

The way students use their memories looks like an enormous waste of time. Before learning mnemonic devices they rely on repetition, which is very bad. An average student needs to repeat a sentence six times or more to remember, and then to read it once a week for six weeks to keep the data in memory.

After learning the mnemonic devices, students revisit their memory structures. This is actually not that bad, but still effort-intensive. Learning 100 words can be as simple as filling one memory palace. Then revisit the palace after a day, after a week, after a month, and after three months. And that should be enough for most data sources.

Yet this way there are still at least four revisits. The memory structures from different articles do not connect with each other. And there is no output to the effort. Also, we are not likely to remember memory palaces after 10 years. I tend to forget them after two years. We are not even likely to revisit. We are much more likely to remember and revisit our own essays. Even if we write a lot.

Memory experts cheat

To be honest, many memory experts remember their structures years after and possibly as long as they live. Probably not due to spaced repetition software. They use the data they know if private meditations, public speaking, or even in writing. This is a cheat, in a way.

You do not need a complex mnemonic structure and spaced repetition if you think, speak, and write the subject. Mnemonics and writing methods might reinforce each other, but otherwise, they are redundant.

We kind of use the mnemonics to remember the articles we read for a couple of days. This is the delay needed to collect enough information to daydream or write an article. It is good, helpful, and instrumental, only it has very little to do with spaced repetitions.

Intense repetition

When we work or a project, we rarely use spaced repetition. We remember the information because we repeat it in multiple situations. First, we search for it. Then we read it. Then we write down our ideas. Later we reread to see that we did not miss anything. And we edit our own results. Possibly we use the information to pitch the project to collaborators. We repeat it so many times, that it becomes a part of our autobiographical memory.

And then there is information that does not need conscious repetitions. When we are shocked by something, not necessarily negative, it is often played in our head in a loop. Or we may daydream about the information in multiple contexts.

What is the effect of this? Reading something six times is enough to remember it for a short while.  Using something twenty times in different contexts is usually enough to remember it. Especially if there is a strong emotional effect. And usually, this happens almost spontaneously: without software or cue cards.

State dependency

Our memory is contextual. We usually work with spaced repetition in a narrow context. For example, we use one program on one device in one pose. Now, this narrow context helps us to ace the program and minimize the exposure of the relevant information. This part is great. However, when we want to use the same info in a different context the recall is not that good.

When we use the information actively we create more recall opportunities: in different contexts, associated with other information, in different poses. We are more likely to recall the information this way.

Suboptimal even for languages

Did you notice that I emphasized the use of spaced repetition with rare words? For a very good reason! When using common words, immersion, writing, or debate practice is much better.

In fact, even in language learning, spaced repetition should be the last resort. Use it when you are not likely to encounter the word otherwise. If this is a common word, do not waste your time.

Now if this is a rare word, you may need it if you are fluent. Novice or intermediate language users will not need to learn rare words.

If you are a professional linguist, even rare words will not be meaningless. They will remind you of something else that you know. You might remember etymology. There will probably be some cool linguistic context…

So while spaced repetition definitely works… May I ask WHY do you insist on using it, rather than consider the alternatives?