The ninth rule in 20 Rules for Formulating Knowledge is “avoid sets”. What does this mean in terms of study methods, and how can it be applied to language learning?

A set means some sort of question or prompt where the goal is to recall all the members of a group. For example, trying to list out the uses of all three de particles in Chinese grammar, the different types of le or a particular set of vocabulary. Outside of language learning, it might be a prompt like “what are the constituent countries of the United Kingdom?”

This type of memorisation prompt is generally a bad idea, whether it’s for learning Chinese or something else. Here’s why.

Maximum information

Rather than following the minimum information principle, set-based learning tries to get you to remember information in large, unwieldy chunks. This slows your learning down dramatically.

When information is broken down into small parts, you’re able to focus on the problem areas and spend less time on the material you’re fluent with. Sets put a big obstacle in the way of that and cause frustration, as one difficult item will repeatedly disrupt your learning of all the others.

Sets aren’t realistic

Another problem with sets is that they are rarely a good representation of how you use knowledge in the real world. Take a Chinese vocabulary list as an example: you almost never need to recall a list like that in real world situations. Instead, you need fluency and association for the individual items in their own contexts.

Sometimes badly designed tests will require you to memorise sets. This is a shame, as it forces the learner to waste time brute-forcing knowledge that won’t help them much with the real skill they’re trying to acquire. If there’s no choice, then in these situations the best approach is to use mnemonics targeted at recalling the set (see below).

Sets are likely to be inaccurate or change over time

In the same way that they tend not to reflect real world skills, information sets are more likely to be inaccurate or change over time than more bite-sized knowledge.

For example, imagine if you’re tasked with memorising “all characters with 亻 (rén zìpáng)” in your first year of learning Chinese. The task is really to memorise all the 亻 characters covered so far in your textbook, and the set is always going to be incomplete. It will only ever be useful on the exam.

There’s little value in memorising such a set, but there is a lot of benefit to studying related material like that to build some useful mental links. Studying and considering a set of 亻 characters is helpful, but trying to memorise the set is not.

How to deal with unavoidable sets

Sometimes, you’ve got no choice but to memorise a set. The most common situation is that this is required by a textbook or exam. How should you deal with this?

The first thing to do is see if you can convert the set into an enumeration. An enumeration is a set with a specific order. Enumerations are not ideal either, but they are slightly better than sets because the specific sequence helps in recalling the whole thing.

For example, Chinese characters can be seen as small enumerations of component parts. Stroke order is emphasised so much in learning to write hanzi because it makes them easier to remember (as well as helping you to write more balanced and consistent characters).

Whether you have to deal with a set or an enumeration, mnemonic techniques are the way forward. You can often make up a song or story that covers the elements of the set. Even more powerfully, the memory palace technique allows you to memorise surprisingly large sets of information.

In short:

Try to avoid sets altogether, as they are a hindrance to acquiring knowledge.
If you can’t avoid having to learn a set, make it an enumeration (the component parts of each hanzi being a good example).
Use mnemonic techniques such as mini stories and phrases to recall the enumeration.