Who creates mnemonic methods?
Mnemonic methods were originally developed for public speakers. In the middle ages, they were used by clergy to remember scriptures and by philosophers to hide things from the clergy. As paper and writing became ubiquitous, memory arts were used by preachers and after-dinner speakers. The next round of development was done by memory sportsmen and magicians, rivaling with computer scientists.
Computer science approach
Traditional memory arts used mental palaces as the most natural form for the human mind to remember things. The computer science approach utilized mathematical theories to help machines remember things.
Computer scientists try to represent knowledge in the way most natural for the knowledge itself. The theory of graphs, information theory, and programming paradigms are often used to describe knowledge templates.
Trees and hyperlinks are basic tools used to represent hierarchical knowledge and connections between various kinds of knowledge. Mindmaps are one of the early attempts to adopt this approach for human memory.
Big data
To be honest we use computers in our work. Basically we need to remember what to look for and where we can find it. Then we can access the right information. On the other hand, we need to be able to modify what we know as new information is available. These operations can be easily done on mindmaps. Mindmaps have operations like replacing a node, adding or removing branches, associating some node with a node on another mindmap. Doing something like that on a mental palace would be difficult.
However, when we discuss things in a conference room or talk in front of people, we do not have the luxury of taking 5 minutes to reread an article. For these situations, we can use mental palaces as a way to remember individual articles and navigate within.
To deal with this situation I suggest using “mental cities”, entire visualized cities where the roads are mindmaps and the buildings are mental palaces.
Formulating thoughts
A very different challenge is formulating our own thoughts. When we write text, it can change all the time. So we can effectively represent it with a mindmap. However, we write many documents side by side. On a computer, we would put these documents in folders.
If the data structure of each document is a tree, a good representation of multiple documents would be a “mental forest”. To be honest, this representation is usually also available in computerized form.
In my masterclass, I explain how to actually build and remember mental forests step by step. While the things we wrote appear to be above the ground, there is also a complex underground tree of our reasoning. Trees have roots, you know…
You can use a different representation.
Reading order
When we read we usually take a two-step approach. We can preread, allocating the data structures, and then fill in the details while reading. Alternatively, we can read and fill in the data structures, with optional rereading and refinement step.
Prereading is something we typically do when facing a formidable article or a book. For smaller pieces, rereading is more likely to happen. This means, we read an article and filled in the mental palace, and then we want to make a very detailed representation of a particular detail, so we add a mental map on the wall of the mental palace or place a mental palace in the node of the mental map.
Even experts do not usually realize that this is an option. We feel somehow more comfortable placing portals in mental palaces and hyperlinks in mindmaps. Adding details to visualizations is a standard exercise. But we can also take a more direct approach. Why not?
More complex visualizations
We could layer hybrid structures if we have to. A mental city may have a garden which can work as mental forest, and on the branches and between the roots of the trees we can put fairy treehouses. On the wall of such a house can be a map. You can use your imagination.
The more complex structures add further flexibility, but the visualization and navigation get slower and less reliable. If we do not use the same mental structures every day, we will not be very skilled using them, and who honestly can use something very complex every day?
Multiple dimensions
To simplify things most of our visualizations do not use zoom and operate just on one height level. Of course, we could make things more interesting using multidimensional structures, for example for creativity training. Some more complex literary works use layers of stories and substories. Computer graphs and data structures allow a very complex representation of information. I remember before the age of neural networks, working with support vector machines, I actually visualized multidimensional manifolds.
Usually, the easiest way to work with such monstrosities is visualizing separate two-dimensional cross-sections, while understanding how these cross-sections happen. This means that the same information will appear several times in the relevant memory structures, with various meanings based on the way we analyze the situation. Mechanical engineers usually draw the same part from three positions to represent it. Electrical engineers draw multiple layers of their circuits with connectors between the layers. Why not do the same with memory structures?
A simple answer: because these structures are hard to get used to. If you are not an engineer or a biochemist, you might never need this skill.
Chunking and linking
Our information may be very simple or very complex. Our methodology is usually very human. We group together similar items and build associations between groups. These operations are called chunking and linking. The faster we perform them, the faster we process our information into mental structures. Since we consume a lot of information, it makes sense to become experts in chunking and linking.
Mindmaps are an exercise in linking. We build interconnected hierarchical structures by linking all the information we get.
Mental palaces require very precise chunking. We usually have PAO visualization, which means trigrams with optional details. Then we get four or eight of such trigrams per room. And we do this along fixed itineraries, Adding or removing info in mental palaces after the initial memorization is very inconvenient. Reducing the flexibility we can increase speed and accuracy both of encoding and of recall.
Templates
If we can predict the information we encounter we can use templates. These are premanufactured mental houses or mindmaps, maybe even mental forests and cities. We do not need to create new chunking but simply populate the existing structures in fixed locations.
Assuming zero flexibility, templates make our work very easy and very fast.
When we allow infinite flexibility we get something we cannot use efficiently.
No use case for something else
Something like the major method could help you remember a random linked collection of alphanumerical information. I do not even have a use case where this skill is more effective than the alternatives. When we get to the point where we need to remember alphanumerical info, it can usually be easily chunked for effective PAO and mental palace implementation.
Flow diagrams and engineering data can definitely be more complex than any mindmap or mental palace, but then we do not have mnemonics for this… If the data gets too complex, we remember it as we design it, including the design software quirks.
Go to the basics
In Russian folk takes, in a deep forest there is a house of Baba Yaga. For me, this is a cool demonstration of a mental palace within mental forests. If this is not too hard for Russian kids, why would it be hard for memory experts like you?
