The best thing about mindmaps is the ease of rearranging the information. Entire branches can be moved with one operation. You do not get that with mental palaces. Yet, you need a computer to enjoy it… Or not? How did notetaking and mindmaps look like two hundred years ago? People used to research back then. You may start your own research from here.

Personal knowledge base

We should not fully trust our memory. It is very useful, but it can occasionally play tricks with us. Something important should also be written somewhere. The collection of all of the notes we take while reading, researching, brainstorming… How do you call it?

There is no consensus. You can make your own personal wiki, or a folder on google drive. Some people use professional software…

Two hundred years ago there was no software. People relied on index cards in slip boxes instead… To be honest, I discovered the slip-boxes part of the story only recently.


Literary meaning slip box in german…. Slip boxes or pull boxes are kind of large drawers in the offices.

Unlike the regular drawers, we put a couple of notebooks into, the slip-boxes are dedicated for one task only. They are strong enough for quick manipulation even when full of paper. And they contain index cards. The index cards are of uniform size and they are somewhat thicker than regular paper: rigid enough to stand in the box and be manipulated individually. You can buy them new for example using the keywords “Lattice Desktop Drawer Storage Box”. Only two hundred years ago they were better, more suitable for quick examination of the box content.

The origins of the index cards

Once we have boxes full of paper, how do we find anything within? It is a lot of information to browse through… This is the basic issue faced by any librarian, so there are very good and very old answers. We can still observe some parts of the index of the books in the ancient library of Alexandria: Callimachus’s Pinakes…

The idea is to maintain effective cross-referencing between the boxes. It was initially implemented as a library catalog by Conrad Gessner (1516–1565) and improved since. The sociologist Niklas Luhmann (1927–1998)  built up a zettelkasten of some 90,000 index cards for his research and credited it for enabling his extraordinarily prolific writing including over 70 books and 400 scholarly articles,

In my speedwriting course, the focus is often on how to organize information in computer media. Yet, I bet that even with the best computers very few of you could beat Niklas Luhmann or Agatha Cristie in quality content creation. And these people used index cards for their research and for the outlines of their books.


Each note or a card in the drawer is filled as soon as the information is acquired. Yet, some space is usually left for adding further data in the future.

The most important part of the metadata is the number, which I will address soon. Before the number, notice the tags. Tags are the ~5 keywords and combinations of keywords describing the piece of information. Just enough to understand what the document is about and associate one document with another. These are basically your associative connections that can be used both for data synthesis and data retrieval.

Next, there are cross-references to other documents. They allow following an idea across multiple boxes.

Here is an example. A zettelkasten is made up of notes containing numbers, tags (blue) and cross-references to other notes (red). A tag index (bottom right) allows topical cross-referencing.

The content of the card itself is usually the location of a book or a folder and file within. This allows complex manipulation of the files without changing the basic numbering.


Basically, all of our declarative knowledge can be ordered as one huge document with a complex outline. Then hierarchically step by step we open new areas within the system. Today the outline is recalculated automatically any way we want. Two hundred years ago, it was fixed. The numbers were damn hard to change. In a way, it was harder to fix a mistake, but it was also significantly easier on one’s memory.

As we progress to increasingly specific subjects, we get more longer hierarchical number. We have the basic scientific discipline, within we have paradigms, and then we have specific areas of research, themes within the area, and so on. The number corresponding to a specific paper is often within 7 layers of hierarchy. The PAO version used by Jordano Bruno had 7 layers, like the number of planets known during his lifetime.

And once the number is set, moving it is not a real option. So if you need to modify the content, you better have a folder within.


While the inner organization of the database can easily be a set of indexes, to navigate properly it is best to use some sort of visualization. Mindmaps are much faster to visualize and navigate, than trying to address complex hierarchical numbering in your mental palace.

Yet, if you want to follow an association to some other mindmap, you should have some sort of coordinate where you point. The most common process in index cards is adding a new piece of information and linking it to existing information. Adding branches is less common. Moving branches around can be a challenge.

Notice that while mindmaps are easy to manipulate, there is no good reason for that except for a massive paradigm change.  If this happens, all associations need to be reviewed and multiple cards need to be replaced. Not very convenient. Even worse with our own stable visualizations.

You can change anything you want while creating the visualization. Then you have about a week to finalize it. After that, any change requires complex and costly relearning with spaced repetition or equivalent…

Quite often color schemes are added to facilitate long-term navigation. After all, you do not want to mix supporting data, pros and cons communication, research methodology, and funny facts about the researchers. So do not allow your visualization to run wild if you need long-term memorization. They need to preserve some color schemes.

Scale and organization

Complex associative indexing is usually effective for large and constantly growing bodies of knowledge. For a single document, a linear process with a fixed outline works better. For several documents, a folder with files can be ideal. Only when we get to larger systems the approaches change.

In a way, we can talk about a mental palace. A library has a given number of rooms, a given number of cabinets within each room and shelves within a cabinet. This is essentially a fixed mental palace. If a book is no more needed, it is moved from the library to some amorphous archive. Good luck finding anything there…

Each layer of knowledge benefits from a dedicated methodology because we use them differently. I mean, how often do you discover a new faculty in a university and venture into its library? And withing your library, you probably spend most of your day in the same rooms. As for specific books, you probably read and discard many of them and go back only to a selected few.

And it does not matter if the library is physical like it was when I was a child or virtual like it is today. The way we address knowledge did not change drastically. Should it change?

Men vs computers

Computers can find information in ways that are very complex for humans. Searching a very long list is easier for computers. Going through complex chains of associations is easier for a human. In computers, the information is often almost flat without the deep hierarchy we (humans) often enjoy.

Programmers are slowly ditching relational databases for something less organized. At the same time, the documents themselves become complex hierarchical structures rather than simple tables.

We are moving more structure into each specific piece of knowledge and away from the systematic organization of knowledge, possibly because everything changes faster than ever before. It is very hard to maintain large systems when paradigms change every year.

To be honest, zettelkasten is often the last-ditch backup in the unlikely event of a cyber catastrophe. And a cyber catastrophe can happen. Unlike nuclear weapons, cyber weapons are notoriously hard to control…

What is in it for me?

  1. Try to maintain some sort of backup for some of your critical information. I know this is hard, and none of us is good at this sort of activity.
  2. If you need to maintain a large body of knowledge, do not neglect cross-referencing and associative connections.
  3. Consider a fixed structure and numbering policy for your knowledge.
  4. Understand that fast research requires a very good organization of acquired knowledge.


Get 4 Free Sample Chapters of the Key To Study Book

Get access to advanced training, and a selection of free apps to train your reading speed and visual memory

You have Successfully Subscribed!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.