Memorizing tables is hard. Fortunately, there are not many people that need to memorize many tables and not many tables that most of us need. There are some adaptations for regular mental structures and some specific memory structures which work best with tables. Please consider the pros and cons of each approach.
Tables of contents are actually mindmaps
The word “table” might be a bit misleading. The most useful tables do not have a tabular structure.
A table of contents is a mindmap. It is a highly structured hierarchical data that can be altered when we refactor our documents. We remember it like a regular mindmap. Do not try to use mental palaces for that. Even ancient Roman orators had specific structures for the divisio of the content, that were unlike the method of loci used to remember specific facts.
Table rows as long lists
In many tables the rows are used for the specific data, while the columns are used for the properties shared by all rows. Here the methods of loci work very well. We place a visualization defining the row and typically located in the first column in the middle of the first room. Often the first room is unique for that object. And then we progress along fix itinerary. The last room is usually reserved for comments, which are often placed as a mindmap on the wall. Quite often we use a color scheme, like different colors for odd and even rows.
The limitation of this approach is a lack of 2D navigation. For that purpose, you need to have separate mental structures for the rows and for the columns. This allows dual coding. Typically the relevant mental structures are placed in some mental landscape, like the first row and the first column of chessboard structure,
For sparse and not very large tables, mental chessboards are extremely comfortable. They can work up to 4 chessboards, where a chessboard is 8×8 squares. Within the relevant squares one can place mental palaces, often shaped as chess figures.
I have a separate article on mental chessboard construction. The best part of it: you can choose your itinerary. Notice that in chess not all squares are used. Using too many squares makes mental chessboards confusingly crowded.
For some multidimensional tables, the mental palaces can grow up into skyscrapers, with each floor serving as a separate mental palace. Chessboard structures can be nested, so that a chess square encodes an entirely new chessboard or a mental skyscraping. This nesting can be useful for pivot tables and some other multidimensional structures. Honestly, I never met such a table that required memorization, as we usually process them on our computers.
Many rows few columns template
The most useful tables often have very few colors and a lot of rows, like irregular verbs. They usually have only 3 forms, but hundreds of verbs to remember. Different languages have their own tables. You can compare the tables of irregular verbs in English and in Russian or in French.
The mental structure used is typically some mental palace with a very long itinerary and shallow depth. For example, you can put the visualization as an itinerary in a park, with 3 rows of trees as placeholders of 3 shapes. Or as many similar cabinets in an office building, with 4 corners of each cabinet used for each of the forms (the 4th for translation or notes).
If you need a bit more depth, you can use large offices. I personally use the Uffizi museum structure I learned in Florence,
Mendeleev himself visualized the periodical table as a large traveling bag. or organizer. Opening it we see a very special structure and color of shelves. Within each shelf, there is an object encoding an element. Take out the object and on it, you visualize attached on a label a set of properties – possibly as a very shallow mindmap.
This sort of curios is very useful with tables that we access a lot, and that have strange but still roughly tabular structures. Typically these tables are used in physics and chemistry. For example, subatomic particles like quarks have their own periodic table of 17 elements with colors and greek letters.
For very sparse multidimensional data we use tables of coordinate and value. The coordinate may often require 3 entries (XYZ) and the value 3 more entries (RGB or reflectivity/confidence/artifact). To be honest, such entities are usually not memorized, but you may need to visualize them. For example, you may want to remember the locations of key constellations in the night sky and the prominent stars within each constellation.
As the raw data becomes complex, typically it is best not to encode it beyond the level you need. Since the new data arrives constantly we often divide the space into segments. For example, we may remember some main constellations in the sky, and then add further secondary constellations with respect to the main constellations. The properties of each star are then stored in a separate mindmap, with anchor marker for the mindmap being the location of the star in the constellation.
I personally do not remember stars and galaxies, but I have many years of experience with Lidars.
I hope that by now I presented enough examples of tables most of us see every day. However, there are professional tables with very specific sturctures. For example, corporate accounting generates very specific accounting tables. Typically professionals do not need dedicated structures to remember this data. After sufficient exposure, they know their tables really well, like chess players know situations from different games. They use face recognition memory. Sometimes I call this “logical marker”. You need years of experience, but if this is your profession experience is gve. Use the information as-is, and you will be surprised how familiar its handling becomes after only two years.
Typically tables are either used as-is or encoded by combination of mindmaps and loci. Mental offices, chessboards and curios are specific variations of loci. If you are good with mental palaces, learning these additional techniques will be extremely easy. And yet you might not want to memorize tables at all, as data is usually processed via computers.