Problem memorising textbook


I’m having a problem that some of you have no doubt encountered before. I’ve started to memorise a textbook, but I’ve noticed that it isn’t as straightforward as just memorising a simple list of numbers =D

I’m getting on ok with it so far after reading this really useful thread, but before I go too far with my efforts I thought I’d ask in here for some clarification.

My approach within each chapter is to read through it and create images for the key concepts/points that are made. This is working quite well but these images are quite complicated, given the complexity of the information they represent. By complexity I mean that they represent a range of different types of information - names, dates, contributions to the field (within which there will probably also be other numbers still), maybe a list of other associated information as well. As I say, it’s ok but in my mind it feels quite messy like this. For example so far I have no way to distinguish between dates and the other images involved, so I just have to make an educated guess based on the fact that the particular image is part of my major system…

Does anyone have suggestions for maybe organising each point/locus? Does anyone have a particular approach to dealing with dates, for example I’m often encountering 18xx and 19xx, but I’m not very creative in finding a method for dealing with this except for simply using my major system. That feels a little bit overkill for dates because the same images are appearing for 18 and 19 obviously!


Hi, AnInsect

You basically have two problems: (i) image complexity; and (ii) representation of dates.

For the first, while there will never be a perfect solution, I suggest decomposing complexity by using a hierarchical approach. By using “triple links”, you can mentally represent any arbitrary tree as a binary tree and keep all information organized (I’ve just talked about it on this thread, if you care to read). Higher-level images are not just mnemonic representations of headings in a textbook but they actually summarize the information of the level below, thus assisting in the memorization process per se. By “triple link” I mean that a given image must be able to connect to a right-sibling, i.e., another image in the same hierarchical level, and to a first-child, i.e., the first image one level below. There can be a multitude of methods to do so. I tend to make right-sibling links by “seeing” a story unfold in space. This is not the same thing as artificial memory palaces, but I do like to place the images in a geographical area, even if a very approximate, blurred, or fictitious one. First-child links I create by going “inside” the image or doing other crazy associations that don’t extend across space. This is only approximate – method can only go so far in madness – but, inserted into a spaced repetition regime, it has worked well so far.

With respect to your second problem, I suggest you rely a little on your natural memory and understanding of the subject-matter when possible. For instance, I have just memorized a small book on the Enlightenment. While the philosophers mentioned span from 1500s to 1800s, the subject-matter itself and the order the book is laid out help me easily guess the “16xx and 17xx” I encounter.

That said, I do have a “system” to remember approximate birth/death dates of important historical figures. Note that I said “approximate”. I find it important to situate people in time, but usually, I don’t care about exact dates. What I do is that I use one single image of my 00-99 number system, and that can be either the Person or the Object (I use a PO system) or, sometimes, just a small attribute of either the P or the O. If, for instance, I want to memorize that Voltaire lived from 1694 to 1778, I use the image for 97, meaning that he was born in the 90s of a given century and died in the 70s of another. Since I know he was a key figure of the Enlightenment and that he could not have possibly lived more than 180 years, I can have a pretty good notion of his lifespan.

Now, when precise dates are needed, I see no other option than using two 00-99 images or the like. About it being hard to discern the sub-images within a given locus as to which is a number and which is not, unfortunately, I believe you will either always need to put some effort into decoding it (which is greatly minimized by spaced repetition) or you will have to keep a logical organization of your images (which, for me, is not desirable). I have a long thread on this forum about trying to conjure images that convey exact sentence structures by their visual arrangement, but I have long abandoned that project. I love order and method, but mnemonics is greatly hampered by too much of it, in my opinion.

I hope that might provide you some insights.