I spent the week prepping for finals. One is a year-long cumulative closed-book chemistry exam that I haven't had much time to practice for. I was worried about memorizing a few things:

  • Periodic trends and exceptions
  • The form and application of approximately 100 workhorse equations and various forms of measurement (molarity vs. molality vs. mole fraction).
  • Equations that get used rarely in homework or on exercises, but might be used as "gotchas" on the test.
  • Some concepts that I found either confusing, or so simple that I didn't bother to remember them the first time.

My anxiety wasn't just my ability to recall these ideas when prompted:

"What's the two-point form of the Clausius-Clapeyron Equation?"
ln(P2 / P1) = - Δ Hvap/R * (1/T2 - 1/T1)

Nor was I unable to perform the calculations.

My real concern was that I had spent the year treating my chemistry textbook like a reference manual, a repository for concepts and equations that I could look up when needed. I just memorized the few bits I'd need on any given quiz. Looking back at 1,000 pages of chemistry, I foresaw myself reviewing chapter 5 for a couple hours, but forgetting that review by the time I got to chapter 19.

The sheer volume of work that seemed to be involved in memorizing a textbook seemed unreasonable. I hate using Anki, and I spend far too much time in front of screens as it is.

So I decided to try something different - experimenting with the memory palace technique.

I perceive myself as having a poor visual imagination, but I've been trying to practice improving it lately, with some success. Gwern points to expert opinion that visual thinking ability might be second only to IQ in terms of intellectual importance. My experience is that when I'm using psychedelics, or deliberately practicing my visualization abilities, I do improve far beyond my perceived abilities. We're stuck with our IQ, but if it's possible to improve our visual thinking skills through practice in adulthood, that's important.

I want to describe my attempts and the outcome.

First Room

I tried this both with a single calculus textbook chapter, and my entire chemistry textbook. The results were similar but different. I'm going to focus on the chemistry palace here.

I close my eyes and allow myself to picture nothing, or whatever random nonsense comes to mind. No attempt to control.

Then I invite the concept of a room into mind. I don't picture it clearly. There's a vague sense, though, of imagining a space of some kind. I can vaguely see fleeting shadowy walls. I don't need to get everything crystal clear, though.

I mentally label the room as the "Ch. 14 room," or the "rates room." That means doing lots of things to make the label stick. I speak the words in my head. I picture a banner with them printed on it hanging from the ceiling. Or if I can't see it clearly, I picture a banner-like thing and just know that it says "rates room." I picture hourglasses sitting on furniture - the image comes to me much more easily than a banner with text.

I imagine the crucial equations sitting on columnar pedestals. Again, they are easier to picture for some reason. I make sure that I can visually see each piece of the equation. I imagine a label on the pedestal - one says "t1/2" for the half-life equations; the other says "Integrated rate law," with an hourglass made out of two intertwined integration signs.

I look up a picture of Svante Arrhenius and picture him in the room. He takes on a life of his own. I can tell he's proud of his equation, which appears in bold letters at the back of the room, with a sort of curtain around it. He's the keeper of the room. It takes on a calm atmosphere here. He's also the doorman. I have to tell him how to calculate the overall reaction order in order to enter. But if he knows that I know how to do it, I don't have to explain it in as much detail. We have a psychic relationship.

Second Room

Moving backwards to Ch. 13, I once again imagine a new room, the Solutions Room. Standing there, I can still see the entrance to the first room - I can even picture some of the things inside, from a distance. I start populating the room with symbols, objects, equations, and the chemists they're named after. They are happy to explain things to me as many times as necessary.

Abstract concepts that the book presents in words, still images, or equations get visualized in new ways. Partial pressures become two beakers, one with yellow steam and the other with red steam emerging. They get mixed into a single beaker that now emits a mixture of yellow and red steam, somewhere in between the amounts that the yellow and red beaker emit on their own. François-Marie Raoult is standing by to demonstrate his law to me. There's a bottle of Coke with Henry's Law printed on it.

The solubility rules are accessible when I glance at the periodic table on the wall. Rather than seeing a list of rules, I see the individual elements, which take on a life of their own. The alkali metals, ammonium, and nitrate zoom around the room, not interested in talking to anybody, on their own adventure. The halogens are too cool to talk to anybody except silver, mercury, and lead, who are immensely popular. Silver had a falling out with acetate, who's a communist and not interested in money. Be sensitive! Chromate is a rich chick in an expensive chrome-hubbed car cruising around, looking for a boyfriend. Sulfur is bicurious, so she'll bond not only with the transition metals but with astatine, arsenic, bismuth, and lead.

I practice traveling back and forth between the first and second rooms. They stay remarkably stable. Unlike recalling flash cards or the textbook, when I'm in my memory palace the ideas come almost unbidden. The elemental relationships I've used to conceptualize the solubility rules come bursting out of the periodic table.

Further rooms

I continue this for 6 chapters over the course of several hours. I am shocked and delighted at how easy and pleasant it is both to create the memory palace and to access the memories stored there. Not everything goes in - just the bits that I tend to forget. If I'm not sure about something, the famous chemists who populate the rooms will remind me, literally by talking me through their ideas.

The presence of the chemists is also helpful for keeping me focused. I suspect that my brain is recruiting my social motivation. If the only people in my environment are genius chemists who are delighted to keep me interested in chemistry, then why would I get distracted by the internet?

I find it deeply reassuring to stand in the Intermolecular Forces room and know that just by walking a few rooms over, I can get back to the Rates Room, where all the equations are stored. Perhaps I've built a path through the mental mountains? The next day, it's pretty easy to get back to the memory palace, and everything is as I left it. I just have to close my eyes and wait for a moment to get back in.

Concerns and questions

I also did a memory palace for calculus. I did it day-of because I felt more confident about calculus, it wasn't a comprehensive exam, and it was open book. I'll describe it another time. Mostly, it helped me feel more confident that I understood the breadth of the material. I found it much more convenient to refer to the textbook when necessary.

But for tomorrow's, I'm very glad that I now have a store of chemical facts in my memory palace. The anxiety that had been plaguing me this week has vanished. I'm not certain that it will really help. But I do anticipate continuing to use this technique in the future. I think it helps not only my memory but my synthesis of learning.

For example, our chapter on Lewis Structures also introduces the topic of electronegativity and formal charge. Anyone who's taken first year gen chem knows they're related: any negative formal charge should go on the most electronegative atom.

But when I would stare at the electronegativity pages in the textbook, I would focus on the rules offered there: the range of EN difference that characterizes a covalent vs. ionic bond, the periodic trend in EN, and how to calculate net dipole moment. Likewise, in the formal charge section, I would focus on how to calculate the charge.

It took seeing Linus Pauling holding a symbol for electronegativity in one hand, and a symbol for formal charge in the other, to more deeply understand that these are not just two different calculations to do. They're deeply related ways of modeling how molecules are structured. They go together like yeast and flour.

I also see how much faster and more intuitively I think about both chemistry and calculus when I can visualize them. It's just no comparison. Trying to remember Raoult's Law by remembering a verbal description or picturing the equation is just no comparison to looking at those yellow and red steaming beakers. Similarly, it's so helpful to picture a 3D mountain range and see a tiny little yellow gradient vector surfing up and down it on the steepest slopes.


I'm a true beginner here, so I don't want to make any grand claims about how to learn or how useful these techniques are. But I'd give a few pointers so far:

  • If you think you can't visualize, you might be wrong.
  • Start by just closing your eyes and allowing your brain to produce images without trying to control them. It seems important to have a relaxed, accepting attitude toward my own brain's way of visualizing.
  • The way to add control is to take a gentle, experimental attitude. Go with what's easy. Let yourself be surprised by what's do-able and what's useful. Is it hard to picture letters on a banner? Try visualizing Isaac Newton and ask him to say the thing you're trying to remember. Maybe you don't even need to do that - maybe it's enough to picture a vague stripe in the sky that you just know is a banner, and you just know has the words "rate room" printed on it.
  • It takes a while to navigate the rooms and get the information you need, so this might need refinement or practice if you've got to remember things quickly.

New to LessWrong?

New Comment
16 comments, sorted by Click to highlight new comments since: Today at 5:54 PM

Liked this post a lot. Tangentially, I'm interested in what other things you've been doing to train your visual imagination

Here are some additional benefits I'm discovering as I go along:

I am interested and able to memorize long lists of things that I might formerly have glanced at for two seconds, such as a table of common ligands.

Also, the several hours of visual practice I've put in so far is allowing me to visualize complex 3D molecules, such as Co(en)3^3+. Even now, my first impulsive response is to turn away from it, as the visual details seem too complex for me to make sense of. But instead, my visualization system is going so strong that I make the effort to picture it in my mind. I am able to not only replicate the 2D image of the molecule, but translate it into a 3D representation that I can inspect from different angles.

It's not the same as having a 3D graph of it on my computer that I can rotate and inspect with perfect fidelity. But the advantage is that I can pick out salient details of the 3D structure while holding their larger context in mind. For example, I can see the relative angles of the three ethylenediamene ligands and how they are rotated relative to each other, so that each molecule makes room for the others around the central metal ion.

After picturing the molecule in my mind, when I look at the molecule on the page, it looks different. Instead of a jumbled mass of colors and shapes, I can see the three-D structure, and it feels ordered and sensible.

This reminds me of Mary Carruthers's comparison of characteristics people emphasize when they talk about Einstein vs. Thomas Aquinas and the way in which memory palaces played an integral role in not just recall but composition for medieval thinkers. Unfortunately, it's too much text to copy, but I've taken screenshots of the relevant pages and uploaded them here.

Happened to look this post up again this morning and apparently it's review season, so here goes...

This post inspired me to play around with some very basic visualisation exercises last year. I didn't spend that long on it, but I think of myself as having a very weak visual imagination and this pushed me in the direction of thinking that I could improve this a good deal if I put the work in. It was also fascinating to surface some old visual memories.

I'd be intrigued to know if you've kept using these techniques since writing the post.

Not strictly as described, but my reliance on visualization has grown with time. I'm currently in a biomedical engineering program, so much of the subject matter is visual. In the past, I would tend to learn by remembering descriptions of things. This didn't work very well, either for understanding or memorization.

Now, I read much more slowly, and focus on building a picture in my mind as I go of what the book is describing. Usually, it's more of a cartoon or schematic, which is all that I need.

One framework for working memory is that it's composed of two distinct mechanisms: an audio loop and a visual sketchpad. My interpretation is that in the past, I relied almost exclusively on the audio loop, and almost not at all on the visual sketchpad. Now, I use the sketchpad much more than the audio loop. I think it improves my understanding of the material, and that this leads to more robust memory formation as well. I also will spend time, after reading a chunk of material, going over it several times in my mind from memory, to make sure that I understand the relationships between the ideas.

However, I no longer think it's a particularly sensible goal to try and build long-term, highly-detailed memories of an entire textbook, just for its own sake. Working in an engineering lab, that's just not how you move things forward. It's good to understand the textbook, know what's in it, and be able to use it as a resource for going deeper when necessary. However, choosing to memorize every detail in a textbook, for its own sake, now strikes me as folly.

So is it well worth developing your visualization abilities in service of creativity and the ability to learn and understand? With time, I feel more strongly that this has been, and will continue to be, a high-impact skill for me.

Is it worth memorizing a textbook? With time, I have come to feel like this is a bad idea.

Thanks for the reply! I also feel like I rely heavily on the audio loop currently, hoping I can boost the visual sketchpad side.

I have tried visualization mnemonic techniques (not the palace one. I don’t like palaces.), and they work decentish for biology-like stuff. They don’t work as well for mathematics/CS, because there either it is easier to “understand” the equations/rules, or it is very hard to visualize them. Of course, I do use visualization in trying to understand some math stuff, but it’s a thinking tool then, not a mnemonic. All in all, the big bug is that studying with visualization takes time, and it also can lower comprehension, as the visualization is an overhead in some contexts.

Beware that mnemonic-enhanced memories get forgotten, too. It’s slower, but still happens. If something is really needed in your longterm memory, spaced repetition is essential. Note that most everything are not needed in your longterm memory and you can sustain only a limited amount of spaced repetition. My heuristic is that anything that needs Anki most probably doesn’t belong in the longterm memory in the first place. Things that we really need, we use, and so we remember them.

I suggest memorization via ”upgrading.” That is, you learn A (for example, the concept of a 2D line). Then you and learn B that uses A internally (e.g., the concept of an ellipsis). This pushes A into your subconscious expertise, and grows your knowledge. Vanilla Anki-style repetition will not add anything to you, and you’ll start to hate that you need to grind just to stay the same.

I found the peg technique useful for chemistry. You decide on an image for numbers and elements, memorize them (e.g., via anki), and now you can visualize a lot of stuff easily.

Also, a word of caution. In navigating the educational system, it's hard to say whether you're trying to:

  • Feel confident that you've learned something.
  • Retain the material you learn.
  • Be able to apply a small subset of the material in new contexts.
  • Be fast an accurate at repeating the material on a timed exam.

All of these are valuable, but I'm not certain which of them a well-practiced memory palace technique is best for. My guess so far is that it could be useful for rapid recall, but that for a beginner, it's better for slow but broad retention.

I'm also not certain which of them is best to prioritize under what circumstances.

So it's not yet clear to me that the memory palace is a good strategy for learning. It's just an interesting approach that I'd neglected until now.

I like your idea of deliberately practicing visualization.

I found Brienne's posts here (https://agentyduck.blogspot.com/p/noticing.html) to be extremely helpful in practicing memorization.

Thanks for the recommendation! Your link formatting got messed up somehow. Here's a fixed link.

Have you considered trying SuperMemo instead of Anki? Anki is a pain to use because it gives you tools for memorizing but nothing else. SuperMemo works way better because it has tools for learning/processing/managing material then memorizing it, when it's reasonably learned.

Video I made on SM recently: SuperMemo for Wizards

I'll look into it! Thanks for the recommendation.

This is cool as all hell. How long ago did you do this? If you think of some way to test this, I'd be super curious to learn how much of this you can still remember in a month. I expect it to be pretty decent. I've never just... sat down and tried to do this for a big topic, and I might now.

This is just from the last three days. It also makes learning much more enjoyable. What would you rather do, read and re-read dry textbook writing about diffraction? Or stand in a room while Einstein himself shows you an epic, 3D simulation of wave diffraction at any speed or angle you like?

It's suddenly easy to get myself to read my textbook and I'm really happy about that. Best video game ever!

Great post.

Gwern points to expert opinion that visual thinking ability might be second only to IQ in terms of intellectual importance.

Do you happen to remember the source for this? A quick Google search didn't help.

EDIT: Found it (I think):

an emphasis on spatial reasoning & visualization ability was one of the reasons behind SMPY choosing to use SAT-M for screening, as one of the theses is that, after general intelligence, visuospatial reasoning (as opposed to the more academically-prized glibness & verbal ability) may be the next most important requirement for major STEM achievement.

Yep that's the one!