Memory palaces are an ancient technique, rumored to be a powerful way to improve memory. Yet what a strange concept. It’s hard enough to remember the facts you want to recall. Why would placing the added cognitive burden of constructing and remembering an memory palace assist with memory?
I think the memory palace is an attempt to deal with the familiar sensation of knowing you know a fact, but not being able to summon that fact to your conscious mind. Later in the day, by chance, it pops into your mind spontaneously. Just as you suspected, the memory was there, somewhere—you just didn’t know how to find it.
My thesis is that the utility of a memory palace primarily that it gives you a structure for controlling which linkages between arbitrary memories, and a way to deliberately navigate between them. Consider your childhood home, or any other location with which you’re intimately familiar. No matter how you represent it in your mind—as a kinetic sense of spatial awareness, as a vivid 3D “memory reality,” or some other way—I bet that it’s easy for you to navigate it, deciding where you are in the room, what you’re facing, what angle you’re at, and so on.
Personally, my mental representation of my childhood home is not nearly as vivid as actually being there. It’s also not up-to-date: the dining room oscillates between the old and new chairs, the kitchen has the old and new linoleum, the apple tree we’ve long since cut down is visible through the back window. However, the basement is the new renovation, not the old dark space full of spiderwebs, and my parents’ current cars are parked out in front.
I can float through the house rapidly, change direction, examine things, and so on. It’s a bit like playing a video game that doesn’t simulate gravity and doesn’t block you from walking through walls.
I can even do this, to some limited degree, for locations I’ve been in only once, even if I wasn’t paying careful attention to my surroundings at the time.
There seems to be a mental faculty for remembering and navigating locations that is more powerful than the ability to remember discrete bits of information.
Yet I find it very difficult to take abstract bits of intellectual information—mathematical equations, arguments, and so on—and represent them in any sort of visual form. It’s much easier to access these bits of knowledge directly. Trying to place them around my childhood home is a lot of extra work, for no clear payoff—and where would I put all those equations anyway? The memory house would rapidly become cluttered.
But I like the core idea of the memory palace—the idea that we should deliberately construct and practice navigating between our thoughts, not just treat individual facts and memories like atomic objects out of context or relationship. We should try to make the technical knowledge we work so hard to understand accessible and navigable, whether or not we’re sitting down with a textbook or practice test and studying it actively.
This is difficult, of course, because if we have a strong memory of a particular set of facts, there’s little point in going over them more. What’s needed is practice remembering what doesn’t spring readily to mind, and here we face all kinds of challenges: we might remember incorrectly, we might not be able to conjur up the facts we want, we might get distracted, we might waste time.
However, I’ve reaped great gains in my ability to control my mental imagery by allowing that imagery a lot of freedom to “manifest itself.” Rather than trying to precisely control the exact image my mind “paints” on its inner canvas, I give it a suggestion and let it do the work on its own, much like giving a prompt to DALL-E. The results have improved with time.
I think a similar technique is useful for remembering facts, arguments, and other pieces of left-brain STEM knowledge. Giving my brain a suggestion of the sort of facts I’d like it to recall tends to lead to productive results. If I’m preparing for an exam, suggesting the content of a certain chapter tends to lead to my brain recalling more facts than I knew that I knew.
As more facts come to the top of my mind, I gain an ability to reconcile them, relate them, and recall more along that theme. At the beginning, I only had enough facts at the top of my mind to request my brain remember “facts about torque.” But as those facts come trickling in, I’m able to make requests for ever more detailed information.
These are early days, and I’m not certain about whether this is a reliable and useful effect for practical purposes. But it points in the direction I’m interested in, and I think an especially nice feature is that this is precisely the sort of mental practice that is especially hard for external tools to replicate, yet might offer outsize returns to deliberate practice.
Explaining the Memory Palace
Memory palaces are an ancient technique, rumored to be a powerful way to improve memory. Yet what a strange concept. It’s hard enough to remember the facts you want to recall. Why would placing the added cognitive burden of constructing and remembering an memory palace assist with memory?
I think the memory palace is an attempt to deal with the familiar sensation of knowing you know a fact, but not being able to summon that fact to your conscious mind. Later in the day, by chance, it pops into your mind spontaneously. Just as you suspected, the memory was there, somewhere—you just didn’t know how to find it.
My thesis is that the utility of a memory palace primarily that it gives you a structure for controlling which linkages between arbitrary memories, and a way to deliberately navigate between them. Consider your childhood home, or any other location with which you’re intimately familiar. No matter how you represent it in your mind—as a kinetic sense of spatial awareness, as a vivid 3D “memory reality,” or some other way—I bet that it’s easy for you to navigate it, deciding where you are in the room, what you’re facing, what angle you’re at, and so on.
Personally, my mental representation of my childhood home is not nearly as vivid as actually being there. It’s also not up-to-date: the dining room oscillates between the old and new chairs, the kitchen has the old and new linoleum, the apple tree we’ve long since cut down is visible through the back window. However, the basement is the new renovation, not the old dark space full of spiderwebs, and my parents’ current cars are parked out in front.
I can float through the house rapidly, change direction, examine things, and so on. It’s a bit like playing a video game that doesn’t simulate gravity and doesn’t block you from walking through walls.
I can even do this, to some limited degree, for locations I’ve been in only once, even if I wasn’t paying careful attention to my surroundings at the time.
There seems to be a mental faculty for remembering and navigating locations that is more powerful than the ability to remember discrete bits of information.
Yet I find it very difficult to take abstract bits of intellectual information—mathematical equations, arguments, and so on—and represent them in any sort of visual form. It’s much easier to access these bits of knowledge directly. Trying to place them around my childhood home is a lot of extra work, for no clear payoff—and where would I put all those equations anyway? The memory house would rapidly become cluttered.
But I like the core idea of the memory palace—the idea that we should deliberately construct and practice navigating between our thoughts, not just treat individual facts and memories like atomic objects out of context or relationship. We should try to make the technical knowledge we work so hard to understand accessible and navigable, whether or not we’re sitting down with a textbook or practice test and studying it actively.
This is difficult, of course, because if we have a strong memory of a particular set of facts, there’s little point in going over them more. What’s needed is practice remembering what doesn’t spring readily to mind, and here we face all kinds of challenges: we might remember incorrectly, we might not be able to conjur up the facts we want, we might get distracted, we might waste time.
However, I’ve reaped great gains in my ability to control my mental imagery by allowing that imagery a lot of freedom to “manifest itself.” Rather than trying to precisely control the exact image my mind “paints” on its inner canvas, I give it a suggestion and let it do the work on its own, much like giving a prompt to DALL-E. The results have improved with time.
I think a similar technique is useful for remembering facts, arguments, and other pieces of left-brain STEM knowledge. Giving my brain a suggestion of the sort of facts I’d like it to recall tends to lead to productive results. If I’m preparing for an exam, suggesting the content of a certain chapter tends to lead to my brain recalling more facts than I knew that I knew.
As more facts come to the top of my mind, I gain an ability to reconcile them, relate them, and recall more along that theme. At the beginning, I only had enough facts at the top of my mind to request my brain remember “facts about torque.” But as those facts come trickling in, I’m able to make requests for ever more detailed information.
These are early days, and I’m not certain about whether this is a reliable and useful effect for practical purposes. But it points in the direction I’m interested in, and I think an especially nice feature is that this is precisely the sort of mental practice that is especially hard for external tools to replicate, yet might offer outsize returns to deliberate practice.