The 3 Keys To Using Memory Cues
Use The Power of Load, Latency, and Familiarity For Better Learning Outcomes
Table of Contents
The 3 Properties of Cues: Load, Latency, and Familiarity
3.1 Load
3.2 Latency
3.3 Familiarity
Improve Your Memory By Using These 3 Properties
On Reactive Memory
My memory practice is fairly laidback.
I don’t try to memorize thousands and thousands of items. I’ve done that before, and it sucked. Instead, I focus on a few key points, usually “structural” markers like book titles and chapters, and then I reconstruct the book’s arguments by harnessing my reactive memory, the term I use for what the scholastics would have called intentio, which is sort of similar to what modern psychologists sometimes call spontaneous retrieval.
This reactive approach to memorizing doesn’t involve a lot of upkeep and the set-up phase is also minimal.
It is tricky learning how to do it in the first place, but like riding a bicycle, once you have it down, it becomes second nature.
The Limits of Reactive Memory
Reactive memory has certain limitations. In order to understand those limitations, you need to have an idea of how the cue-response loop works.
A cue is a trigger that evokes a response, or a memory (habits are also a kind of memory). All cues have responses. It’s a yin-yang thing. You can’t have one without the other.
Mnemonics are just cues memory practitioners create in order to evoke the things they want to remember.
Memory practitioners make a mental painting (a cue), and they place that painting in a memory palace (a meta-painting, a cue that collects other cues), and that mental painting evokes the idea they want to remember. In other words, the memory palace evokes their memory of the mental painting, and the the mental painting evokes the memory of whatever they want to recall.
The cue-response loop is a biological process that requires energy. There’s no free lunch in biology. There’s a physical cost the body pays each time a cue evokes a response. It’s a small cost, but it’s a real one, and it sets limits on what you can expect from your memory practice.
In this essay, we’re going to cover 3 properties that cues have that are important to master in order to create strong reactive memories. In other words, if you want a cue to effortlessly evoke a memory so you don’t have to engage in effortful retrieval, you’ll want to get a grip on these properties.
The 3 Properties of Cues: Load, Latency, and Familiarity
Cue Load
The ancients had all kinds of spacing rules, like how far mental images should be from each other, and those rules reflect something important about the limits of cueing.
In modern psychology, they talk about cue overload. Cues can interfere with each other if you load too many of them at once. Think of it this way: if you plug in a hundred electronic devices into a single outlet, you’ll blow a fuse. In the same way, there’s only so many cues you can activate before they start shorting each other out.
Cue latency
Cue responses degrade with each association retrieved. It’s like running, the brain’s cue-response loop loses steam the more you use it, just like your body gets out of breath the longer you run.
The fan-effect helps explain some of this cue-response degradation. The fan is the number of associations a cue has. The larger the fan, the slower the retrieval. That’s not something you can really get around, and it’s one reason older people have higher retrieval latency. They have decades more associations to handle, which slows things down.
However, it’s not just the total number of associations a cue has that impacts latency, but also the number of associations that get activated. The more you activate, the higher the cost. Simply put, if I ask you to recall a hundred memories that feature bananas, the first ten memories you come up with will come to you a lot faster than the last ten.
Low latency means fast retrieval, and my method is all about making retrieval spontanneous. That’s the whole beauty of “intentio”, it’s a natural response to a cue that doesn’t require effortful retrieval.
Cue Familiarity
Cue familiarity determines if the response a cue evokes is general or specific. The study, “Spontaneous memory retrieval varies based on familiarity with a spatial context” covers a bit of what I mean… sorta. We’re not really saying the same thing, but there’s a bit of overlap. Anyways, in my use case, a response is general if it has a lot of associations, and specific if it doesn’t.
General responses reflect fluency, while specific response reflect isolated islands of experience that are less embedded in the larger fabric of our understanding of the world.
General responses give us knowledge, specific responses give us data. It’s the difference between cueing the forest and cueing the trees. Both matter, but the forest is where the magic happens.
When I read a book, I want to turn its contents into general knowledge, not have it siloed off in the land of episodic memory. I think this is one reason I’m not a huge fan of flashcards, which is about creating specific prompts that evoke specific responses. Flashcards are only useful when they’re a stepping stone to embedding those specific responses into a more general context.
Improve Your Memory By Using These 3 Properties
Familiarity, latency, and loading, are three important properties of memory cues. In mnemonics, all we really do create cues to evoke memories, so it’s important to understand how these forces impact our own memory practice.
Use Specific Mnemonic Cues Sparingly
Years ago, I put in a lot of effort into building a useless “antipalace”, a palace that contained thousands and thousands of facts that I never engaged. My experience with the antipalace convinced me to try the opposite approach: creating much smaller memory diners where each mnemonic cue evoked multiple associations instead of a single association.
It didn’t take long for me to realize that my reading comprehension improved significantly with this new approach.
And that makes perfect sense if you think of it in terms of cueing specificity.
In my old “memorize all the facts!” approach, I focused on creating specific cues for specific facts, while in my more modest “memorize just a few big important things” approach, I focused on creating general cues that evoked my general understanding of what I was reading.
This is worth repeating: general cues evoke general knowledge, specific cues evoke specific facts.
A title or a chapter heading is a general cue, and it should evoke your understanding of the chapter’s contents.
Specific cues have far fewer associations. These can be useful under certain contexts, but they should not be the majority of your mnemonic cues.
In your mnemonic practice, you should only create specific cues if you have a specific purpose in mind for whatever they evoke. You might want to create mnemonic cues for each biblical verse, the way many scholastics did, however you don’t need to go that far for most of what you read unless you intend on approaching that book the way medieval monks might approach the bible.
Use specific cues for specific purposes, use general cues for everything else.
Beware The Fan-Effect
The fan-effect places limits on how many associations (or memories) you can easily retrieve from a single mnemonic cue.
I think mnemonic cues should be used sparingly, because the smaller your memory palace, the more likely you are to actually sit with whatever you put in it. My palaces have fewer cues, but those cues have more associations. The pros outweigh the cons, but the cons are very real.
Cue-Response retrieval degrades as the number of associations increases. I don’t have much trouble with the fan-effect when reading popular books written for a general audience, but it kicks my ass whenever I try to use a simple “titles only” memory palace for a heavier book.
Books that cover more territory require more cues.
How many cues you create for these more complicated works is going to depend on the book itself and what you want to get out of it. I usually either use the author’s subheading as additional cues in denser works, or create my own subheadings if none exist. Creating subheadings is just grouping pages by topics, which you can then assign mnemonics cues.
Avoid Interference
Space your cues in your memory palace so that they don’t interfere with each other. How that might look is going to be different for everyone.
Most of my palaces are small so spacing isn’t much of an issue, but that’s something I learned how to do from experience.
I came up with the dinner party rule for my memory palaces, and it’s pretty straight forward: if your palace was a real event in a real location, would you be able to comfortably enjoy a conversation with each and every “person” (i.e: mental painting) there?
I don’t want my palace to feel like a rave or a mosh pit. I want them to be cozy, relaxed environments with plenty of breathing room.
This rule isn’t hard and fast, but it works for me and so it might be a decent heuristic for other people. Whatever the case, you should get a feel for how much breathing space you need between cues to prevent interference.
Wrapping Things Up
You should be mindful of the latency, load, and familiarity of your mnemonic cues designing your memory palaces. Aim for low latency, low interference, and high familiarity.
Focus on creating cues that play with those 3 properties to get a feel for what works and what doesn’t.
I also highly recommend creating bad cues deliberately (i.e: experiment with high latency, high interference, and low familiarity) just to learn first hand they don’t work.
In future essays, I’ll share some of my “do bad things on purpose” experiments. I once built a palace using randomly generated cues that had zero to do with what they were meant to evoke. It was basically pavlovian mnemonic cues. That was an experiment that would make a lot of memory artists angry.
The point is, you shouldn’t just take my word for what makes a cue bad, you should experience it for yourself. Try creating different cues with different properties to get a feel for what works and what doesn’t.
I think my opinion on low-familiarity cues is the one that would be the most controversial among veteran mnemonists, so that’s the big one you might want to focus your experiments on. Create palaces where you memorize All The Things in a book, then create palaces where you only use general cues, then experiment with mixing both specific and general cues together, and see what works for you.