I was told ages ago that humans divide our spatial memory into discrete spaces, and then store our memories of objects relative to the space in which we saw or thought about the object. That's why when you need to get something from another room, and you walk through a doorway to get there, so often you'll realise you've forgotten what it was that you needed to get. And then you walk back through the doorway and you'll remember.
I'd guess this is also why memory palaces work - you're basically allocating a data structure to store your facts in.
It's called the doorway effect and I also believed it to be established science already. Perhaps this study drills more into the details and particulars of the effect.
I would recommend Kevin Horsley's Unlimited Memory[0] over Foer's Moonwalking with Einstein.
Foer is a lot of hype and unfocused, whereas Horsley almost immediately gets down to brass tacks after a brief introduction meant to energize you, and provides a great general overview that leaves you with the knowledge of how to continually improve beyond what the book provides.
You can almost think of memory palaces more as maps or pointers than the data structures themselves. You're piggybacking off of existing neural connections in your memory circuits and placing new connections in between or at the ends, and activating those new synaptic connections by activating the ones you're already familiar with which are connected to them.
The reason we use memory palaces specifically is because our visual memory is incredibly strong. Many techniques take this a step further and add an element of time, allowing you to make use of your chronological memory as well, filling in the gaps between "scenes". It works exceedingly well, I can memorize a pack of cards in about 2-3 minutes with this technique.
There’s a psychotherapy approach called Internal Family Systems, IFS, that is the theory that we’re all naturally made up of individual, multiple personalities. Each part of you having it’s own anxieties, fears, goals.
It’s why you could seemingly be one personality when you encounter love, hungry, horny, etc and seemingly different person while in a different context.
Is this psychotherapy approach generally respected among therapists, or is it kinda fringe?
Because from that description, it sounds like it would probably be a fringe thing, but I don't know the field in the slightest, so I can't be confident in that.
Also, that description at the end of the post, that you say it explains, doesn't seem to match my experience.
Also, I don't think I see how it relates to the parent comment?
I don't work in neuroscience, but I've taken a few courses in undergrad. I'll chime in until a professional can answer your question more definitively.
There's some evidence that the left and right brain hemispheres have their own desires and goals, based on some studies conducted by Roger Perry in the 1960's. Sperry received the 1981 Nobel Prize in Physiology or Medicine for his split-brain research.
Aside from the left-right split, I think there's a large category of well-supported models that conceive of the mind-brain as modular, but with sub-units dedicated to basic processing tasks (memory retrieval, sensory integration, etc.) rather than being fully-formed, discrete "personalities" in themselves.
Never pushed the investigation but I had managed to remember several poems and songs verbatim using this method. Something that I otherwise really struggle with.
Memory palace works essentially by storing arbitrary things in your spatial memory. Most people have well developed spatial memory, as there is no practical method that could replace it. It's much better long term if you actually push yourself and train to use your textual or whatever memory properly, instead of using your spatial memory for storing things that don't belong there.
I have a very good memory for places. But I am terrible at remembering where places like relative to each other.
I grew up on a coastal town, where essentially every 'area' lay on a single highway.
And after living there for ~15 years on and off, I could name almost every business in three towns, which ones had shut down and re-opened as something different, what had been swamp a few years before.
Everything.
But if I got in my car, and tried to drive to a specific location, I was completely useless.
I could tell you what the building looked like, what was accross the street, what was next to it on either side.
But I woukd be completely unable to tell you if it was north or south of me.
I often had to navigate by mentally walking from building to building.
That's weird. For me it's the other way around. I can't remember things if I can't put them in the map. Watching places of my hometown in local TV, I have a bad time identifying them, even if they're very familiar, because I see them "out of context".
> I often had to navigate by mentally walking from building to building.
That's similar to how I navigate if I have to find out if there's a certain kind of business in proximity. I quickly "fly" down the street in one direction, see if there's what I need – say, a convenience store – for about a bus stop or two, then the other way. I'm looking in the direction I'm scouting, but my mind is wandering way further.
I'm also quite capable of finding my way back to somewhere I'd been once, but I won't be able to tell you where it is on the map unless I find something memorable on it and orient myself similarly to above.
"We find it easy to remember the position of many items as one unit when arranged in large open spaces. Hence, large corridors, roads, and entrance areas that provide a broad overall view enhance wayfinding."
I think this is something that web design has lost. Most modern web design is optimised for the mobile and as a result, navigation and menus are hidden behind hamburger menus, swipe left menus, or behind mysterious icons. Instead of a broad overall view, the user has to enter each corridor and room to find out what is in there. Where once websites were about having a broad overall view of the site perhaps now it's about getting the focus on a piece of content on it's own.
I suppose another example that comes to mind is API Documentation on one page, versus split up into many little pages in a structured hierarchy.
Looks like they put people in two different types of VR environments. One was an open room ("vista space"), the other was more like a simple maze with a few corridors ("environmental space"). There were various objects throughout the environments that the subjects were asked to memorize and then "point" towards using a joystick.
The subjects were allowed to learn the locations of the objects in various ways such as from a few vantage points, or moving around along various paths through the space, whether the objects appeared one at a time or all at once, etc. They then measured how long it took the subjects to complete the "pointing" (latency in seconds) and the error in direction (degrees).
They predicted that in the open room condition, subjects would perform better when the object was 0, +/- 90, and 180 degrees relative to the body orientation, but for the corridor case performance should be best when the object was +/-45 or +/-135 degrees from the body orientation. I didn't figure out why they predicted this.
They found that people in the "corridor" condition took longer to point at the object and made more errors, and this increased the more walls that were in between the subject and the item. They also found that the performance was largely as predicted according to the relative prediction of the object (better at +/- 45/135 in the corridor condition, and vice versa for the open room condition).
They conclude: "memory differences between vista and environmental space originated mainly from the spatial compartmentalization which was unique to environmental space learning."
Seems like a pretty flimsy connection between data and theory to me. If anyone can figure out why they made that prediction about body orientation please comment.
25 comments
[ 0.28 ms ] story [ 24.0 ms ] threadI'd guess this is also why memory palaces work - you're basically allocating a data structure to store your facts in.
Foer is a lot of hype and unfocused, whereas Horsley almost immediately gets down to brass tacks after a brief introduction meant to energize you, and provides a great general overview that leaves you with the knowledge of how to continually improve beyond what the book provides.
[0] https://www.amazon.com/Unlimited-Memory-Advanced-Strategies-...
The reason we use memory palaces specifically is because our visual memory is incredibly strong. Many techniques take this a step further and add an element of time, allowing you to make use of your chronological memory as well, filling in the gaps between "scenes". It works exceedingly well, I can memorize a pack of cards in about 2-3 minutes with this technique.
It’s why you could seemingly be one personality when you encounter love, hungry, horny, etc and seemingly different person while in a different context.
https://youtu.be/LuJLv98ks-I
Because from that description, it sounds like it would probably be a fringe thing, but I don't know the field in the slightest, so I can't be confident in that.
Also, that description at the end of the post, that you say it explains, doesn't seem to match my experience.
Also, I don't think I see how it relates to the parent comment?
There's some evidence that the left and right brain hemispheres have their own desires and goals, based on some studies conducted by Roger Perry in the 1960's. Sperry received the 1981 Nobel Prize in Physiology or Medicine for his split-brain research.
https://embryo.asu.edu/pages/roger-sperrys-split-brain-exper...
Aside from the left-right split, I think there's a large category of well-supported models that conceive of the mind-brain as modular, but with sub-units dedicated to basic processing tasks (memory retrieval, sensory integration, etc.) rather than being fully-formed, discrete "personalities" in themselves.
https://en.wikipedia.org/wiki/Method_of_loci
Well, the photo at the top of the article shows them wearing an Oculus.
https://en.m.wikipedia.org/wiki/Sheaf_(mathematics)
I have a very good memory for places. But I am terrible at remembering where places like relative to each other.
I grew up on a coastal town, where essentially every 'area' lay on a single highway. And after living there for ~15 years on and off, I could name almost every business in three towns, which ones had shut down and re-opened as something different, what had been swamp a few years before.
Everything.
But if I got in my car, and tried to drive to a specific location, I was completely useless.
I could tell you what the building looked like, what was accross the street, what was next to it on either side.
But I woukd be completely unable to tell you if it was north or south of me.
I often had to navigate by mentally walking from building to building.
That's similar to how I navigate if I have to find out if there's a certain kind of business in proximity. I quickly "fly" down the street in one direction, see if there's what I need – say, a convenience store – for about a bus stop or two, then the other way. I'm looking in the direction I'm scouting, but my mind is wandering way further.
I'm also quite capable of finding my way back to somewhere I'd been once, but I won't be able to tell you where it is on the map unless I find something memorable on it and orient myself similarly to above.
I think this is something that web design has lost. Most modern web design is optimised for the mobile and as a result, navigation and menus are hidden behind hamburger menus, swipe left menus, or behind mysterious icons. Instead of a broad overall view, the user has to enter each corridor and room to find out what is in there. Where once websites were about having a broad overall view of the site perhaps now it's about getting the focus on a piece of content on it's own.
I suppose another example that comes to mind is API Documentation on one page, versus split up into many little pages in a structured hierarchy.
Looks like they put people in two different types of VR environments. One was an open room ("vista space"), the other was more like a simple maze with a few corridors ("environmental space"). There were various objects throughout the environments that the subjects were asked to memorize and then "point" towards using a joystick.
The subjects were allowed to learn the locations of the objects in various ways such as from a few vantage points, or moving around along various paths through the space, whether the objects appeared one at a time or all at once, etc. They then measured how long it took the subjects to complete the "pointing" (latency in seconds) and the error in direction (degrees).
They predicted that in the open room condition, subjects would perform better when the object was 0, +/- 90, and 180 degrees relative to the body orientation, but for the corridor case performance should be best when the object was +/-45 or +/-135 degrees from the body orientation. I didn't figure out why they predicted this.
They found that people in the "corridor" condition took longer to point at the object and made more errors, and this increased the more walls that were in between the subject and the item. They also found that the performance was largely as predicted according to the relative prediction of the object (better at +/- 45/135 in the corridor condition, and vice versa for the open room condition).
They conclude: "memory differences between vista and environmental space originated mainly from the spatial compartmentalization which was unique to environmental space learning."
Seems like a pretty flimsy connection between data and theory to me. If anyone can figure out why they made that prediction about body orientation please comment.