Generate many new loci with ease: The Layered Nook and Cranny Method



Hi everyone,

I wanted to share a method I’ve been testing the last couple weeks with a good deal of success. You’re likely familiar with the Nook and Cranny Method of generating 10 loci per room. The Nook and Cranny Method works by turning the four corners and four walls of a room into loci, then adding the floor and ceiling, yielding a nice round collection of 10 loci. Building from that technique I have been able to generate 26 loci per room using what I think of as a Layered Nook and Cranny Method.

The Layered Nook and Cranny Method works by identifying a column of three loci in each of the four corners and along each of the four walls. When the floor and ceiling are added, this results in 26 loci. This is just like the Nook and Cranny Method, except that there will be a top layer to the room, a middle layer to the room, and a bottom layer to the room.

For example, along the left wall there might be (1) a lamp at the top of the left wall, (2) a mirror in the middle of the left wall, and (3) a dollhouse at the bottom of the left wall. These three loci now serve as a replacement for what would have been only a single loci (likely the mirror) using the original Nook and Cranny Method.

I don’t use the corners and walls themselves. Rather, I find distinctive features of the room that would serve as good loci (e.g., art, furniture, etc.). In this way I have been able to systematically pull loci from rooms in a predictable way rather than just hoping a given room will yield sufficient loci for me to use. When I mentally move through the loci I do so from top to bottom of each column.

One of the added benefits is that not every room is limited to just four corners and four walls. Some rooms have extra protrusions, creating new corners. Or some walls, especially in larger rooms, are particularly “busy” with objects and allow me to identify more than one column of loci. This means it is not uncommon to walk out of a room after having identified 29 or 35 new loci.

Having 26 loci might sound like a strange number, especially compared to the original Nook and Cranny Method’s nice even number of 10 loci. However, considering that many card systems require 26 loci (like PA) and that only two rooms in succession house 52 loci (perfect for one card per loci), I’ve found it to be very convenient.

I’m sure this technique is not original with me. No doubt that many people using the Nook and Cranny Method or just identifying raw loci have stumbled upon this method themselves. But I appreciate the simplicity and pragmatism of the Nook and Cranny Method and have found this expansion on it to be incredibly useful for my own memory work. Now after having put the technique to the test over time I have no reservations in sharing it. Perhaps others looking to build their first memory palaces will find it useful.

(Josh Cohen) #2

I will give that a try. :slight_smile:

I wonder if one could create pre-designed grid patterns and then apply them to room sections.

Section: one of the 10 areas of the room, always in a certain order:

  1. near left corner
  2. left wall
  3. far left corner
  4. far wall
  5. far right corner
  6. right wall
  7. near right corner
  8. entrance (where you are standing)
  9. ceiling
  10. floor

Then, any of the preset grid overlays could be applied to each section:

Overlay one:
Choose one location in that section.

Overlay two:
Choose three locations in that section, like mentioned above in your post.

Overlay three:
Suitable for windows: apply any of the overlays to the window or door to get additional locations.

Overlay four:
For two-dimensional objects, like windows: four corners, four edges, and a surface. Additional overlays could be applied to the surface, if it’s interesting enough.

Additional overlays:
Divide up the section in other ways…


  1. left near corner: overlay one, since that corner is bland and you'll only put one image there
  2. left wall: there is a window there, so you could apply overlay #2 to the wall, making three locations -- and there is also a window there, which makes up the 2nd location, so you could apply overlay #4 just to that window. Total locations in the left wall would then be 11 (top of left wall, window with 9 locations due to overlay #4, and bottom of left wall)

Each overlay type could have its own mnemonic image so that you can remember which type of overlay has been applied and how many images you expect to find there.


so what you are saying is you take 3 objects at each of the locations such as the wall or the corner and then use each of those objects as a loci? Well what if there are less than 3 objects? What if there are more? Do you make up those objects?