Loci vs Linking?

When I read about memory techniques most people seem to use more loci, and less linking.
For example many put 1-3 objects at every lokus and link them together, and everywhere I look people seem to use that system.

My question is if loci really is better than, for example, linking objects in a chain?

The extreme case of loci is when you put one object at each lokus, requiring many memory places and settings, and the extreme case of linking is that you make a list of all objects and then have to create links between every object.

Personally I use kind of a hybrid when I memorize. I start with a familiar place and then start linking the objects together while traveling through the room. Often i follow the walls or have kind of a system on how to move through a room, but I haven’t created any memory places, instead I let the list interact with whatever is currently close.
Often I choose about 50-100 objects in every list, and after that I choose another room or place for the next list.

For me this method is a lot faster and requires less preparation (I don’t have to create any journeys) and I feel more creative without being bound to a journey.

What do you think? What are the pros and cons with the journey system and using linking?

I reckon it depends on time constraints and what you are using it for.

At the moment i am loving linking method instead of loci/memory palaces because i need a lot of information memorized quickly for a short period of time. When i say a lot, i mean a lot. Memory palaces and others take ages to prepare but they are more suited for long-term, not that they are no good for short-term, linking is easier if you are in a rush like me.

I’m linking everything from a huge book and creating my own little stories that are far more digestible to my mind. Anyway gotta dash!

loci vs linking? I think it is down to personal preference. Jonathon Hancock won the WMC in 1994 using the linking method.
One disadvantage of linking is that it is much more difficult to locate a certain piece of information at a certain point e.g. if you needed to randomly know the nth piece of information you would have to run through your links counting them until you reached the nth place whereas with the loci method you should more or less go straight to the nth place.

I decided to try using loci instead and it seems that method works a lot better for me than linking. Last time I tried the result was the opposite. With linking my record is 4:50 for a deck of cards, and with loci I’ve managed in 2:50.
So I’ll develop my loci system now and see if I can get it even faster.

Though I think linking is better (at least for me) for large amounts of information that you want to remember for a long time. Linking takes longer time for me, but creates more vivid scenes, and I also don’t need any pre-defined lokus to associate the images with.
So for my 10,000+ PI decimal project I will use linking instead of loci.

I think that I tested Loci in a different way, the location itself or/and the objects (around me) become a binding agent between the subject and the object. Not sure if this would work in the long term.
I just put the word, phrase or the number on/in a Cabinet, sofa…etc. sometimes imagine it in large print. This would do the job for me, at least for now! The rest is just reviewing my subject in the same order.

Linking is only reliable when you do not have any duplicates. For example, you can use linking to reliably remember a single deck of cards, but linking would not work to remember 30 decks of cards. Just ask yourself, “what comes after the jack of hearts?” If you only memorize one deck, then you know the answer. If you try to memorize 30 decks, you will have almost 30 different answers, so you don’t know for sure.

It’s possible that some people can link with some repeat cards. Like maybe someone can memorize 2 or 3 decks using linking. But the point is, you are much more limited in your accuracy with linking if you are trying to remember a long list. Can someone link 2 or 3 decks? Maybe. What about 10 decks? Or 50? 100? 1000? At some point you cannot use linking. Loci does not have that problem. With loci you can remember unlimited number of things as long as you have enough loci.

That means that linking for memorizing pi will not work. Digits will repeat too often and you can only remember maybe 10-30 digits. If you want to remember more you will have to use loci.

After thinking about this for some time, I believe the main challenge in memorization is continually adding more loci to keep up with everything you want to remember.

One method that is interesting is using natural associations to create a list of loci on the fly. For example: spoon, bowl, cereal, milk, cow, grass, lawn mower…and so on. But remember that this is still a chain, so you cannot repeat anything. If you use “cow” in 5 places or 5 different lists, you will not remember correctly.


I do not agree with that rtr.

I started memorizing with linking and managed to memorize 2000 decimals of pi without any problems with mixing images up. The only problem I’ve found so far is that linking is slower, but do not require any loci (obviously). The problem with duplicates is not a problem for me because I see every object as a part of the link and a part of the environment/context I memorize. When I memorized the 2000 digits I could take one two-digit number and then tell where in the linking chains (I used 20 of them) the number occurred.

I don’t see why linking would result in more duplicate images than loci. Care to explain, rtr?

That’s interesting spenen. Tell us more details about your link method. It sounds like you have images for 00-99 and link 2-digits per link. Do you link 1000 links in a row (for 2000 digits)? What do you mean when you say “part of the environment/context” and “I used 20 of them”? Does that mean you have 20 chains? Or do you have one long chain with 1000 links? Thanks for sharing!

Yeah, I used a system with images for 00-99 (now I have a system for 0000-9999), and 100 images for every chain, 20 chains for 2000 decimals. So what I do is that I make up 20 different stories that all take place in a “set”. A set is similar to what you have when using loci, it could be my house, the library, a specific room, inside a computer or whatever, but the difference is that I don’t create a path in advance but instead make it up as I go along.
In that way I can differ between links that consist of the same images.

If we take the number 8415 in two different sets, where 84=a lighthouse and 15=a tree, and set1=in the kitchen, set2=in a forest.

In set1 the link could be a lighthouse poking hole on a tree laying in the kitchen sink.
Set2 could be that a tall lighthouse is hugging a tree.

In that way I don’t mix up the links, and if I want to make them more distinct I could imagine a leaf tree in set1 and a christmas tree in set2.

The best thing with this technique is that after I’ve memorized the images I can remember the path I take through the set, and I’ve created a loci to use as a backup if I can’t remember the link.

So my method could be called loci-linking maybe :slight_smile:

The big advantage is that I don’t have to memorize any paths, I just improvise. And the drawback is that linking, according to my own experience, is much slower than ordinary loci, but tend to stick longer in my memory.

1 Like

Ah I see :slight_smile: So you are using a form of loci combined with chains. My point is, if you only use a single chain, you cannot remember a large number of things if the things repeat (like numbers). So to the original question, that is one of the main differences between loci and linking.

what would people recommend for studying?
Because i really never put much effort in linking because i always thought of loci as superior.
but i have this mental block to make many new routes. i think? i sometimes improvise a route when i need to memorize something… which also works…
but find it hard for my study to get myself to make the new routes! and linking works also quite wel i guess?

well i would like to know what others use for study!

I did not know that Jonathon Hancock had one the 1994 world memory championships using the linking method. I am extremely intrigued by this, does anybody know more about it? Or about any of the methods Jonathon Hancock uses?

About keeping track of the order of certain pieces of information using the linking method, there is a pretty simple solution. Just throw a peg in your link every so often, for example every 10th item or so. That way if you need to know the location of a piece of information you can instantly narrow it down. If you really want to get carried away you can throw a peg in every 5th item or so. But in my opinion that is a little overkill.

My own method or learning large amounts of information is also kind of a hybrid of the two. I turned the information I want to know into a list of basic objects and then use a story to link them together. Then I take the first object in the story and I place it in one of the loci of a journey. That way I always know where to locate it. I like to use journeys which I have formed out of movies for this, since it is easy for me to pick a movie which is relevant to the subject I’m learning. In this way each loci of my journey will contain an entire list of information. And one complete journey is able to hold a vast amount of information on a particular subject. Sometimes if I wish to expand upon a certain peace of information, I will use a link within one of my stories as the starting point of its own story, in this way my links branch out like some sort of mnemonic mind map.

I would be very interested in knowing if anyone else uses this type of method and if so how well does it work for you? I personally have had good success with it as long as I keep the stories distinct and used a good review method.

That is similar to what Jonathan Hancock does, at least according to his books. In his more recent books (Brilliant Memory Training), he mentions the combination of story method and loci method, eg. a starting point of each information will be placed in each location. So every location will be a scene or a story. So each location is a point, and the story would be the supporting details. I think in competitions, Jonathan Hancock used method of loci instead of just pure story method. But he does love story method. This is even more apparent in his first book (The Mindpower System).

It is interesting that we use link method to mean either the Link Method (Harry Lorayne’s 2 at a time link) or the story method.

I wonder which one is better? I would think 2 at a time link would save more time. Harry Lorayne did disapprove the story method, saying that it causes confusion and that we would be juggling too much information at a time. Another question is how do we use 2 at a time link (or Chain method) in a location?

What do everyone think?

Here is my problem with the two at a time link method. Simple example. So the first three items are elephant-watch-butter.

I link elephant and watch. Great.

I link watch and butter. Great.

Now, three months later, I can remember my elephant to watch link.

OK so now I am trying to remember what links with watch. Well, elephant actually…!

I must be missing something, right?

Links should not be symmetrical.

Elephant - Watch - Butter

Elephant is wearing a big watch on it’s trunk.
Watch is made out of butter, so it melts like in that Dali picture.

If watch linked to elephant, it would be a watch whose hands were elephant trunks pointing to the time.

So Elephant -> Watch is very different from Watch -> Elephant.

It is this asymmetry which stops you getting suck in the loops you fear.

1 Like

Hmm, I’m still not quite there. So would you be incorporating a story sequence at some level here with the actual watch on the trunk then melting?

Or is there some kind of rule built into your image links that define whether item X is the ‘linker’ or ‘linkee’ - poor terminology I think but I hope my drift is clear…? I understand that watch-elephant and elephant-watch are different in your example, but I’m not clear how that approach would translate to other situations.

The way I’ve always did the link method is to make things follow on in a story. Elephant -> watch -> butter -> bus -> salt -> cow -> gravy -> pistol -> flag

I’d see an elephant wearing a watch made of butter that melts onto a bus, that I pick up and use as a salt shaker on a cow, whose udders squirt gravy onto a pistol which fires a BANG flag.

Am I similarly wrong?

Wow, I never know there’s such a rule. Very interesting. May I know where did you get that from?

I would actually just go: An elephant wearing a giant watch around its trunk. A giant watch is eating butter. So elephant reminds me of the watch, watch would then reminds me of butter. or if i go backwards, butter reminds me of watch and so on.

I think people use Link Method interchangeably. In my mind, I see your way as Story Method as it is a ‘story’. Whereas for Link, I would see it as an association between 2 items at a time.

So in your example:
Elephant, Watch: Elephant wearing a watch on its trunk
Watch, Butter: A watch eating Butter
Butter, Bus: A yellow bus is melting like a butter
Bus, Salt: A bus is filled up with salt rather than passengers
Salt, Cow: A cow appears from a pile of salt
Cow, Gravy: A cow swims in hot gravy, enjoying a bath
Gravy, Pistol: As you fire a pistol, its gravy flowing out of it
Pistol, Flag: A flag is erected from the tip of a pistol

When recalling, elephant would remind me of watch, watch reminds me of butter and so on. The advantage is that, you only focus on 2 at a time. Story method, sometimes I find, if the words are so random, can consume a bit more time to make up a story to link them together.

1 Like