SEM Cubed

SEM cubed can be described as a filing system for the mind. It allows you to memorize over 10,000 pieces of information or rules on any subject, or range of subjects, perfectly. At the same time SEM cubed improves creativity imagination and other skills. Tony_Buzan? is the creator of SEM cubed and MindMapping, these techniques being some of the only major advances in memory techniques since the mid-seventeenth century. The description of SEM cubed can be found in Master Your Memory by Tony_Buzan?.


SEM cubed brings together two main principles, the MajorSystem which is a method of replacing numbers with images and a Greek discovery that memory is on the whole an associative process. The MajorSystem can be used to remember, with great efficiency, any multi-digit number. The rather abstract concept of number is replaced by an easy to remember image. This is very useful for memorizing phone numbers or PIN numbers. As a prerequisite to learning SEM cubed you should ensure you have learned and used the MajorSystem first.

By combining the 100 MajorSystem images with the 100 SEM cubed images a 10,000 unit PegSystem is created.

SEM cubed is a matrix of 10 rows by 10 columns giving 100 cells. The matrix is self enhancing since each row emphasizes a different subject including the senses which are important for successful memory. Each of the 100 SEM cubed images can be combined with each of the 100 Major System images to form a compound image which can be used as a memory hook. In this way 10,000 memory hooks can be formed providing the mental filing system which allows instant recall of at least 50,000 items or rules.

To understand the way in which SEM cubed works it is best to provide an example, in this case the memorization of the chemical elements. First of all a section of the matrix needs to be allocated to this set of data, say the locations 8001 to 8105 with 8001 being used to memorize facts about Hydrogen (atomic number 1) such as the atomic weight (1.008), the year of its discovery (1766) and its chemical symbol (H). Continuing in this way through to element Unnipentium with atomic number of 105 being placed in the matrix at location 8105.

The 8000 row of SEM cubed is the rainbow category and the range of 8000 to 8099 is the colour red so the hooks in this range should emphasize be combinations of the colour red with the the Major System images. In this way the hook for hydrogen at 8001 is 'Red Day'. Use your imagination here but I prefer to imagine the start of a new day with a huge red sun rising over the horizon, emphasize the 'redness' of the sun. I then realize that it is in fact a large hydrogen balloon not the sun. This provides my link to hydrogen.

Once you have the primary link you can then link in other data, in this case the atomic weight the year of its discovery and the chemical symbol. As you can see each of the 10,000 locations can hold more than one item or rule by using the link system so the capacity of SEM cubed can easily exceed 50,000 items.

I saw your write-up on the SEM Cubed system and have read it a few times, but as I am not familiar with the periodic table of elements, I am finding it hard to connect the 8001-8099 numbers etc... I have Dominic_O'Brien?'s book and have an image for all numbers from 00 to 99 and have a few Journeys worked out too.

Just imagine you have a list of more than 100 items to memorize. Since the MajorSystem and DominicSystem both provide only the numbers 00-99, how do you do it?

Tony Buzan's answer is SEM Cubed. You can think of this as a hotel with 10 floors (0-9), each of which has 10 suites. Each floor has a theme, such as Rainbow or Animals, and each suite on that floor is in some way related to the theme. For example, on the Rainbow Floor, the first suite is Red, which is the first colour of the rainbow; the second suite is Orange, and so on. On the Animals Floor, the first suite is Zebra (z=0 in the MajorSystem), the second suite is Dog (d=1), the third is Newt (n=2), and so on.

Each suite in the hotel contains 100 rooms corresponding to the 100 pegs of the MajorSystem. Here is why the "room" 8001 as given in the periodic table example corresponds to "Red Day": Floor 8 in Hotel Buzan is the Rainbow Floor. The first suite on that floor is 80, the Red Suite. Room 01 of that suite is represented by the word "Day" in Buzan's version of the MajorSystem (d=1). So combine 80 (Red) with 01 (Day) and you have a Red Day.

For more information on SEM Cubed, consult Tony_Buzan?'s books, or see the link below from the World Wide Brain Club, which contains a complete list of the SEM Cubed images. However, since you have already memorized the 100 people and action pegs of the DominicSystem, you have everything you need to memorize 10,000 items already -- Room 8001 in Hotel Dominic would be represented by whatever combination of person and action you have designated for HOOA. This will save you time learning another 200 memory pegs (100 MajorSystem pegs and 100 SEM Cubed pegs), which might be time well saved if you are cramming for exams.



My guess is that the DominicSystem could be used to memorize 10,000 items or more with less fuss. (Details on that page.) Anyone care or dare to compare? We need some science here. --Ron Hale-Evans

Personally, I am very sceptical of SemCubed. Most of the top memorizers, such as world memory champions Dominic O'Brien, Andi Bell, and Jonathan Hancock, all report using places ("Journeys" or "Loci") to store their images. I've never heard of anyone using a SemCubed or a similar system to successfully memorize a large amount of information. I learned SemCubed when I was starting out with learning memory systems, and found it confusing and ineffective; now I use a mix of techniques all based around using places to store images. If anyone does have a report of someone successfully using SemCubed, I would be interested to hear it. Until then I remain unconvinced!

Also, for anyone learning 10,000+ items, repeatedly revising the items in order to make them stick into memory becomes a major task in itself (Some memorizers claim to be able to recall material after just one sighting without revising. I am very sceptical of these claims; many top memorizers such as Dominic O'Brien don't claim to be able to remember things forever without revision, and those that do claim they can are often "stage memorizers" with an interest in exaggerating their ability).

(moved to FlashCard)


I must say I agree with everything in ThufirHawat's comment above. I too have tried the system and am very skeptical if anyone has actually used it to memorize anything like 10,000 items. It just seems to me that too many of the themes are too similar and would therefore cause confusion before long. Take for example the recommended theme for locations 8000-8999: rainbow. The numbers 8400-8499 have blue, 8500-8599 have indigo, and 8600-8699 have violet. How the heck are you supposed to keep the images for blue, indigo and violet apart for all the images in this range?

This is exactly the sort of question that should be answered in Tony Buzan's books. But there is nothing like this, no real examples, 'here is how you might go about learning the periodic table', 'be careful you don't make this sort of mistake' etc. Rather it is basically explained as 1) here is the Major system, 2) here is how you can increase it (SEM3), 3) away you go!

One thing I found curious is that in his book 'Master Your Memory' he has a list of recommended SEM3 locations, eg Artists in SEM3 section 1200-1399, Composers 1400-1599 etc. There is no explanation for why these locations were chosen--and why is this placed in the Foreword at the front of the book and not mentioned anywhere else???

I could be wrong, but I have the nagging suspicion that this is a 'theoretical system'. ie not something that Mr Buzan or others he knows has actually used (as documented) in a serious way. If you really want to expand the Major system beyond the basic 100 pegs, I would suggest that you use some of his other suggestions--he gives a list of peg words up to 1000 for example. Another suggestion he makes is to use the idea of themes again, but using 10 of them this time rather than 100. The reduced number of themes leads to much less scope for confusion between images. (Both of these can be found in the book 'Use Your Memory').

When I first looked at memory systems I was initially attracted to this system for a couple of reasons: 1) The large amounts of data it claimed you could remember. 2) The 'random' access to this data. 3) Being of a mathematical background, being more comfortable with numbers rather than journeys--an idea that seemed very strange to me initially.

However I quickly found 1) to be a spurious claim. Also 2) was not as important as it seemed, and the journey method was more flexible in this regard than I thought at first sight. For example it is easy to traverse journeys backwards. Also--if it is important--I will label every 5th or 10th stop with the position number to 'index' the journey. That way I can jump straight to the 15th item for example.


If someone wants to associate ten colours perhaps a good list would be red, orange, yellow, green, blue, black, white, gray, glass(transparent/colorless), and the tenth slot could be a flexible "other". Brown and purple seem too weak to reliably support a full colour category on their own, but could easily be used in that tenth "other" slot when the imagery supports a strong brown or purple theme. Polka dot, plaid, striped, or rainbow-hued could also be used at will in that tenth slot.


Regarding distinguishing between blue, indigo, and violet. These colors may have some association for you. For example, I read somewhere that Elizabeth Taylor's eyes were violet. I might associate indigo with ink and blue with the sea. So instead of just thinking of the color violet, I might incorporate Elizabeth Taylor into the vision. I too think just associating color is difficult. While I can picture blue, indigo, and violet, I have no confidence in remembering the distinction between a blue, indigo, or violet book for example. But remembering a book in association with Liz Taylor, spoiled with spilled ink, or floating in the sea (or water), I can clearly picture the distinctions.

Other Systems

SEM Cubed is interesting with the latter part of the book's interest in learning real life facts - although Shakespearean characters might not be to everybody's taste; more challenging is the idea that a theme can be merged with one of those basic list of images in order to produce a unique new image that can be distinguished from other images. For me personally, regarding increasing the library of images by a multiplication factor, the images formed by using idea A + idea B would be too similar to an image of idea A plus idea C. In the book, a 'the same dinosaur' has different situations which count as different pegs. Maybe many people can do this visualization effortlessly and others can not; in which case, the latter need another system. So, it makes sense to test the system and your own ability to use it before committing time to learning the entire system. M Curtis.

It would be useful to have a page on picking appropriate mnemonic techniques for the appropriate situation, . Perhaps you would like to write it? TheRightToolForTheJob?, or maybe MemoryTechniquesArentForEverything?, or maybe even LiveLife?, or DontMemorizeEverything?.

Usually when people are memorizing things like Shakespeare's characters, though, I think they're just doing it to practice their memory technique. Sort of like archery: You're probably not going to shoot things with arrows in day to day life. But you might like to refine your skill.

(That said, I have a friend who memorized all the lines of The Crucible in one week, and hence got the lead part.)

-- LionKimbro [[DateTime?(2004-06-18T19:22:39Z)]]

SemCubed is neither effective nor simple. And creating peg words for numbers above 100 seems to me another step in the wrong direction.

This is what I've done: After 99, no more images for individual numbers. Instead, I've simplified it by creating one image for the 100 group (100-199), then one image for the 200 group (200-299) and so on. Then I created one image for the 1000 group (1000-1999) and for the 2000 (2000-2999) group and so on.

So, for all numbers in the one hundreds (100-199) I'll have 2 images: one for the particular hundred group itself and one for the last 2 digits (the standard peg image from the 00-99 group).

For all numbers in the thousands, I'll have 3 images (thousand group, hundred group, standard peg).

For example: for the 100 group I use Danny Devito as the Penguin from the Batman movie. So that any number from 100-199 will have the Penguin. The image of the Penguin will tell me that the number I'm trying to recall will have the digit 1 in the hundreds column. For the last 2 digits I use a standard peg image. Therefore, 122 = Penguin + Mother Teresa (22 = nun), which is then associated with whatever you want to remember. 162 = Penguin + chain (62 = chain). 152 = Penguin + lion.

Same thing for the thousands. Each thousand group gets ONE image. For the 4000 group, I use Superman (there are at least 4 actors to choose from). So for 4162 = Superman + Penguin + chain. 4122 = Superman + Penguin + Mother Teresa.

The beauty of this system is that it makes no difference what story or general image you create, you can NEVER mix up the numbers they represent, since each image stands for ONLY ONE place (hundreds, thousands, tens/ones column). This means that the images don't have to be linear--they can be all mixed up since they each represent ONLY ONE PLACE (thousands column, hundreds column, last 2 columns).

This is especially effective when you want to remember phone numbers (or any kind of number). When you recall the images and want to translate them back into the number, each image can represent ONLY ONE place in the line. When you see King Kong (or whoever else you use for the 300 group) you AUTOMATICALLY KNOW that the 3 belongs in the hundreds column of the number you're trying to recall. And anytime you see your standard peg images, you KNOW that they belong in the last two places (ones and tens columns).

For phone numbers, I associate only one standard, perpetual image (Green Lantern) to remember the area code and another image to remember the exchange. Example: for the area code 312, I make up the standard combination that I would use for 312 = King Kong (my image for the 300 group) + Tin man (12 = tin). Then I combine these two with Green Lantern, so that when I recall the imagery and see Green Lantern with King Kong and the Tin Man, I know this number MUST BE THE AREA CODE because the only time I use Green Lantern is to remind me that the number is an area code. For the exchange (the 3 numbers following the area code) I use another image.

To repeat: the little story or general image you create doesn't have to be linear. You can put the Tin Man first, King Kong second, and Green Lantern third. When you "translate" or "reconstitute" the imagery back into a number, you ALWAYS put Tin Man at the end (the last 2 digits) because the Tin Man is ALWAYS your standard peg image for the last 2 digits of ANY number. You MUST put a 3 in front of the 12 because King Kong is ALWAYS your standard image for the three hundred group (any number from 300-399). In other words, you MUST end up with 312 all the time, no matter where in the story you place your images. And when you see King Kong and the Tin Man associated with Green Lantern, you KNOW that 312 MUST BE the area code.

The whole point is to GROUP the numbers into broad categories (100s, 200s, 300, 5000s, 7000s, etc.) with standard images and NOT use the phonetic alphabet. Using images in this way means that you NEVER have to translate each and every number into a word (in order to get an image). If you did, you'd have one thousand words or images for every single number from 1 to 1000! And that's just the FIRST thousand! You don't have to first translate the number into the phonetic alphabet, and then try to remember which word you used (or haven't used if it's a new number). Each place in the number (ones, tens, hundreds, thousands) will automatically conjure up the standard image you use for that place.

In this way, all new numbers translate as instantly into images as your standard peg images. when you see 4132, you immediately think Superman + Penguin + Moon, which you then tie into the thing you're committing to memory.