Olli System = Convert numbers AND syllables into images

General Memory Chat
1

In this post I am going to outline what I call the “Olli System” ;p

My name is Oliver so there you go :wink: I am still getting started with it, so I hope we can improve the system together.

The KEY IDEA is to kill two birds with one stone by combining the learning of numbers with syllables.
This should allow the encoding of numbers AND words/ parts of words (and therefore sounds) to images with the same system instead of using separate systems or only a system for numbers. With this system you can create images from digits as well as from multiple letters (i.e. syllables).

First, I will give a short introduction for people who are new to this:

One thing mnemonics can be used for is to remember numbers, e.g. hundreds to thousands of digits in an hour.

I am personally not interested in that but I use the techniques for remembering historical dates.

Mnemonists come up with ways of turning abstract numbers into vivid images. This can be called “encoding”.

For example, instead of the number 2, you could remember a swan due to the similarity in shape.

To be more efficient, it makes sense to have images for the numbers 0-9 and 00-99 (Total: 110 images).

This also increases the size of your Peg System which you can use with - or instead of - a Memory Palace for memorizing lists.

One system for encoding numbers to images is called the Major System and another popular one is the Dominic System.

Both can easily be expanded into a PAO cluster system.

A 3rd level Major System or something like the Ben System which assign images for numbers 000-999 is beyond the scope of what I need.

Here is a good post on how to memorize historical dates:

http://www.memory-sports.com/2009/10/how-to-become-a-memory-champion-part-7/

(I plan to use “1.5 overlapping” mentioned in the article)

(People new to all of this, make sure you read through the links above before continuing)

There have previously been ideas about having images for syllables. It might not make sense to only remember vocabulary with a syllable system but having images for syllables could definitely help to jumpstart the brain when it comes to remembering any (foreign) word or concept.

I realized that a good mnemonist will have 110 images associated with numbers. I also noticed that you get the same number when you take 22 consonants and 5 vowel sounds.

Vowels:

A
E
I /Y
O
U

By the way: Oxford Languages | The Home of Language Data)

Consonant sounds:

B
C/K
D
F
G
H
J/Y
J (guttural sound like in Spanish)
L
M
N
P
Qu
R (English R)
R (French/German R)
S/C
SH/SCH
T
V
W
X
Z

Personalization: You can of course do this differently by choosing your own set of vowels (e.g. ä,ü etc.) with different consonants.

Check the IPA for inspiration: International Phonetic Alphabet - Wikipedia

So why not combine numbers and these sounds when creating the 110 images?

In comparison to, let’s say, the Major System, this seems more systematic and useful to me. Plus, it kills 2 birds with one stone by allowing the encoding of digits as well as letters.

For example:

1=Ba=Brian Adams
2=Be=…
3=Bi
4=Bo
5=Bu
6=Ka
7=Ke
8=Ki
9=Ko
0=Ku

01=Da
02=De
03=Di
04=Do
05=Du
06=Fa
07=Fe
08=Fi
09=Fo
10=Fu
11=Ga
12=Ge
13=Gi
14=Go
15=Gu
16=Ha
…
99=Zo
00=Zu

(Note that by splitting it up this way you always get a new consonant every 5 steps. For example: 01-05 starts with D, 06-10 starts with F, 11-15 starts with G)

Now, when assigning images one could either choose any type of image, e.g. “Ba” could be British Airways and the image of a plane or you could use people in order to expand into a PAO cluster system.

The only drawback I see here is that you would have all last names starting with the same set of vowels. This could be avoided by braking the “First Name, Last Name structure” and allowing just the first or last name, e.g. Ba would still be Brian Adams but you may also have Hu = Hulk, Ma = Madonna, Na= Napoleon, Pi = Picasso etc.

(Note: If you would like to add vowels in front of the consonants or do anything else that might expand this in the direction of a 3rd level system (i.e. covering 000-999), I’ll leave that up to you for now :wink:

Please let me know what you think of the “Olli System” =)

If you would like me to post a word doc with a full PAO cluster system based on the Olli System, please let me know. If so, please direct me to any already existing PAO systems for inspiration.

Such as this one: https://artofmemory.com/forums/major-system-based-pao-with-pictures-1028.html

Hope this helps,

cheers,
Oliver

2 Likes
Having trouble with trying to decide which memory method to use
Major System based PAO with pictures
2

Nice post, it reminds me my pain to have a major system and a dominic system with a different digit-letter convention :smiley:

I was looking for something like that as a “last solution” for learning english vocabulary.
Sometimes, especially for the short words, I do not find good associated image.
With this system, I can get an image by syllables decomposition. Then if I store the corresponding image into a well-chosen loci in my palace, I would expect to remember the sense of the word targeted. Just a question : Are you uninterested by “inverted syllables” like “AB, IT, etc…” ?
I am wondering if an alternative would be to keep the classic association for consonnant Sound (0=s/z ; 1=t/d ; 2=n ; 3=m ; 4=r ; 5=l ; 6;f/v ; 7=k/c;8=ch;9=b/p) and use the 50 first digits in your way, and the 50 others in an “inverted” way.

Cheers,

Bruno

3

I have syllables for all of my images too, but it’s setup a bit differently.
(0-9, 00-99, and 000-999)

Every digit is a vowel:
0 = o as in row
1 = i as in bee
2 = u as in two
3 = aa as in cat
4 = a as in father
5 = ai as in five
6 = ih as in six
7 = e as in seven
8 = ei as in eight
9 = uh as in lullaby

Double digits are counted like this (cv-system):

00 = so
01 = si
02 = su
03 = saa
04 = sa
05 = sai
06 = sih
07 = se
08 = sei
09 = suh
10 = to
11 = ti
12 = tu
etc.

I use image modifiers to switch the order:
00 = os
01 = ot
02 = on
03 = om
etc.

Cards get the same syllables (see attached image).

Triple digits are consonant-vowel-consonant as in the Ben System:
000 = sos
001 = sot
002 = son
003 = som
etc.

Here are some blog posts about the way I do it:

1 Like
4

Hi Josh,

Do you mind elaborating on this part a bit?

I use image modifiers to switch the order:
00 = os
01 = ot
02 = on
03 = om
etc.

…I didn’t quite understand how 00=so=os and 01=si=ot

Would you mind giving an example of how you use image modifiers?

(Side question: Is there a collection of image modifiers somewhere?)

I will expand on this post later on by referring to what I will have learnt from studying Bigram Frequency:

http://www.cryptograms.org/letter-frequencies.php#Most_common_bigrams_.28in_order.29

(Side note: As for building a PAO system, I found it quite helpful to google “famous people named X” For example:
http://www.biography.com/people/groups/famous-named-ben/ )

Hope the info from cryptograms.org might spur some interesting innovations =)

Best,
Oliver

5

Sometimes I make images that modify the meaning of other images. The only purpose of the image is to change the meaning of another image. I call them “image modifiers” or “tags”.

My image modifier to reverse my number syllables was to encase them in a block of ice, but I’ve been setting them on fire lately.

Another example of an image modifier or tag would be to attach a specific image to a foreign vocabulary word to mark it as masculine or feminine.

Here’s a PDF that shows the consonant and vowel assignments:

0-9 are vowels
00-00 are consonant-vowel
000-999 are consonant-vowel-consonant

0 is s/z or o
1 is t or i

Usage:
00 is “so”
10 is “to”
01 is “si”

Reversed:
00 is “os”
10 is “is”
01 is “ot”

That gives me 1,210 syllables to work with based on my number images. The 000-999 images are being reworked at the moment. A few weeks ago, I decided to redo them from scratch.

I’ll check out the links you posted…

1 Like
6

Hi Josh,

Thank you for explaining! =)

I now understood the way you do it.

For me personally, I think it would be too confusing to have two different sounds for a set of digits.

For example, reading “zo” when I read the number “ninety nine” would only be faster if my brain didn’t have to hesitate a second to figure out whether it was “zo” or “oz” this time.

Also when seeing the modified image, I think it might take a few mental leaps to get to the syllable or number. But that’s just my personal impression.

So, now to my analysis of how to get the most frequent 2 letter combinations into your phonetic number/ syllable system.

I had a look at the frequency of letters and bigrams in English.

http://www.cryptograms.org/letter-frequencies.php#Most_common_bigrams_.28in_order.29

Letters:

The most frequent consonants are (in that order):

t
n
s
h
r
d
l
c
m
w
f
g

I listed 12 because one might want to neglect w and h because they most likely appear together most of the time (e.g. in where, what, who etc.)

The Major System puts some consonants together (such as 1 = t and d) and by doing this covers all of the above plus even covering p/b.

  • So that’s good to know I guess ;p

Bigrams:

For looking at bigrams, I would use the Top 39 listed in Wikipedia

(The bigrams on Wikipedia and the ones on cryptograms.org are the same but the order is a bit different.)

One should note that some bigrams made it onto the list because they are words (such as in,it,is, on etc.).

In order to get this absolutely right, one would need a bigram list where bigrams were not counted when they were standing alone i.e. forming a word.

Anyway, back to the analysis:

Leaving out the bigrams consisting of two consonants, two vowels (and thus forming a new vowel sound) or the bigrams starting with a consonant (the latter is already covered by my system), you end up with the following 18 bigrams.

in 0.94
er 0.94
an 0.82
at 0.59
on 0.57
es 0.56
en 0.55
em (no frequency stated)
ed 0.53
it 0.50
is 0.46
or 0.43
as 0.33
et 0.19
of 0.16

al 0.09
ar 0.04
ur 0.02

(Note: “em” had no frequency stated but I assumed it would be more frequent than al,ar and ur)

I think it’s safe to neglect the last three (all under 0.1% frequency) which leaves you with 15 bigrams starting with vowels, which are more frequent than the 15 ones formed by using the least frequent letters of the alphabet (x,q,z).

In alphabetical order, these are the 15 bigrams that remain:

an 0.82
as 0.33
at 0.59
ed 0.53
en 0.55

em (no frequency stated)
er 0.94
es 0.56
et 0.19
in 0.94

is 0.46
it 0.50
on 0.57
or 0.43
of 0.16

So, in order to cover the most frequent bigrams in the English alphabet, one would have to insert the 15 bigrams above into the system and kick out some of the less frequent ones.

However, all in all, given the fact that half of the bigrams above form words by themselves and therefore have unreliable frequency numbers, I decided that it wasn’t worth breaking the logic of the “Olli system” outlined in my first post.

One could still create images without numbers for them later - if necessary.

Finally, I did end up making one slight modification. Instead of using two different Rs for my consonants, I reduced it to one general R and inserted ST as a consonant sound - since ST actually is quite a frequent bigram ;-D

7

WoW. Thanks a lot Josh !
I understood that your system is not specially made for word encoding, but if you had to encode words into numbers, would you use this technique of “image modifier” to encode words with two close consonnants sounds (like “word”, craven etc…) ?

8

Next, I’ll post the PAO system I end up with.

Cheers everybody :slight_smile:

9

It was made to be used for words too. I have certain rules, but it might sound strange:
If it’s a 3-digit number that has 6 in the middle, I can drop the 6.

164 is “tihr” can also be tr…
Actually, in my accent, “tr” is pronounced “ch” (like in cheese), so maybe that isn’t the best example. :slight_smile:

Not every sound can be represented with these syllables, but I have mnemonics for letters too. I think that mnemonics don’t need to completely replace natural memory, but can often just act as a trigger in order to get past the “tip of the tongue” problem.

I don’t need to encode every sound. Usually one or two syllables will trigger a memory.

There could be faster ways to do it. I don’t have a real problem with reversing things, because I also read numbers as consonant-vowel-consonant:

10 is “to” or “ot” “is”. [Edited 2014, since I think that was a typo.]
101 is “tot”
110 is “tis”

I’m familiar with reading 101 as “tot” and 110 as “tis”, so 10 is sometimes “ot” and sometimes “is” even without the 2-digit numbers.

I learned how to read a few different foreign scripts while traveling, and I noticed that the same symbol can have different sounds depending on a filter that gets switched on in the mind.

Example:
The letter “v” is pronounced “n” in Greek. I don’t have any problem reading it that way as long as my brain recognizes that the context is “Greek” and not English. I think my mind usually scans for a distinguishing letter like sigma or theta, and then uses the Greek or English filter for letters like “v” or “H” based on that.

If I’m memorizing some text and need to remember the word “in”, I will picture Obi-wan (62) on fire and then I know that the context is to read it backwards. If it were 162 it would be “bihn”, so it isn’t a stretch for me to see the “in” in 62 instead of the “bu”. :slight_smile:

10

Hi,
This looks very much like a simplified BEN system to me!

11

Your system reminded me of one I had read about previously.
Indeed I found a few references to similar systems.

Here is a posting from a German memory site by I think Ulrich Voigt.
I quote it below in Google Translate’s mangled English.

–quote
Vera and other Birkenbihl recommend working on a “rural-urban flow” with alphabetical lists of words and keep the seemingly even better and more progressive than the Major System.
Birch Biehl (The inner Archives, 2nd edition, 2005) are arranged alphabetically by headings on 75 such lists. For the letters X and Y, it has very often no ready words, sometimes not for C, J, U, W.

Although the opinion, ABC-lists are superior to the Major System, is quite naive, so that does not mean that such lists could not have their value.
But I think it is questionable from the above reason, in the ABC hold slavishly. Rather, I would recommend it to scale down a 20-letter ABC.
This has two advantages: First, we find a sufficient number of words to each letter, secondly, can the number easier to remember the letters, so you have not only a character-wardrobe, but a numbered Buchstabengarferobe.
The oldest approach in this direction, I know, dates from the year 1503 and was often ignored in the subsequent literature, particularly of chub (1707).

The reduction is carried by C = K = Q, I = J = Y, V = U = W. The memory is so ABCDEFGHILMNOPRSTVXZ ABC
I have in Donkey World (2001), p. 103 f. described, 168, 187 et seq. There is also a method is provided, to remember the numbers of letters.
Compared to the other and now used by Birkenbihl simple urban-rural river system, this old method seems to me clearly advantageous.

Chub maintained the unfavorable letter X, which I would not do. Rather, I would put Q = C = K = X and hold onto U , so would recommend
ABCDEFGHILMNOPRSTUVZ.

No matter: The idea was then, this letter based on the principle-a-s-i-o-u expand, creating thus a predictable 100-wardrobe.
Each word is thus transformed into a number between 1 and 100

For example, Aa = 1 Adam Allah / Ae = 2 Alete Adele / Ai = 3 high school Achill / Ao = 4 in A minor, Adolf / Au = 5 Aluminum Battery
For example, Na = 56 Napalm Napoleon / Ne = 57 = blacks Negligee / Ni = 58 Niger Niebühl / No = 59 = grade Nora / Nu = 60 = zero Nubia

Now you just need that any clever Meier such a system cleverly marketed as a brilliant reinvention."
–end quote

I feel that the idea of trimming the consonants down to 20 is a good idea. If each is matched to one of the five main vowels then you have a phonic basis for deriving 100 images which can be associated with numbers or used as pegs. Whether this is better than the Major, Domenic or Ben systems can still be debated however. I guess it depends on how you adapt it for your own uses.

Another reference to an alphabet system is at Mentat Wiki: Alphabet System
Yet another is at http://www.memorybloke.com/images%20for%2000%20to%2099.htm but this system of Michael Curtis looks less intuitive to me.

12

I use a modified Ben System. :slight_smile:
The full description is here [PDF].

The only difference from a Ben System, other than the consonant and vowel assignments, is that I gave up on making 2,704 images for cards, and that I have images for 0-9 and 00-99 that are based on my modified Ben System phonetics. The images are always referred to by single syllables, even if it doesn’t match the English word. Maybe other Ben System people do that too – I don’t know.

Examples:

10
pronounciation: TO
image: Otis Redding

11
pronounciation: TI
image: At-at

That way the system can also be used for words, etc. There are some examples in this blog post:
Memorizing Shakespeare Using the Method of Loci

13

I didn’t quite understand the Google Translated text, but will reread it on the train later. :slight_smile:

I use 10 vowel sounds, but if I want to encode letters, I use my mnemonic images for letters, which are different than the ones for sounds.

I still don’t have images for syllables with certain diphthongs like oi and ou, but can still encode them. “Toy” would be 10-1 (TO-I). I use a glottal stop to represent a break in digits/images, so if it were a sequence of digits like 10-1-5-54, it would be pronounced TO’I’AI’LA in order to encode the dashes. I don’t really have a practical use for that yet, other than holding IP addresses or passwords in working memory for a few moments. :slight_smile:

14

Wow, deep stuff, some nice ideas too. I use modifiers myself, though not for vowels or syllables, just as a personalisation of the SEM.

I thought “Olli’s system” would be a variation on the Skipnum system, but it seems to be entirely different. Good work lads.

15

Hey everybody,

I’m done creating the first version of my system. (Will most likely still get improved as I start using it)

http://rapidshare.com/files/4208052312/Olli%20System.doc

Feel free to share, post in wikis or whatever.

I personally find that it is quite intuitive, especially since I managed to find an alliteration for pretty much every P-a-o.

I updated the description of the system at the beginning of the doc.

Cheers

1 Like
16

I tried opening the document in Libre Office and it wasn’t working. Does it have macros in it?

(By the way, you can also upload .doc files to the forum, or to Google Docs and then link to the Google doc.)

17

Thans for letting me know Josh!

I’ve uploaded it to Google Drive this time.

https://docs.google.com/open?id=0B1RpWhsfWa9HSlFfeldtcjRzR0U

Let me know if you have any trouble accessing it.

Best wishes,
Oliver

18

Strange – it still isn’t working after I download it. Does it let you upload it to Google Docs and convert it into Google Docs format? Or can you save it as a PDF?

19

Random question: does anyone know how long Nelson has been doing this?

20

“Late 2009” according to this:
http://climbformemory.com/mission/