Human Memory Function

The memory function of the brain is only incompletely understood. There is no proof for the existence of "photographic memory" in the strict sense of the phrase, though I have used this terminology in this book. All memory is associative. Thus when we visually "memorize" a Monet painting, we are actually associating the subjects of the painting with something deeper in our memory, not just a two dimensional picture composed of so many pixels. This is why great paintings or unusual photographs are easier to remember than similar images of lesser significance, though both may have the same bandwidth (number of pixels). As another example, if you take a photograph of a circle on a sheet of paper, the photo will be accurate; the diameter and location of the circle will be exactly correct. But if you make a "photographic memory" of the same circle in your mind and then try to redraw it on another sheet of paper, the diameter and location will be different. This means that you memorized it conceptually (associating with some previous knowledge about circles and approximate sizes and locations). How about photographic memory of the music score? I can actually see it in my mind! Isn't that photographic? It is easy to prove that this, too, is associative -- in this case, associated with music. If you ask a musician with "photographic" memory to memorize a full page of random music notes, he will have great difficulty, although he may have no trouble photographically memorizing an entire sonata quickly. This is why there is no better way to memorize music (photographic or otherwise) than from the standpoint of music theory. All you have to do is to associate the music with the theory and you have it memorized. In other words, when humans memorize something, they don't store the data bits in the brain like a computer, but they associate the data with a basic framework or "algorithm" consisting of familiar things in the brain. In this example, music theory is the framework. Of course, a super memorizer (who may not be a musician) can develop methods for memorizing even a random sequence of notes by devising an appropriate algorithm, as we now explain.

The best evidence for the associative nature of human memory comes from tests on good memorizers who can perform incredible feats such as memorizing hundreds of telephone numbers from a phone book, etc. There are numerous memory contests in which good memorizers compete. These good memorizers have been extensively interviewed and it turns out that none of them memorize photographically, although the end result is almost indistinguishable from photographic memory. When asked how they memorize, it turns out that they all use associative algorithms. The algorithm is different with each individual (even for the same task), but they are all devices for associating the objects to be memorized with something that has a pattern that they can remember. For example, for remembering hundreds of numbers, one algorithm is to associate a sound with each number. The sounds are chosen such that they form "words" when strung together, not in English, but in a new "language" that is created for that purpose. Japanese is a language with such a property. For example, the square root of 2 is 1.41421356 which can be read as a phrase that translates roughly to, "good people, good people are worth looking at", and the Japanese routinely use such algorithms to remember strings of numbers such as telephone numbers. To 7 decimals, the square root of 3 reads "Treat the entire world!" and the root of 5 reads "On the 6th station of Mt. Fuji, an owl is crying." The amazing thing is the speed with which good memorizers can map the object to be memorized onto their algorithm. It also turns out that these good memorizers are not born that way, although they may be born with mental capabilities that can lead to good memory. Memorizers develop after much hard work in perfecting their algorithms and practicing every day, just like pianists. This "hard work" comes effortlessly because they enjoy it.

A simple, but less efficient, algorithm is to map the numbers into a story. Suppose that you want to memorize the sequence of 14 numbers 53031791389634. The way to do it is to use something like the following story: "I woke up at 5:30 AM with my 3 brothers and 1 grandmother; the ages of my brothers are 7, 9, and 13, and my grandmother is 89 years old, and we went to bed at 6:34 PM." This is an algorithm based on life's experience, which makes the random numbers "meaningful". What is so intriguing is that the algorithm contains 38 words, yet it is much easier to remember than the 14 numbers. In fact, you have just memorized 132 letters and numbers with greater ease than the 14 numbers! You can easily test this for yourself. First memorize both the 14 numbers (if you can -- it is not easy for me) and the above algorithm. 24 hours later, try to write down the numbers from memory and from the algorithm; you will find the algorithm to be much easier and more accurate. All good memorizers have devised incredibly efficient algorithms and have cultivated the art of rapidly transferring any memory job onto their algorithms. You can do a similar thing with piano music by analyzing the theoretical, structural, or musical story of the composition.

It is now possible to understand how memorizers can memorize many pages of phone numbers. They simply end up with a "story", instead of a string of numbers. Note that a 90 year old man may not be able to remember your name, yet he can sit down and tell you stories for hours or even days from memory. And he doesn't have to be any kind of memory specialist to do this. Thus if you know how to use your brain, you can do things that seemed at first to be utterly impossible.

So then, what is it about associations that actually enable us to do something we otherwise cannot do? Perhaps the simplest way to describe this is to say that associations enable us to understand the subject to be memorized. This is a very useful definition because it can help anyone to do better in school, or in any learning endeavor. If you really understand physics or math or chemistry, you don't need to memorize it, because you can't forget it. This might seem pointless because we just shifted our question from "what is memory?" to "what is association?" and then to "what is understanding?". It is not pointless if we can define understanding: it is a mental process of associating a new object with other objects (the more the better!) that are already familiar to you. That is, the new object now becomes "meaningful".

What do "understand" and "meaningful" mean? The human memory function has numerous components, such as visual, auditory, tactile, emotional, conscious, automatic, short term, long term, etc. Therefore, any input into the brain can result in an almost infinite number of associations. However, most people make only a few. Good memorizers have brains that continually make numerous associations with every input, in an almost automatic or habitual way. The large number of associations ensures that even if some of them are forgotten there are enough left to maintain the memory. However, that is not enough. We saw that in order to memorize, we must understand, which means that these associations are connected and ordered in some logical way. Thus a good memorizer is also good at organizing these associations so that when you get one input (for example, the name of a person) you can immediately find what you want (his phone number) by following these interconnections. If the associations are not well ordered and interconnected, you may not be able to recall the number even if it is in memory somewhere. Thus good memorizers habitually make large numbers of associations, they reinforce them continually, and are able to organize these associations into logical structures so that they can be recalled. The brains of good memorizers are constantly seeking "interesting" or "amazing" or "mysterious" or "outrageous", etc., associations that make recall easier. These same principles apply to memorizing music.

The associative nature of memory explains why keyboard memory works: you associate the music with the particular motions and keys that must be played to create the music. This also tells us how to optimize keyboard memory. Clearly, it is a mistake to try to remember each keystroke; we should think in terms of things like "RH arpeggio starting from C, which is repeated in the LH an octave down, staccato, with happy feeling", etc., and to associate these motions with the resulting music and its structure; i.e., memorize groups and families of notes and abstract concepts. You should make as many associations as possible: Bach's music may have certain characteristics, such as special ornaments and colliding hands and parallel sets. What you are doing is making the action of playing "meaningful" in terms of how the music is produced and how the music fits into your mental universe. This is why practicing scales and arpeggios is so important. When you encounter a run of 30 notes, you can remember it simply as a section of a scale, instead of 30 notes to memorize. Learning perfect pitch or at least relative pitch is also helpful for memory because they can provide additional associations with specific notes. The most common associations musicians make are with emotions evoked by the music. Some use color or scenery. "Born memorizer" is a phrase without a definition, because every good memorizer has a system, and all the systems appear to follow some very similar basic principles.