Archive | Ebbinghaus RSS feed for this section

Learning the Indian alphabet

2 Jul

One of the key messages of this site is big learning in small doses. In this post I want to give you an example of using this principle by explaining how I learned to read the Indian alphabet, more correctly know as Devanagari.

One of the languages I am trying to learn is Sanskrit, the ancient liturgical language of Hinduism. Many of the source texts of both Hinduism and Buddhism are written in this language using the Devanagari script. Here is the word yoga written in Devanagari:


Devanagari is used in many modern Indian languages including Hindi. So my first step was to purchase a set of Hindi script flashcards.

flash pile

The cards show a Devanagari letter (which represents either a vowel sound or a consonant and a vowel sound) and an image that is supposed to be a mnemonic for the letter. The letter in the card pictured above stands for the “ga” sound.

Here is the back of a card

flash card

I did not find the visual mnemonics helpful, indeed they were a nuisance because I wanted the letter to be the cue for remembering, but I solved this problem by placing a card over the bottom half of the card.

fash cover


Rather than try to learn the entire deck at once, I started with only three cards. Each morning, while walking on my treadmill, I would see if I could say the sound of each letter in the small pile. When I could correctly identify the sound of every card in the pile I would add one or two more. I would only add cards when I felt I has mastered the cards in the current pile.

When working with flash cards it is very important shuffle the deck every time you review. This is because of the serial position effect in list learning discovered by  Hermann Ebbinghaus. In learning any list we tend to have better memory for the beginning of the list and the end of the list and poorer memory for the items in the middle. However, we want to remember all the letters equally well and we do not want the order of the cards to be a cue for remembering the letters. So when working with flashcards, shuffle the pack for every review.

Over the course of a couple of months, spending a very few minutes everyday, I learned the basic Devanagari alphabet. I now review the entire deck of flashcards only three times a week just to maintain my skill. If I had tried to learn through intense cramming I think I would have failed, and I certainly would not have retained the information.

If you take a long term perspective and break a big learning task into small steps, you can succeed in mastering seemingly difficult material.


Hermann Ebbinghaus and the forgetting curve

29 Jun

It’s surprising that it didn’t drive him mad. Everyday for months on end German Psychologist Hermann Ebbinghaus would memorize and recite long lists of meaningless syllables. What ever the tedium of these exercises, it was worth trouble. For with his experiments, conducted in the 1880s Ebbinghaus, forged the modern science of memory.

Ebbinghaus was the first psychologist to perform rigorous memory experiments. After receiving his Ph.D. in 1873 Ebbinghaus spent two years teaching in England, where he came in contact with the associational psychology promoted by British philosophers. Also, while in England he came across a copy of the Elements of Psychophysics by Gustav Fechner, an early advocate of experimental psychology.

The great philosopher Immanuel Kant claimed that it was impossible to measure psychological processes. Fechner’s work was a direct assault on this pessimistic view. He was able to find law like relationships in the way people responded to stimuli. Ebbinghaus was deeply impressed with Fechner’s arguments and felt that he might be able move the study of learning from philosophical speculation to verifiable research. He used himself as a subject and many of his early conclusions have been supported by subsequent research.

Ebbinghaus noted that the study of memory presented a central difficulty. If memory is affected by the degree of association, how could you control for past associations? Previous exposure might contaminate experimental results. If asked to memorize a line of poetry, previous exposure, even unremembered exposure, might affect your current ability to commit the line to memory. To solve this problem Ebbinghaus hit upon the use of three letter nonsense syllables, often called trigrams. Each trigram consisted of two consonants separated by a vowel, such as MIB or LAJ.

In a typical experiment, Ebbinghaus would create a deck of thirteen cards with a single trigram on each card. He placed a blank card on the top of the deck.
On the first trial he would go through the deck turning over each card, trying to learn the cards in order. After the first trial the he would turn to the deck over and start again, but this time when he looked at the blank card on the top he would try to name the following card. He would then turn over the blank card and see if he was correct. Looking at the first trigram card he would try to guess the name of the second trigram card and so on through the deck. This is called the method of serial anticipation, because he had to recall, that is anticipate, the next card from looking at the card that preceded it.

Even though Ebbinghaus was working with nonsense syllables the method of serial anticipation, essentially a type of list learning, did resemble many important memory tasks. For example, learning the alphabet, the correct spelling of a word, or the lines of a poem, requires us to recall information in a particular order where each item is a cue for the next one.

When Ebbinghaus could work through the deck with out error twice he considered the deck learned and recorded the total time it had taken him to learn the deck.

After a deck was learned he would wait some amount of time, sometimes minutes, sometimes days, and test himself again. Typically he would make some mistakes and he would work through the deck a number of times to in order to relearn it. As with initial learning he considered the deck relearned if he could he could work through it twice without error.

Not surprisingly it took less time to relearn the deck. The amount of time it took to learn the deck minus the time it took to relearn he called the “savings,” a measure of how much work had been saved in relearning the deck over initial learning. Dividing the savings by the amount of time it first took to learn the deck allowed him to represent the information saved in memory as a decimal fraction.


By repeating this experiment many times with many different list over many different time intervals, Ebbinghaus made several important discoveries that bear on our goal of memory improvement. Perhaps the most important was a pattern that emerged over his many experiments, a relationship between initial learning, time, and forgetting that we have come to call the forgetting curve. The forgetting curve is essentially the opposite of the learning curve; the downward slope of the curve shows us how much information is lost over time. Studying this curve will lead us to understand those factors that increase resistance to forgetting.

%d bloggers like this: