The Orange Juice Analogy for Incremental Reading


Around 2006 I began using SuperMemo and it changed my life. It wasn't just the software, but the writing of Piotr Wozniak and the dialogue that we began and continue to this day.

I discovered a whole new way of learning, as well as a new philosophy and new ideals to strive for. Excited by my discovery, I began to try to explain these new ideas to people, but they fell on deaf ears. The ideas were positively alien to everyone I knew! Those who did get mildly interested looked up SuperMemo and interpreted it to be nothing but a "flashcard" program; an ugly and warped description of its true power.

After a couple of years I discovered the power of metaphors for helping people to understand such new and different ideas. Since then I have come up with many metaphors for explaining various parts of the philosophy. However, one of the earliest ones that I came up with, and one that remains a core part of my philosophy, is about orange juice.

Making Juice

Suppose we race to make some orange juice. We each have a jug and plenty of oranges, and the race is to see who can fill their jug first. We each take different approaches to win the race:

  • Approach 1: Squeeze each orange one by one until it's dry, and then move on.
  • Approach 2: Squeeze each orange as long as the juice is pouring fast, but move on to the next orange as soon as the juice slows down, long before it stops.

While the second approach is likely to leave behind a large pile of half-squeezed oranges, the jug will fill much faster. Since approach 2 is much faster, how can we use it in learning?

Incremental Learning

In learning, our information sources (such as books, life experiences and other people) are the oranges, and the knowledge gained is the juice. The goal of learning is to squeeze as much knowledge from our sources as possible!

Incremental reading is a process pioneered by SuperMemo designed to do just this. In incremental reading, you read part of an article, find a good idea or two in there and then move on, leaving the article unfinished. You repeat that process for one article after another, until you've had enough for the day. When you return the next day, you repeat the process, with many half-read articles returning along with totally new ones.

While this approach can initially sound crazy, the orange juice story shows why it actually works amazingly well. Instead of focusing on the articles that you leave behind, half-read, you can begin focusing on the knowledge gained along the way. You stop focusing on "finishing" the book, and start focusing on the insights you are gaining.

The real effects of such learning go much deeper. Not only do we learn faster, but when we read many related articles in this way, we begin to easily see patterns and conclusions across articles, thereby building our own understanding in a much more sophisticated way than if by focusing on a single author's models.

A better way to learn

SuperMemo was one of my inspirations for becoming a teacher. I realised that the incremental process it offered was something that could fundamentally change how people learn.

If you're interested in experiencing incremental learning you can always try the process manually by simply keeping a bunch of books or articles and going through them at your own pace. Once you build up a large enough collection though, you'll need a more systematic approach, which almost definitely means some type of software.

Unfortunately, I consistently found that introducing SuperMemo to other people was a painful process. That is why I created Dendro. If you'd like a friendly introduction to incremental learning, please check it out.

Last modified: