The Spaced Repetition Party
So you’re at a party. It’s not some crazy kegger, it’s just one of those social mixers you go to every once in a while to meet people. A homely guy walks up to you and introduces himself as Craig. He’s a financial consultant. He soon moves on.
A few minutes later, he walks up again, and asks, “Remember me?”
“Uhhh, Craig, right?” you reply.
“Yes,” he says. “And what do I do?”
“Uhhhh,” you say intelligently as you draw a blank.
“Financial consultant!” he says snippily and walks off.
A few minutes later he’s back again. He walks up to you and looks at you. “Hey, Craig the financial consultant,” you say. He nods and moves on.
He shows up again an hour later, and then one more time before the end of the event. He’s satisfied you know who he is.
The scene described above is a fictional dramatization of how spaced repetition works. Just like you forgot unmemorable Craig’s profession only 5 minutes after meeting him, you forget most things you learn. That is, unless you’re reminded. And it turns out that there are optimal times to be reminded, and that the more you’re reminded, the less often you need to be reminded. This is the “spacing” of “spaced repetition,” and its rules been pretty well figured out.
The famous Pimsleur language learning system is based on the principle of spaced repetition. It was designed for a time when static audio recordings were cutting edge, however, and the latest adaptation of the spaced repetition principle is spaced repetition software (SRS), which has been refined quite nicely in recent years by a Polish man named Piotr Wozniak.
With SRS, you “join the party” by starting up the software. You’re presented with various “cards” or “facts” which you want to remember. Some of them, like Craig, aren’t particularly memorable, and when they come up again, you may falter. No matter; SRS is infinitely patient. The more you have trouble with a fact, the more often it shows up in your review cycles, until eventually you get it down pat and it gets spaced out to the point where you hardly ever see it again.
Sound like fun? In my experience, the idea of efficiently offloading the work of memorization to a computer program tends to appeal mainly to programmers. I was introduced to it by programmer friend John Biesnecker, who was seduced by SRS evangelist and blogger Khatzumoto (also a programmer). I’ve seen another programmer friend, Mark Wilbur, go fanatical about SRS. Meanwhile, linguists and language teachers tend to go, “meh.”
Personally, while I have my misgivings about SRS (a topic for another post), I think it’s a fantastic concept. The idea that, through science, we can understand how we forget, describe it in algorithms, and then systematically counteract it through software and learned behaviors is nothing short of amazing. The problem is that most of us aren’t willing to simply plug in and “trust the machine.” We prefer to live our lives unplugged… or at least not to be ritually spoon-fed our knowledge.
Like any innovative new form of technology, SRS has its early adopters. Those people swear by SRS, daily executing their spaced “reps” with the leading software: SuperMemo, Mnemosyne, and Anki. At the same time, though, something bigger is happening. Behind the scenes, SRS methods are infiltrating other learning software, such as Pleco (a popular Chinese dictionary). Although perhaps not completely obvious, SRS methods are a cornerstone of innovative Chinese character writing service Skritter. Cerego, the company behind another learning system earning lots of praise, Smart.fm, describes itself thusly:
> Based on years of applied research, Cerego has built adaptive, web-based applications that accelerate knowledge acquisition. Cerego’s patented core learning engine is driven by algorithms that generate optimal learning schedules for discrete chunks of declarative learning content, called “items”. This intelligent scheduling is achieved by gathering metadata on individual user performance and modeling memory decay patterns at the granular level of every item.
Guess what? It’s SRS.
The fact is, the average person doesn’t need to learn to change his habits to adapt SRS. As various companies and developers realize the value that SRS integration offers any kind of learning system, they’re integrating it into their existing products and services. It’s starting to appear in more and more products we already use. In the next few years, you can expect the slower ones to join the party as well. SRS is coming to you.