Figuring Out the Most Efficient Way to Practice, One Molecule at a Time

The old cliche has it that practice makes perfect, but what makes for perfect practice?

One of the first scientific bits of insight came over a century ago, when one of psychology's great pioneers -- the insanely patient cognitive psychologist Herman Ebbinghaus -- pulled a move from the Mad Scientist's Handbook,  and ran massive experiments on his own brain, not with strange substances, but with strange syllables. Over weeks and months and years Ebbinghaus teased his own brain with long (and sometimes very long) lists of arbitrary nonsense syllables, like BOK, DAX, and YAT, and recorded how well he remembered them, and for how long. 

Ebbinhaus' quarry was the recipe for a perfect memory -- or at least for the formula for most efficiently learning new information. Whether you are trying to learn a musical instrument, master a foreign language, or just study for an exam, the rate-limiting step is often memory. The faster you can convert new information into new memories, the faster you can master new skills. 

One of things that Ebbinhaus discovered, in the course of bending his brain on tens of thousands of nonsense syllables, was that it was better to space out what you learn a little bit each day, rather than cramming the night before the exam. 

Ebbinhaus's observation -- known as massed-versus-distributed (or spaced) practice law  -- has since become common knowledge, repeated in virtually every psychology textbook that's come since.

Yet nobody has really understood why it is true. Why should the interval at which you try to remember things matter? And what interval is the perfect interval?

To truly answer these questions, psychologists and neuroscientists need to bridge the chasm between their fields, and connect molecules to memories. Two studies, both published in the last month, represent incredible first steps. 

The first, reported in Nature Neuroscience and explained nicely in Scientific American, used computer simulations to look at the dynamics of individual proteins in the brain of a sea slug (memory researcher/Nobel Laureate Eric Kandel's organism of choice).  Neurobiologists at the Medical School at the University of Texas found a new way of practicing that was somewhere between spaced and massed practice, and somewhat more efficient than either.

The second, published earlier this week in Science, looks at the something known as the "mushroom bodies" in the nervous system of a fruit fly, identifying a particular molecular response that happened only after spaced practice, but not last minute cramming. This work helps pinpoint exactly why timing matters to practice, in terms of the dynamics of memory consolidation, and lends further credence to  the notion that that we might be able to use an understanding of molecular biology as a tool for building better schedules for practice.

Neither of these two studies is definitive; sea slugs and fruit flies aren't human beings, and their human neurochemistry almost certainly works differently than ours. 

But not all that differently. As I explained in my earlier book The Birth of The Mind many of the genes and molecules that underlie human brains are quite similar to -- and quite closely related to --  the molecules that underlie the nervous systems of other creatures. In biological terms, there has been an enormous amount of "conservation" of genetic material over evolutionary time. And that means that it's a good bet that these new studies will help us understand human brains a lot better.

If our brains worked like computers or smart phones; we needn't practice to make perfect. Each time we tried to remember something, it would simply stick. All we would need to learn something would be an installer CD or a download link; new skills and new memories would flow directly into our brain. 

Something like that -- efficient, immediate transfer of new information directly into our brains -- might happen some day; until then, the speed with which we can acquire new information is constrained by the clumsy evolution of human brains

For now, we have no choice but to work with the quirky brains that's we've got. The more we learn about exactly how their neurochemistry works, the better we'll able to use them efficiently. As I've written elsewhere, the idea of "10,000 hours of practice" is a bit crude, but often fine as a first approximation. It's not hard to imagine that new studies like these could eventually lead to new training regimes that cut that number by 10% or 20%  -- a truly thrilling prospect for anyone who has tried to learn something new.

Copyright 2012 Gary Marcus. Cross posted on