To Develop Expertise, Motivation is Necessary but Insufficient

Lots of readers of my entry on learning languages have said that the only reason I learned French well the second time (with the Assimil course) is that I was motivated. Here is one example: “Guy, the main reason that you learned French this time was because you wanted to learn it this time.”

Understanding the role of motivation in learning is important for designing productive learning environments — i.e. for learning well — so I would like to discuss it further.

Yes, motivation is important for learning! When I was in high school and training for the U.S. Physics Olympiad team, we heard (maybe apocryphal) stories about how our counterparts were being trained in the USSR: Candidates who didn’t make the cut got sent to the army. This kind of motivation, I thought, would definitely lead me to put in the needed hours.

To agree with the readers’ comments more strongly: For learning, motivation is necessary. However, there is a distinction between necessary and sufficient. Although motivation is necessary, it is not sufficient.

First I’ll give you an example from my own experience; then I’ll discuss a key research result from the study of expertise.

Myself, I’ve long tried to improve at chess. I learned the game from my mother when I was 3. My family tells of an uncle who had come to visit a few years later and was happy to find a 6-year-old keen on having a chess game. After I had won, I consoled him by saying, “If you want to win, you should play my dad.” One conclusion is that I was a little brat. The other is that I’d had lots of time to learn chess. And I was motivated. Sadly, at age 42, my chess skill is not much higher than when I was 6, despite lots of motivation and lots of practice.

Motivation does matter: We must want to learn. But the wanting is only part of a productive learning environment. We must also know how to learn — which I did not. That’s the key conclusion of the research on expert performance (an area of research discussed in this blog in the Q&A with David Shenk).

The how of learning is deliberate practice. For example, in school and college, to develop mathematics and science expertise, we must somehow think deeply about the problems and reflect on what did and did not work. One method comes from the physicist John Wheeler (the PhD advisor of Richard Feynman). Wheeler recommended that, after we solve any problem, we think of one sentence that we could tell our earlier self that would have “cracked” the problem. This kind of thinking turns each problem and its solution into an opportunity for reflection and for developing transferable reasoning tools.

Deliberate practice requires sustained concentration, and the rewards are subtle and apparent only in the long term. Thus, one needs motivation in order to enter into and sustain the hard work of deliberate practice. But the learning happens not simply through putting in the hours, but through doing so intelligently.

For chess, deliberate practice includes deep analysis of grandmaster games. You take an annotated grandmaster chess game — say, from Bobby Fisher’s My 60 Memorable Games — make the first few moves, then cover up the grandmaster analysis and try to figure out the move the grandmaster is about to make. If your candidate move matches the grandmaster move, great — move on to the next position in the game! If they do not match, try to figure out what the grandmaster understood that you did not. If you cannot figure it out, study the analysis in the annotations. Then repeat from step 1 with the next position.

This work is hard. But it is necessary. And it is productive. Canadian chessplayer Peter Biyiasas is said to have become an International Master by studying just two books, one of which was Fisher’s My 60 Memorable Games.

The power of deliberate practice is the message of the paper, “The Role of Deliberate Practice in Chess Expertise” (Applied Cognitive Psychology19:151—165 (2005)). Authors Neil Charness and co-workers studied the effect of different types of chess practice on chess rating. Hours of deliberate practice were a factor of 6 more effective than hours of regular practice (playing tournament games). No wonder I have hardly improved in 36 years. (And this factor of 6 comes after transforming the practice hours onto a logarithmic scale, which means that, in practice, deliberate practice is vastly more powerful than regular practice.)

So, do we need motivation to learn? Definitely. But alone it is insufficient. Deliberate (intelligent) practice plus motivation are together necessary and sufficient. Let’s design our learning environments so that they contain both.


How does one determine the best deliberate practice for an unknown subject? As I age, I find I have a lot less patience with unrewarding rote

Sanjoy Mahajan

That's a great question, and something I wonder about a lot. One answer is to find a great coach. Atul Gawande has described his finding a coach in order to become a better surgeon and teacher of surgeons ( )

My piano teacher, by the feedback and suggestions that she gave, helped me figure out what deliberate practice meant for pianists -- something I would have had no chance of figuring out on my own.

Joel Upchurch

You also have to consider innate ability. Practice and Motivation will help make the most of your innate abilities, but you aren't going to reach grandmaster status if your innate abilities are more at the Commodore 64 level.

D Doan

The difficulty lies in determining the limits of innate ability especially for more cognitive skills, which unlike physical skills are not nearly as constrained by physical limits (height, size, and so on) and are often culturally shaped. Defining and determining what innate ability means for cognitive skills is difficult if not impossible; better time would be spent on figuring out the best methods of practice and motivation regardless of some innate ability factor. A complex cognitive skill often has many components and therefore many avenues to exhibiting that skill. Unless the skill is a very narrowly defined skill there would more likely be multiple innate abilities for the expression of that skill.

Also we're should be comparing average people to one another not average people to cognitively impaired people (Commodore 64).

mike fladlien

You have nailed it. It's not "practice makes perfect", but "perfect practice makes perfect." A student might be motivated to practice, but if the student practices a technique imperfectly then the practice was wasted. Only deep reflection and focused practice will aid learning. To make myself clear, if a wrestler practices a takedown wrong every day, the wrestler will have little success on the mat. The wrestler needed more than motivation to learn. The wrestler also needed instruction.



Thank you for your nice article on freak nomics. It will help me.


harshbir singh

to be very honest this post about how to learn was quite shallow in its approach. i would like to make a refrence from the book "outliers" by malcolm gladwell where he discusses how to achieve perfection in almost anything we do. Of course the motivation and delebration are essential but the exposure to the type of enviornment also makes a huge part of the learning and achieving process. For example in his book he has given the 10,ooo rule in which he has mentioned the way if you have dedicated 10,000 hours of your life to anything which is supervised by professionals (Environment) you can be perfected in it.
Cause these kinda things actually defies success and expertise.

Steve O

I thought the article was great! You have to read the previous post (referenced in the first sentence) to understand his point of view. Mahajan isn't saying "think really hard when you practice and you'll be a grand master!" -- his point is that rote learning isn't as effective as deliberate learning, where your mind is engaged and your brain is making meaningful connections.

Jason Millard

A whole lot of rhetoric that basically asserts you cannot learn unless:

A. You want to
B. you're able to distribute practice throughout the long-term accepting that gratification must be delayed.

So what you're saying is executive function and the marshmallow experiment from 50 years ago is true?

Not qualitatively defining how to motivate or why masses of today's youth are seemingly less intrinsically motivated than thee predecessors is to write vacuousally about learning.

Do you work with students in public schools? Are you responsible for successfully motivating and disseminating interrelated facts, concepts, and skills to adolescents ? If not then I suggest you do a bit more research into what I faced on the front lines of tomorrows future.

Steve O

It seems like all of the initial reactions to this post are needlessly negative... Mahajan's posts on the learning have definitely informed my views on the subject and have validated my experience. I think it is vital that we study how people learn and continuously strive to do a better job of teaching (and learning). We can look at how people learn languages or math or medicine and draw definitive conclusions about one way being better than another.

Steve O

edit: **most of the initial reactions** not "all"


Steve just because you're not current on research doesn't mean those of us in education who'd like new research aren't.

This is regurgitated knowledge from the past 20 years, at minimum.

Secondly, intrinsic motivation and home support accounts for more than anything mentioned here.


I guess you are still not as good as you like in Chess, right? Reading your article it feels still too simple. Deliberate practise is about more then just reflecting on what just happened. Reflecting on your chess games and the ones of a grand master are all just random experience. To become great, you need to find "skill chunks" that you can train seperately. A football player for example will not just play and reflect. A big part of his training will be just running for long times. Another part will be to just run very fast. Another part will be to pass the ball. And so on. This is what deliberate practise means! Yes, the feedback is important. But cutting your training process into small steps that are all trainable just by themselves is what will get you ahead of the average player.
Maybe you didn't talk about that, because for you it is just the natural process to learn. In that case just be reminded that it is not usual for most people to learn this way. But I think it is more likely, that you don't think this way about learning yourself. Give it a try!


jack sparrow

Mahajan completely misses the point... Sure you need a lot of things to learn, motivation, concentration, environment, etc etc. ... but an incredibly important factor determining your ability to learn is genetic. For intellectual persuits like learning new languages and improving chess, IQ is probably the single most important factor which he didn't mention at all. Not every kid can learn chess at age 3 even if Fisher or Kasporov try to teach them because they probably don't have the intellectual firepower to follow!

Mahajan is just being plain glib here...

(Gladwell's Outliers is quite misleading on a lot of issues; i'll suggest anyone to follow Steven Pinkers review of Gladwell in nytimes and their subsequent exchanges where Gladwell starts resorting to ad hominems to argue his case.)


And yet IQ is increased by deliberate practise...(Dweck, 2000)

jack sparrow

Sure, I agree but only to some extent. I read a little of what Dweck has to say on perceived theories of intelligence, but I doubt she would disagree with my above comment.

To a certain limit deliberate practice will help, but then there would a road block defined by your innate ability as measured by IQ (for adults for whom it doesn't improve much). You can teach people things like chess or new languages but it would be a lot more difficult and time consuming for low IQ children or adults rather than high IQ. Deliberate practice will never make a below average or even average IQ kid/adult into a Gauss even if she puts 10,000 hours of "correct" practice (which i doubt she'll be able to do in the first place, as she will most likely not be able to sustain motivation to continue because she will get stuck far too often on basic concepts). He/she will just get too frustrated after some time because she won't be able to match pace with the high IQ kids working on similar stuff...

Mahajan should know that performance in certain pure mathematics courses like Topology, Abstract Algebra, Real and Complex analysis, Stochastic Processes are decided purely by IQ (accompanied with hard work of course... (it seems I am understating the value of hardwork here, but only to emphasize when hardwork becomes relevant. in reality hardwork is also critical in achieving greatness, but people miss the point when forgetting genetics just like Mahajan has done above!)). In fact, if anyone has been able to reach that level of sophistication in mathematics to take the above course their IQ has played a exceptionally big role in their reaching such a level (again i have deliberately left out 'hardwork')...

to summarize: you can't work hard (subsitute it with in "deliberate practice" if you like) on something when you can't even understand what/how to work hard on it !!


Jack Sparrow

Here is an experiment if anyone wants to test what I have beein saying.

Go to a low ranked college where and gather the students who have attained the age of 18 (18 years is important because I think by then the genes have played their most important role by then)...

Then, 1) test their IQ; 2) randomly select students (low and high IQ withing the strata) and try to get them the most brilliant teacher you can find to teach them topology or real analysis ... see what happens.

To save you the trouble, here's what I am almost certain will happen:
before you can even think of teaching anyone what topology or real analysis mean, you will need to teach them a lot of "basic" maths like functions, calculus, and logic. You will realise that by the time you have taught them basic maths 1) all low IQ students have disappeared (this should include almost everyone at a low ranked college);
2) now if there is anyone remaining (you'll see that the sample size will have shrunk greatly) and you have managed to teach them topology or real analysis, retake their IQ test and compare it with their scores on the initial test. If their IQ has improved considerably (i think it is very unlikely), what I said in my previous posts is essentially outright wrong!!



Working Memory is the best predictor to success, more so than IQ. See alloway.

Suggestions would be easy if there was research on motivation in the context of schooling alongside a real classroom setting with all variables that play into it. Psychologists almost never do those studies!

High IQs and useless "data" and "research" is not helping our education system. I fear ED Hirsch is the most articulate on this

WordPress Guru

Its all about enhancing our prepostions.