Scott Young’s Ultralearning: Projects or Drill?

Posted on 19-01-2020

In 2011, Scott Young set himself the challenge of learning MIT’s entire 4-year computer science degree, despite not being a student at the university. He succeeded. His new book Ultralearning is all about how you can take on similarly tough challenges, and master really hard skills in relatively short periods of time.

The book is aimed at adult learners, but if you are a school student or teacher there is still a lot you can learn from it too. The parts I found most fascinating and thought-provoking in this context were chapters 6 & 7, which seem to be putting forward contradictory advice.

Chapter 6 is called ‘Directness: Go Straight Ahead”, whilst chapter 7 is ‘Drill: Attack your weakest points’. In chapter 6, Young recommends that we learn by doing – you have to practice skills directly, in a form as close to the final skill as possible. But then in chapter 7, he says you need to take a complex skill, pull it apart, and practice the specific bits you are weak on – that is, you need to drill.

Young acknowledges that these two principles sound a bit contradictory.

  • “Astute readers will probably notice a tension between this principle and the last. If direct practice involves working on a whole skill nearest to the situation in which it will eventually be used, drills are a pull in the opposite direction. A drill takes the direct practice and cuts it apart, so that you are practicing only an isolated component. How can you resolve this contradiction?”

I was particularly interested in this because you can argue that a similar tension exists in my own work. For example, in Seven Myths about Education, in one chapter I argue that transferable skills don’t exist. You can’t hope to teach students how to think critically about an episode of the Simpsons and then expect them to transfer such critical thinking skills over to, say, the English Civil War.

However, in another chapter I argue that projects and activities are not an effective way to learn – that drill and decomposition are much better. If you want to learn how to use an apostrophe, don’t do so by reading lots of books and writing lots of essays. Learn about the apostrophe in isolation and practice in isolation.

You can – and some people have – accused me of being contradictory. If I’m saying that transfer doesn’t exist, how on earth do I expect some of the decomposed drilling activities to work? What if students learn the apostrophe perfectly in isolated activities but then are utterly unable to transfer it to real pieces of writing?

In a sense, my second book, Making Good Progress, is an extended attempt to answer this critique. My basic argument is that you need ‘a model of progression’: a clearly defined end goal which is the real world skill you want students to attain. Then you need a series of steps and activities of gradually increasing complexity which will allow you to get to that end goal.

Young’s conclusion is somewhat similar.

  • “The tension between learning directly and doing drills can be resolved when we see them as being alternating stages in a larger cycle of learning. “

The reason this tension exists is because of two limitations of the human brain. The first limitation is that we think in incredibly concrete terms, and we really do find transfer really difficult. This limitation means it really is important to get practice, and to practice in ways that are similar to the eventual end goal.

But, pushing against that limitation is another limitation – our limited working memories. We can’t take in lots of new information at once. We need time to develop habits, fluency and automaticity. Complex skills like writing are made up of many moving parts. Trying to learn just by reading good writing and then writing lots yourself makes it hard to pay attention to all the different component parts you will need.

Given these two limitations, the perfect solution is drills that are gradually combined together in increasing complexity. Drills are concrete and specific, but they also respect the limitations of working memory. If they are sequenced and combined in the right way, they can overcome problems with transfer. Young has a nice way of thinking about this process of combining drills:

  • “Even in well-designed drills, there are going to be transfer hiccups owing to the fact that what was previously an isolated skill must be moved to a new and more complex context. Think of this as being like building the connective tissue to join the muscles you strengthened separately.”

To go back to my apostrophe example, one students have learned how to use the apostrophe in simple and isolated contexts, you can then step up the complexity by asking them to apply it in gradually more complex and free-form pieces of writing. This kind of ‘near transfer’ is still difficult, but it is possible. The ‘far transfer’ of expecting a student to think critically about topics as different as the The Simpsons and the English Civil War is so great that any attempt to bridge it would not require a few activities; it would effectively turn into a new subject (perhaps we could call it ‘history’?!)

So aren’t you just saying we need a bit of balance?

The easy conclusion to take away from Young’s book, or from what I have just said, is that ‘we need a bit of balance’. The odd project here and there, a bit of drill here – they’re both important. However, whilst this is an easy conclusion, I don’t think it’s right. For an example of why this is not the case, let’s turn to one of my favourite analogies: marathon running.

If you are training for a marathon, you can err by being too direct, which would be the equivalent of training for a marathon just by trying to run marathons.

You could also err by not being direct enough. There are a number of different ways you could err in this direction. You could spend all your time reading books about marathon training and writing plans, without ever actually doing the running. You could spend all your training in the gym, which might make you strong but ultimately won’t be enough. You could spend all your training doing fast shorter runs, which won’t prepare you for the 26.2 miles either.

However, you can also err by attempting to ‘find a balance’. For example, if you spent half your training sessions reading a book about running a marathon, and the other half attempting to run 26.2 miles, that also wouldn’t be very effective. If you spent half your time doing weights in the gym, and the other half doing 26.2 mile runs, that would also not be great.

The right way to think about things is not that we need a balance, but that we need a sequence. My favourite school example of this kind of sequence is the writing programme Expressive Writing. It begins with short word length activities and ends with lengthy writing tasks. In between are a series of activities which gradually increase in complexity, with plenty of interleaving and spaced repetition of core skills.

Designing this kind of sequence is hard, and this is a key message of Young’s book too. You would think that a book about Ultralearning would be all about finding the motivation and determination to work hard at tough goals, and certainly there is a fair bit of that. But in many ways, Young’s book shows that the hardest aspect of autodidacticism is not the work itself, but working out exactly what it is you need to work on. It’s easy to get bogged down in activities that are not really moving you forward. When you start learning something, by definition you don’t know much about it, and so you therefore don’t know what you don’t know.

Young’s book gives the best advice possible at trying to overcome these problems. It’s a fantastic summary of the research on learning and how we can apply that in our own lives.  One of the reasons the book is particularly valuable is that schools and universities employ few of these tactics in designing their own learning course, and that learning materials are typically not structured in this way. Too often, the onus is on the learner to work all of this out for themselves, even though the learner is not well-equipped to do this. My hope is not just that budding Ultralearners read this book, but that educators read it too – and work out what they need to do to help everyone learn better.

There’s more on this in my new book, Teachers vs. Tech, which will be published by Oxford University Press in March. You can sign up to my mailing list here for more updates about it.