A Novice→Expert Model of Learning

Every artist was first an amateur.

Ralph Waldo Emerson

One of the best understood principles of cognitive psychology is that novices learn and think differently to experts. These labels are domain-specific, not person-specific; I can be an expert at particle physics whilst still being a novice at evolutionary biology. Or skateboarding. Similarly, you could be an expert skateboarder whilst knowing little of nothing about theatre design or ancient Tibetan languages. What this means is that we're all novices at something, and many of us will be experts in at least one domain.

To demonstrate how you think differently as an expert to how you think as a novice, let's try a brief thought experiment. Pick something you consider yourself to be something of an authority on. Expertise is relative, so don't be shy. If you're a teacher you'll be fairly expert at many of the things you teach. For instance, I know a heck of a lot about Macbeth. I've seen four different theatre productions and at least five different film adaptations. On top of that, I must have taught it to perhaps 20 different classes, several times as an A level text (although I've never actually studied it as a student.) I've also read several books of literary criticism. As a consequence, I'm steeped in the bloody thing! Not only do I know the characters and plot inside out, I can quote sizeable chunks of it. I know a fair bit about the context in which it was first written and performed and also about how its critical reception has varied over the centuries. Whilst I'd never have the hubris to describe myself as an expert - there's always someone who knows a lot more - I'm not too shabby. In comparison, despite my grounding in literature as an academic study, I know practically nothing about French dramatist, Nicolas de Montreux's 1601 tragedy, Sophonisbe beyond the fact that it's about a Carthaginian woman who lived during the Second Punic War with Rome.

Now, imagine I had to write an essay about each of these plays. If you gave me any essay question on Macbeth I'd feel pretty confident that I'd have something interesting - although perhaps not original - to say. But faced with writing about Sophonisbe, even if I had a translated copy in front of me, would be tough. Why might this be? Although I'm pretty good at essay writing and know how a literature essay is supposed to sound, I'd struggle to write much worth reading about Sophonisbe because I don't know anything about it. I'd have to rely on guesswork, half-formed thoughts and trite stabs at close analysis. No doubt I'd do better than many other people who've never studied or taught literature, but I'm pretty sure that any literature undergraduate who'd actually read and seen the play would be able to outclass me. Basically, in literature - as in every other domain - specific knowledge trumps general ability. Now think about how poorly I'd be likely to do if I was assessed on organic chemistry, or tightrope walking.

Our ability to pay attention is limited to about four 'chunks'. These chunks can be very small - like the tiny smattering of Japanese vocabulary I possess - or pretty big like my knowledge of Macbeth. Experts 'hack' the limits of working memory by being able to draw on huge reserves of inter-related chunks of knowledge, allowing them to free up cognitive resources to enable them to think about the novel aspects of whatever they're thinking about. Think of it this way: a novice struggles to see the wood for the trees. If trying to navigate through a forest they'd head in and hope for the best but would quickly become confused and lost. An expert would take time to survey possible routes through the forest; they'd think about other occasions on which they'd undertaken similar journeys. They'd plan their path and, if they did get sidetracked, would have a range of strategies both for noticing the detour and for getting back on track. The novice's experience of walking through the forest would be completely different to the expert's.

So, how can we move from novice to expert? Broadly, I think there are two main hallmarks of expertise:

1. Automaticity of foundational procedures
2. Ability to see 'deep' structure within domains of expertise

We need to master various procedural knowledge to the point where we no longer have to think about it so that it doesn't take up space in working memory. These automatised procedures are often so well embedded that experts are not even aware of them. Although we might not actively think about these things, we most certainly think with them. This lack of insight into the source of expertise is sometimes called the curse of knowledge, and can lead us into neglecting the teaching of the vital nuts and bolts on which expertise depends. When you learn to drive, concentrating on your feet, hands, mirrors and the environment outside the car requires enormous cognitive resources: you have to pay attention to everything. When you've been driving for a few years, the basic operations for changing gear and and taking a right hand turn have been automatised. Your working memory is free to attend to road conditions and make predictions about what is likely to happen in the next few seconds. Likewise, when writing an essay, a novice will have to pay attention to such minutiae as punctuation, capital letters, sentence structure and academic tone. With practice, these things can be automatised allowing much greater availability of working memory to think about the content of the essay. This is equally true of any domain of expertise: the less attention we have to give to the basics, the more we can think about what matters.

The second area - the ability to see 'deep' structures - comes with the experience of thinking about domain-specific knowledge. The more we know and the more practice we have at identifying and solving problems within in a domain, the more likely we are to see through the superficial trappings of a problem to the underlying structures beneath. When we become aware of these structures we become increasingly able to transfer our ideas between contexts. The more I know about the domain of literature, the easier I find it to see connections between different texts. For instance, when I first watched Reservoir Dogs, with all its all long monologues and blood-thirsty revenge, and especially the set-piece stand-off in the final scene where everyone shoots everyone else, my immediate thought was, Oh, it's a Senecan tragedy! How was I able to see past all the silly names and ear slicing to see this underlying structure? Well, I took Classical Studies at A level and had read a couple of the Roman dramatist, Seneca's plays, then, as part of my English literature degree I'd been shown Seneca's influence on Hamlet, as well as writing an essay comparing Thomas Kyd's The Spanish Tragedy and Titus Andronicus. You could say I understood some essential dramatic principle but that would add an unnecessary layer of obfuscation - I simply knew more than lots of other people who saw the film and arrived at very different conclusions.

All this is to make the point that expertise, changes how we think. Here are a few of the essential differences:

Novices	Experts
Little relevant background knowledge	Lots of relevant background knowledge
Relies on working memory	Relies on long-memory
Lacks effective mental representations of successful performance	Has a clear mental representation of successful performance within a domain
Has not automatised necessary procedural knowledge	Necessary procedural knowledge has been automatised.
Only has explicit knowledge	Possesses huge reserves of tacit knowledge
Problem solving requires following clear steps	Problem solving is intuitive
Sees superficial details	Sees underlying structures
Learns little when exposed to new information	Learns a lot when exposed to information about which they are already knowledgeable
Learns best through explicit instruction and worked examples	Learns best through discovery approaches
Is more likely to experience cognitive overload as attention is swamped by new information	Is unlikely to experience cognitive overload as attention is buttressed by memorised ‘chunks’ of knowledge
Struggles to transfer principles to new contents	Is able to transfer principles between related domains

These strongly suggests that the continuum from novice to expert is what we most need to understand when designing instructional sequences for students. While our aim will be for them to become increasingly expert, we need to acknowledge that they will be novices for the greater part of their time in school.

This graphic (adapted from something Greg Ashman came up with back in his Harry Webb incarnation) is a neat shorthand of some of the points above:

As a way forward, we might do well to implement the strategies Hattie & Donoghue suggest for surface acquisition with novices, and then, as students become increasingly expert, we might think about cautiously moving to the 'deep' strategies. I suggest that the 'transfer' strategies be reserved for the very end of an extended teaching sequence.

And for those of you who like your models to be a little more complex, the Dreyfus model of skill acquisition may be more to your taste. The Dreyfus brothers proposed that expertise develops in five stages: novice, advanced beginner, competent, proficient, and expert. I'm not sure we need all these categories, but if you like them then I can't see as they do any harm.