Improving Performance: Anders Ericsson’s “Mental Representation”

For the last few weeks, I’ve been reading Anders Ericsson’s book Peak: Secrets from the New Science of Expertise which details his research of expert performers over the last 40 years. Ericsson’s work was popularized by Malcom Gladwell in his book Outliers in which Gladwell coined the “10,000 hour” rule, which is how many hours violinists in top music conservatories had put in to gain entrance.

Ericsson’s book (and research career) centers on how performers of all kinds improve. He argues the surefire way to improve is to follow the steps of “deliberate practice.” However, to follow such steps, a performer must first have an effective “mental representation” of the task they are practicing. Because it’s decently complex, I’ll devote the rest of this post making sense of what a mental representation is, and then in my next post, explore the steps of Ericsson’s deliberate practice.

A mental representation is an idea or visualization in your mind of the task(s) you are performing and what a good performance looks, sounds, and/or feels like. It includes monitoring your performance as you go along, knowing where you can improve, and taking action to improve. As such, an effective mental representation is akin to self-feedback.

Just below, I share what I think a mental representation would consist of in different tasks. Before I do so, though, I want to point out that a mental representation doesn’t need to be perfect and complete for a performer to make improvements. It’s just that, in my reading of Ericsson, the more accurate and complete the mental representation is, the more likely the performer will improve.

A musical example of a mental representation…I am an amateur guitarist. When I’m learning a new fingerstyle piece, an effective mental representation would include hearing in my mind what the piece should sound like. If I can’t read music or have never heard the piece before (common situations for beginners), I should have a good idea about how such pieces in general should sound. I should also have a good understanding of what the piece will feel like to play – for instance, how quickly my right-hand fingers should be working in the fast section of the piece, how the stacking of my left-hand fingers feel in the quick succession of barre chords toward the end. If I know what the piece should sound and feel like, I can take action to improve my playing as I practice the piece. If I can’t hear in my mind what a good performance sounds like, improvement will come slowly, if at all.

A motorsports example…I enjoy taking my sports car – a 2016 Scion FR-S – to racetracks in Colorado and improving my lap time, which is the time it takes to do one lap of the track, say 2 minutes, 10 seconds around High Plains Raceway, my local track. The richer and more accurate my mental representation, the more likely I can shave a few tenths of second off my lap time. A motorsports mental representation is primarily visual and kinesthetic, though does contain some aural elements. I visualize sections of the track in my head – for instance, I see at least two turns ahead of the turn I’m currently on, and the “line” (path) I should take through each. Before I reach the spot, I see in my minds’ eye where I should begin braking for the most important corners. And I feel what the performance should be – the amount of g-force as I take a corner at the maximum miles per hour and am flung almost, but not quite off the track at corner exit. I feel my body lunge forward as I get too aggressive on the brakes and hear the chirp of tires indicating a lock-up. If I can see and feel what optimal on-track performance consists of (and where I’m missing the mark), I can make adjustments to my performance and hopefully slice a few tenths off my lap time.

A software task example… And here, I’ll admit I’m still trying to understand what such a mental representation per Ericsson would consist of in this context. However, since my main gig is overseeing the design and development of software training, it’s something I must do.

Let’s say I’m an HR specialist. And I’m a learner in an instructor-led class centered around the hiring process. I’ve been given time to practice the “create job” task in a training environment. If a mental representation is an idea or visualization in one’s mind of the task being performed, a learner’s mental representation in this example would be primarily visual, maybe map-like or perhaps something like a process flow seen in the mind. Before the HR specialist begins the software task, he likely knows (or sees) the location in the software where he’s to begin. And once he’s kicked off the first task, he likely knows (or sees) the remaining main steps in the process. The duration of the task, though subtle, also likely plays a role in such a mental representation. Though trainers rarely tell learners how long tasks should take, the HR specialist intuits that this task is of moderate complexity and thus should take five minutes or so. If it takes less or way longer than that, the HR specialist knows something is up and we’ll need to adjust his performance.

Which causes me to wonder… perhaps in our software training, we should return more often to the high-level process flows we cover at the beginning of task discussions.

In a nutshell, a performer who has an accurate understanding of what his performance should sound, feel, and/or look like has a far better chance of adjusting in the moment and thus improving. Seems to me, we should spend a bit more time and energy helping our performers (learners) understand what good looks like.