Practice for Performance

The three tasks of performance and targets of practice

When we practice, we say that we do it to learn things. But since the target of practice is procedural performance rather than conceptual knowledge, we need to be very clear about what it is that we are trying to develop.

Three components of performance

Any performance consists of three components or inputs. They all play a different role, require different schemas and different practice. Yet, they feed into each other. Each of the components can be thought of independently but usually they are very interconnected. Different disciplines and tasks with them differ in how prominent these components are within them.

The three components are:

1. Mental map
2. Perceptive eye
3. Fluent muscle

Mental maps

Mental map is what allows you to orient yourself in the situation you are facing. It is your primary sense-making instrument - it governs what you focus on, what you ignore and how you solve problems or react to new situations. Sometimes they are referred to as mental model.

The mental map or model are not the same as concepts about what you are facing. They are entirely pragmatic to help you achieve what you want. Just like any maps, they are not the territory. They can be extremely simplified or even slightly wrong but they are still useful.

For example, the mental map of a situation of you reading in a room with an open window and hearing a noise outside. Your mental map tells you that the hole in the wall is letting more sound in, so you stand up and close the window.

Another example is that you come to a device with a display or with a chord leading to an electrical outlet. Your mental map tells you that such devices have ways of being turned on, so you will look for a switch.

Mental maps also help you solve problems. The mental map you have of books is that they have sequential pages and text that ends on one continues on another. When the text going from one page to next stops making sense, you will check the page number to see if somebody tore out a page.

When you are writing and you cannot think of the next word, your mental map will send you to the thesaurus and will tell you roughly where to look.

To develop mental maps, you have to combine conceptual and procedural learning. Mental maps are not ideas or definitions. They are models of the problems you are trying to solve or the task you are trying to accomplish.

Perceptive eye

Simply having a map to guide you is not enough, if you cannot recognise waypoints or that you have achieved your destination. That’s why you also need an eye that can perceive what your map is orienting you towards.

Your eye can recognise multiple features of a device your map tells you can be turned on as a button.

When reading or writing, you have to be able to recognise the words and sentences as being about the thing you are writing - not just some drops of ink on a page. When speaking a language, you have to be able to hear the words as words and not just as a series of noises.

This may sound trivial because these are examples of things with which you are intimately familiar. But just try doing these things with a new language or a new device and you will soon experience what it is like to have a mental map without a perceptive eye.

You must be able to recognise things as examples of the things you have developed mental maps for. You may have a mental map that tells you to write in an academic style but that is not helpful if you cannot recognise what academic style looks like.

To develop a perceptive eye, you have to experience a lot of examples with feedback.

Fluent muscle

Finally, in order to be able to achieve anything, you have to develop fluency. You have to be able to perform the task quickly enough so that it does not just dissolve into meaningless components.

The fluency very often has some bodily dimension. Writing, speaking, playing chess, playing a piano, catching a ball - all of those require some involvement of the body as part of the activity. Without fluency, the performance would be so halting as to be useless or even stop being what we intend. Playing one note and waiting 10 seconds while we remember how to play the next one is no longer playing a song.

But even in less bodily activities such as reading or doing mathematics, we still require fluency. In fact, we can think of those activities quite usefully as sort of gestures. Postures we adopt as we orient our mind towards a problem. There’s always some muscle and movement involved - turning the page, moving our head to look at a graph, positioning the page to better see it. And it would be a mistake to ignore these. They are a part of the gesture.

As we know, fluency is where working memory plays the most important role because if we overload the buffer, we stop being able to do things at a minimum speed and the performance breaks down. This happens very quickly, sometimes so quickly that we may not even bother getting started.

So, we could say that developing fluent muscle is what is needed just to get started. You don’t start playing a song until you have played the individual notes at some sort of speed with the required gaps between them. You don’t really start speaking until you’ve said a sentence and you don’t start writing until you can write a sentence (or at least a sentence-like pattern) at a time.

But it is important to remember that fluent muscle is just a metaphor for fluency. It does not have to be muscle what is involved. But it is something we can develop.

Disciplines and the muscle / mental map / perceptive eye balance

Different discicplines will have a different balance of emphasis as to the elements of performance. History is very heavily focused towards the mental map balance whereas playing baseball is much more about the fluency of physical movement. But even historians have to have perceptive eyes to recognise archival documents quickly enough as relevant and fluent muscles so that they can write about it.

Perhaps the most balanced of all disciplines is musical performance. A musician has to have a finely attuned ear so that they can tell what they are hearing is good enough and they have to have quite sophisticated mental maps of the music they are playing. But above all, they require a huge amount of fluency which they spend thousands of hours developing through deliberate practice and constantly honing and maintaining.

A musician says: if I don’t practice for a day God knows, if I don’t practice for two days, I know and if I don’t practice for three days, the audience knows. [ref]

Chess is a bit different from music because the fluency does not have to be as embodied and in some matches, it is less important. But in blitz Chess, the fluency of performance is essential.

Writing is in many ways somewhere between music and chess. Unlike music, we don’t have the need to peform without stopping to think but we do have to be able to compose in sufficiently large chunks to get anything done. If we wrote at the speed of chess moves, nothing would ever get written. But writing has more space for developing mental maps and perceptive eye. We have to be able to recognise different types of writing.

Three metaphors of how schemas get from inventory to working memory

So, now we have established two crucial facts about performance:

1. In order to perform most tasks at a higher level, we need to operate with larger chunks in our working memory
2. To perform better, we need to practice to build up an inventory of embodied mental schemas

We can be fairly confident that this is a sufficient requirement to develop a successful programme of practice that will lead to better performance. We have examples of successful programmes that use one or both of these facts as a working assumption. But it would be very helpful for thinking about how to make the most out of these, if we had a good working theory of how we get the embodied schemas from their inventory to working memory and combine them to perform sophisticated and fluent actions.

The simple answer is we do not know. But we can come up with some useful metaphors that will help us in thinking about it and making some useful adjustments to our programme of practice. The metaphors are:

1. Computation
2. Projected overlays
3. Clay impressions

Computational metaphor

I claimed that we do not know how the contents of our minds get pulled into working memory where we can use them for performance. But many people will claim we do. They are most likely to be using a version of the computational metaphor of the mind.

Many people may not have a good model of computation. The easiest way to think about it is as a series of if-then statements. If the I move my pawn to this square, my opponent will be able to take it with her rook. If the shape I’m seeing in front of me is standing on two feet, has a head with hair on top but none on their torso, I’m looking at a human being. And so on.

This is a very seductive metaphor because everybody has experienced that kind of a process in their head when they are trying to make a decision. But it is not actually very useful for learning. That is because we cannot describe all the things that need to happen computationally in this way. There are just too many computations.

In the sentence about recognising human beings, I used a lot of words. But they would also have to be computed. How do I know that what I’m seeing is a head and what I’m seeing on top of this is hair? Every single thing would have to be computed. And computations need operations like is, not, if, then and they need to work on things with clear boundaries. This works for chess boards but not much else - we also have to assume that the inputs into our computation exist as solid objects - and that is not very helpful when we try to figure our how to learn to perform quickly. Because we know that we are not explicitly making if-then statements when we recognise what we see.

Maybe our brain is doing it at some level but not at the level accessible to us - and not in a way that would be useful to performance. In fact, it would actively harm performance. So whether, the computational metaphor is useful to cognitive science or artificial intelligence or not, it is not useful for helping us think about learning to get better at performance.

For that, we can use the metaphor or a transparent overlays which is inspired by the conceptual metaphor theory which provides a sense of what schemas look like and its successor conceptual integration theory which helps us think about how we reason with them [ref - Lakoff 1987, Fauconnier and Turner 2003].

Projected overlays

Imagine two semi-transparent sheets of paper (lets’ call them transparencies). These transparencies have various images drawn on them. Some are very detailed, others just outline sketches, some full of colour, others in black and white, some very sharp and others fuzzy. When you put two or more such sheets together and hold them up against the light, new shapes appear. Some elements of some shapes may get new outlines, others may be hidden, yet others filled in. What happens depends on how well we aligned the sheets, their orientation and also the order in which we put them together.

We can put as many sheets together if we like but there are limits to what can be used. Too many, and previously newly emerged shapes disappear again or are transformed. And eventually light stops shining through.

What is important about this metaphor is that we have a very good bodily depiction of this that anybody can experience. And what we know from that experience is that the new image that comes out of the combination is instantaneous. It does not require a computation. In fact, we would find it difficult to compose a series of if-then statements that would give us that outcome.

This in fact describes how most people go about their life. They quickly pull up schemas and overlay them on top of one another to get new results. Because these overlays are never perfectly aligned, we experience new and unexpected shapes - which gives us creativity, something that is really difficult to capture under the purely computational metaphor.

But it is not exactly true that no computation is involved. When we are trying to go further, we may make some of these statements about the how we position the sheets and what would happened if we did that. That’s what it feels like when we reason. But notice, we are reasoning about the positioning of the overlays (concepts, schemas), not about their contents. Those just come up.

This is a very useful metaphor for thinking about using and combining mental schemas. But it lacks any notion of learning or return feedback. These are crucial for thinking about practice. So, this is where the clay metaphor comes in.

Clay impressions

Imagine a ground covered with clay from which you want to make shapes using molds and cookie cutters. But the clay is not uniform, it reaches deeper in some places and hides rocks in others. It also contains ruts and grooves from previous cuttings. So, when we try to impress our molds or cutters into it, sometimes they go too deep and other times they can’t get in. Or can’t get in uniformly.

This means that sometimes we have to push harder or multiple times and sometimes we may even need a hammer to press the cutters in. Or sometimes, we need to reorient these molds so that they fit around the obstacles. Some of this happens without thinking and some of it requires that we pause and think about what might be hiding under the mud.

To get the shapes we want, we have to use our bodies and our minds, employ both automatic and deliberative actions, repeat the same action multiple times and then change our approach when it is not working. As the consistency of the clay changes, we react to the changes by slow adjustments in pressure. But we may also bring different tools as we want to create more elaborate shapes.

The clay metaphor is useful for thinking about how we develop the schemas we have in our inventory as we interact with our environment. It is not a matter of collecting things that just lie out there and that we can pick up and stack up somewhere. We are actively engaged in their creation. We leave an impression in the world and in return in makes an impression in us.

We have all experienced physical activities like that as children or as adults. But we think of them as separate from the work of thinking and learning. But in fact, thinking and learning is much more like this than it is like a computer. So, as we try to develop our inventory of mental representations we should not be afraid to push against the subject or our discipline, feel the touch of it and adjust our reaction. This will not be nearly as instantaneous as when we push a mold into clay, it the principle is the same.