Initially I put this chapter among the slowing factors. But as the book took shape I came to the conclusion that multitasking is not fully understood yet, and it would be wrong to write it off without giving it a chance. The vast majority of people say multitasking is impossible. I think most of them don't really know what multitasking is and use the word for several different things at once.
If you ask anyone what multitasking is, they'll probably say it's the ability to do several things at once. That's the definition used in other productivity books. From that definition and some experiments they conclude that multitasking is impossible. What's more, in their view it only makes things worse.
To get to grips with multitasking we'll use an analogy with how computers work — an area where I'm a bigger expert than many efficiency coaches. Do you know why today's computers are so much faster than those of 50–60 years ago? One reason is simply more power. The other is the computer's ability to do several tasks at once, switch between them and snap back to the previous context. Imagine a program that has to make two requests to different databases, process the results and show them to the user. Old computers could only do that in sequence: query one database, then the other, and only then process both and show the result. Modern computers can run two queries in parallel, and without waiting for the slower one they can already process the result and even show it to the user.
I've written before that our brain is very like a neural network, only much more complex. And in some ways we can compare a person to a computer. In computing we have concepts like asynchrony, multithreading and parallelism. I'll define them briefly; you don't need to memorise, as we'll simplify anyway.
Asynchrony is when doing one task doesn't block the main flow. Instead of waiting for an operation to finish (e.g. loading data), the program can do something else.
Multithreading is the ability to run several threads within one process.
Parallelism is doing several tasks at the same time, which is only possible on multi-core processors. Unlike multithreading, where threads can run one after another on a single core, parallelism means different threads run on different cores at once, which greatly speeds things up.
So multithreading and parallelism both involve running several threads but differ in how they run, while asynchrony is about how tasks are managed and run without blocking.
To keep things simple we'll keep only asynchrony as a term, and fold multithreading and parallelism into one — let's say parallelism. For this chapter we also need to remember that we have at least two types of tasks: mental and physical, and there are also consumer and producer tasks.
Let's start with parallelism and an example where "multitasking" can really work. We brush our teeth every evening and learn one new word in a foreign language. Is that multitasking? I'd say both yes and no. Yes, because we are doing two things at the same time, both are useful, and it all happens in the same time window. No, because brushing teeth is a habit, and habits, as we know, are actions we've automated. Agreed, brushing teeth wouldn't stop us focusing fully on the word we're learning. In this example the two tasks are independent parallel tasks that fit very well in the same time slot. Why? Two reasons. First, as I said, brushing teeth is a habit and we don't focus on it. Second, here we combined a physical and a mental action. Combining opposite types of actions is, in my view, much easier than doing two of the same type. More examples of this kind of symbiosis that won't harm you: running and listening to a podcast, watching TV and drawing, having lunch and reading, washing the car and listening to English lessons. There are many. The main thing is that the actions are as different as possible in type and use different channels (e.g. eyes, ears, hands).
There are also known cases of doing two physical things at once: mirror writing, signing with both hands at the same time, and so on. Of course that's a skill you have to train.
In all the parallelism examples above I gave either purely physical tasks (mirror writing) or a mix of physical and mental (brushing teeth and learning words). I didn't give examples of purely mental parallelism. So far many experiments suggest that mental multitasking doesn't beat doing two things one after the other. Maybe that's true. But I'm open to the idea that the problem is that no one has really trained their brain for multitasking. Maybe no one thought of it, or no one needed it. The brain is one of the few human organs that still surprises us with what it can do, especially when it comes to deliberate practice and mental training. For a long time it was thought that the average person can't hold more than seven digits in short-term memory — as new digits are read out, they forget the earlier ones. But there are experiments showing that the human brain can be developed a lot. I'll refer to the book "Frog, Elephant and Broccoli. How to live and how not to": Ericsson and Chase summed up their experiments. Before them it was believed that the average person couldn't hold more than 7 digits in short-term memory. But after careful training, Faloona began to remember more and more. At his peak (after 200 hours of practice) he could remember 82 digits. To check it wasn't a fluke, Ericsson trained his friend Dario Donatelli. After five years of training Donatelli could recall 113 digits. Both he and Faloona clearly showed that the seemingly fixed limit of 7 digits could be improved by more than an order of magnitude. At one time the record was memorising 10,000 digits of π; now people can recite 100,000. But that's different — long-term memory, and π doesn't change. Ericsson read different digits each time.
So no one yet knows the limits of the human brain. With enough training, mental multitasking may well be possible.
Now for asynchrony. Remember, that's the ability to do something else while the first thing is in progress. Unlike parallelism, asynchrony already works well not only with physical or mixed tasks but with mental ones too. Say at work you're waiting for your boss's reply on a task. Even if the reply comes fairly quickly, in 5–7 minutes, that doesn't mean you should sit there doing nothing. Switch to another task and continue it. When the boss replies you can return to the first issue. You don't have to return the very same minute — it's better to give the second task another 5–10 minutes (or more) and get it to a natural stopping point.
To sum up the chapter. Multitasking is always a tricky topic and often misunderstood. By multitasking we can mean two things: parallelism — doing different things at the same time; asynchrony — switching between tasks to avoid idle time. Parallelism doesn't work very well for simultaneous mental tasks, but it works fine for mixes of physical and mental actions. I don't rule out that with enough trained intellectual capacity, effective mental parallelism is possible. Asynchrony works for any type of action, but don't use it if the switch interval is too short — just a couple of minutes. If the intervals are 5–10 minutes or more, asynchrony will help you avoid downtime and fill empty slots with useful work.
