Have you ever wondered what happens inside our brain whenever we are thinking about something? How some people have great ideas, are able to learn fast, and are incredibly productive, while others are struggling with finalizing even a simple task? There are many components to this process, and in this article, you will learn about one of them.
Working Memory is one of the several memory types where information is processed and manipulated before it is either discarded or committed to the long-term storage. Understanding its nuances can allow us to be more efficient in using it and be a bit more productive at what we do.
And since it is at the core of information processing by our brain, the things you are about to learn are independent of the domain you are working in.
Irrelevant Example First
I don’t know how many articles online make you do exercises, but let’s do one right now. Moreover, I count on your integrity to follow the exercise’s directions since I cannot check you play by the rules.
You will see below two blank boxes. Upon clicking on “Show Letters,” you will see two sequences (one after another) of random characters for 7 seconds each.
Rule # 1: after you click on the button, you have 7 seconds to read and memorize the first sequence, and 7 seconds for the second one (the sequences will display automatically, you only have to click on the button once). Rule #2: do not use any note-taking software (doc, excel, text editor, etc.) or hardware (pen and paper, pen and table, pen and walls; a pencil if fine though!). Well, you can, but you will miss the whole point of it.
Great! How it went? Sorry if you tried to click on a box the second time, and it didn’t show up. I put some sophisticated locking mechanisms in place to prevent that.1
Now, there are two more boxes below. Click on each of them, input the characters you memorized, and see how good you have performed on the task.
If you got neither of them correctly, do not worry. 7 seconds is quite fast, and, as you will see next, the capacity of Working Memory is limited.
If you did remember only the first one, though, how did you do it? Most people usually memorize it as three distinct subsets:
wxyz. It is relatively easy to remember it because of several reasons:
- instead of 12 units of information (characters, in this case), you had to memorize much less (3 letters and the length of the groups–4);
- you have inherent knowledge about the alphabet (if you, dear reader, are older than 6 years);
- you have advanced pattern recognition abilities that help you discern and group units of information into collections;
But what about the second sequence?
How did you attempt to memorize it? Maybe you have split the characters into chunks of 3 or 4 letters and tried to remember the groups, or perhaps you made words out of different adjacent letters and memorized those, or even you came up with a totally different strategy. Anyway, it is usually harder to remember the second sequence. And that is because the amount of information exceeded the capacity of our Working Memory. You had to memorize 12 units of information (or had to search anxiously for some associations to find a way to group those letters), while you have much less space to process at once in such a short period of time.
Let’s do one more similar exercise, but this time with numbers.
How did it go? You can verify your answers below.
There are two main approaches to store the first sequence of digits. The first is to combine them by 2, getting 19, 39, 19, 45, 19, and 69. You have 19 repeating three times, a 39 and 69 that are divisible by three, and a 45 that you just have to remember. But this approach is suboptimal. A more efficient one is to group them by four and use some knowledge of history. You can represent that sequence as three numbers:
1969. The first two are the start and end of the Second World War, and the third is the year of the Apollo 11 moon landing. So you just need to remember three chunks of information rather than 12. It is another demonstrative example that shows how our brain not only seeks to encode and compress information (as in the previous example with the first character sequence), but also it heavily relies on existing knowledge (as in the case with the WWII and moon landing years). The approaches you take may also vary in their efficiency.
The second one does not have any specific pattern to follow. Maybe you divided the sequence into groups of two or three or even four digits, or perhaps you tried to find a rule to group them. For instance, 82 add up to 10, then 64 add up to 10, but it’s not consistent. You’ll arrive at 92, and it doesn’t add up to 10. The approaches can be very different here, but the conclusion is the same - the second sequence is harder to remember because there is no exact rule or compression algorithm that would facilitate the remembering.
The Working Memory Capacity
We saw that it is hard to memorize a sequence of 12 items, which implies that its capacity should be smaller than that, as otherwise, you would effortlessly remember those 12 pieces of information. But how
big small is it?
The first quantification of the working memory capacity was provided in 1956 in a paper by George A. Miller titled: “The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information." It is one of the most highly cited papers in psychology2 with over 33400 papers that cited it (at the time of writing), according to Google Scholar. In his paper, Miller claims the working memory capacity to be 7±2 “chunks,” to which I’ll get shortly.
But since 1956, the Working Memory was actively researched and studied, and other theories of its capacity emerged. One such prominent hypothesis is that of Nelson Cowan, who in 2001 published the paper titled “The magical number 4 in short-term memory: A reconsideration of mental storage capacity," in which, you guessed, he suggests a smaller capacity of four chunks.
What Is A Chunk Anyway?
Since we can store a very limited amount of chunks in the Working Memory for processing, how does one define a chunk?
Unfortunately, there is no concrete answer to this question because it is a relative measure that depends on the type of information and the pre-existing knowledge of the person who operates with it.
It can vary greatly:
- If you are not familiar with the Russian language, the sequence
плюшечкаwould result in a chunk being the size of an individual letter (or even less), since you don’t have implicit knowledge about it. However, it would represent a single chunk for a Russian speaking person since it is a real word (in a diminutive form).
- Distinct words such as job, emotion, productivity may represent single chunks.3
- Combinations of words like “The Fellowship Of The Ring,” “The Two Towers,” “The Return Of The King” can also be considered individual chunks.
- Entire sentences can also represent chunks: “Today, I went to the seaside.” or “I am a responsible citizen, and I will vote.”
- When thinking about them holistically, even entire chapters and books may represent single chunks, as in the Lord Of The Rings trilogy mentioned above.
As you can see, there is no exact size for a chunk, as it depends on different factors, such as:
- familiarity with the topic;
- complexity of the concept;
- one’s mental agility;
But chunks are not represented only by words on paper or screen:
- Chess players store groups of pieces together when they are asked to reproduce a chess arrangement from memory onto an empty chessboard.4 Therefore, a chunk represents the position of several figures at once.
- Solving a simple math equation like
8 * 3 - 42 / 2 = ?requires an iterative approach to load the operands with the highest priority (8 and 3) and the operation to perform (multiplication), then solve that simple equation and store the result in the WM (24), repeat with next pair of operands (42 / 2), hold the result (21), and finally, repeat one more time, by accessing the results of the two simple operations in the WM and doing the subtraction (24 - 21) to get 3.
- Attending a meeting requires keeping different presented ideas in mind to reason about them, make connections with one’s own thoughts, or doing tons of other stuff with them.
Navigating The Hierarchical Structure Of Information Within A Complex Network Of Associations
You might wonder how a chunk may be represented by a single word as well as a whole book? After all, the amount of information differs immensely.
The idea here is how data is represented in our long-term memory and the agility of our Working Memory to combine information. The processes I have described run seamlessly and fast. When we read a sentence (not a random list of words), we expect a specific structure that we are very good at understanding. We know there is usually an action, an object, and a subject in a sentence.
So, when we read words, our Working Memory combines parts of the sentence into chunks and stores them in one of the four available memory slots. Then, once you are done reading a sentence, WM extracts all the relevant pieces that belong to the sentence you just read, combines them into a superchunk, and stores it in a single memory slot or even enter it into the long-term memory. Once you have read a paragraph, you combine several sentence superchunks into one, store it, and repeat the entire process. You thus create a hierarchy of information.
You can navigate it up and down. Say, you discuss the events described in the “Lord Of The Rings: The Fellowship Of The Ring” book with your friend. Right away, you load the most prominent events that you remember. For instance: the surveillance and chase for Frodo and other hobbits by Nazgûl, the events in the Inn Of The Prancing Pony where Frodo meets Aragorn, and The Council of Elrond. If your discussion switches to the details of the events in the Inn of the Prancing Pony, you might “zoom in” on that specific part and load more details about it, and so on.
As mentioned in the brain basics, our memories are encoded in complex neural circuits, parts of which are shared between different memories. This allows us to search among similar concepts and switch to those as well. For instance, when speaking about Legolas, the Elf archer, you might “zoom out” and talk about the elven race in the Middle Earth, after which you could compare Tolkien’s Elves with those from the Warcraft universe.
3 Rules To Increase Your Productivity
Since you have only four memory slots in the Working Memory, it is critical that you use them for the things that can push you closer toward your goal. Here are several productivity tips that will allow you to perform better.
1. Use Pen And Paper
In the thinking process and working process, make sure to optimize memory allocation. Since a lot of work we are doing today is based on abstract concepts (functions, patterns, and more in programming, concepts and ideas in copywriting, marketing campaigns, etc.), free up your Working Memory capacity by writing down the concepts you are operating with. This will offer you more capacity to manipulate with the relevant ideas rather than use some of the available capacity to simply store information that you are going to use later.
Besides optimizing memory usage, using a visual aid involves the visual sensory channel. That provides more opportunities to make better connections between ideas since the visual neural circuits have their own set of associations in place.
2. Operate On Manageable Chunks Only
Since not only the amount of memory slots is limited, but the size of the chunk is also finite, it is essential that you work on tasks that are appropriately sized. And if an assignment is too big for you to comprehend, consider splitting it into several smaller tasks until you get them of a size that you can easily accomplish. Huge tasks have too many details, and our Working Memory, with its limited capacity, does a poor job when trying to process it all at once.
Instead, divide the assignment into smaller, bite-sized subtasks so that you can focus and do it faster and more efficiently. This will also implicitly create a hierarchical structure that you can easily navigate and reason about.
If this sounds obvious to you, think how many people you know who actually jot down their ideas and create an action plan to follow.
3. Multitasking Is A Myth
With the newly acquired knowledge, I hope you can see why multitasking is extremely inefficient. To make it clear, though, by multitasking, I mean doing more than one task that requires cognitive effort. Doing many things on autopilot, like riding a bicycle and listening to music, does not have any penalties, since the Working Memory does not participate in that.
But when doing more than one task that requires your attention and though, as Working Memory has only four available slots (that fill up quickly), you are doing a lot of loading and unloading of information related to those competing tasks.
You load the relevant chunks to do Task 1, process them, store the result, unload them, and load relevant pieces to do Task 2. After processing those, you keep the result of Task 2, unload the information related to it, and switch back to Task 1. Rinse and repeat.
Thus, you don’t do these tasks in parallel, but instead frequently switch between them, spending a lot of time on loading, storing the intermediate results to keep track of the progress, and unloading the information.
To Be Productive Is Hard, Although Productivity Rules Are Simple
Our Working Memory’s limited capacity is a fundamental limitation that, unfortunately, we cannot overcome. It is a cornerstone of productivity that you know now how it works and are able to adjust your actions to be better.
But it is not the only thing that contributes to your efficiency. In future articles, I will dive into the other areas of how you function and how to improve your productivity. But until then, if you liked this article, make sure to subscribe below to receive new posts directly into your inbox and also follow me on twitter (@iuliangulea).
Although you can simply refresh the page and hack my extraordinary cheating-prevention mechanism. ↩︎
Interestingly, when researching the capacity of the Working Memory, scientists found that multisyllabic words (photosynthesis, polycarbonate, unimaginatively) are represented as more than one chunk in WM (Cowan, Baddeley et al., 2003). That is explained by the fact of how experiments are run. The tasks in these experiments do not leave people time to consolidate and rehearse long words, hence these long words consume more capacity. If, however, let people rehearse, long and short words perform the same. ↩︎
Nelson Cowan — Working Memory Capacity book, Chapter 3: Capacity Limits and the Measurement of Chunking, Subunit 1: Past Approaches to the Measurement of Chunking ↩︎