Cognitive Bandwidth

Another word of power that I have learned to master in the last two years is cognitive bandwidth.


Cognitive bandwidth is the maximum amount of thinking that’s available per unit of time. It’s pretty close conceptually to bandwith in telco networks: there’s the actual amount of thinking done per unit of time, which is the cognitive throughput, and then there’s the maximum possible, which is the cognitive bandwidth.

In the same way that it is hard to understand the concept of network bandwidth when one has never reached it yet (for example, someone with a slow computer or a simple telephone connected to fiber doesn’t use enough bits to really feel there’s a limit), cognitive bandwith is only a concept once experiences first-hand after a good amount of life experiences that require regular/continuous thinking over a significant period of time.

I first became aware of cognitive bandwidth a couple years ago, when I noticed that I would forget what I was thinking about earlier in the day after I’d have worked on multiple projects throughout the day. This was new, because until then I would pretty much go to sleep every evening with a very clear recollection of every conversation and every thought I had during the day. In fact, in more quiet days (including “down” week-end days) I was still able to do that, so it was not like my cognitive ability was lowering with age. After a period of transition, when I was slowly becoming aware of this memory issue, I went to more or less consciously experiment with different activities to gauge how this worked. It turns out that if I focus on a single complex task for a fixed number of hours per day (say, 6), I’ll remember all of it afterwards; whereas if I focus on three complex tasks for the same number of hours total (say, three times 2), I’d forget about the details of the first task by the time I completed the third.

At that point, that understanding was vague, semi-subconscious, and I did not yet possess the true word to designate what was going on.

Instead, I figured out soon that I could take notes during the day about what I was doing and what was left to do, and use these notes to keep track of complex tasks over time. So I started doing that.

Without realizing this consciously, I had expanded my cognitive bandwith a little with my notes.

Some time afterwards, I also found out that I couldn’t keep too many notes each day, because without extra work the notes became too busy / complex to understand the day after, and if I did spend extra work to take good notes then I would spend less time on the tasks (lest work hours spill on the rest of my life). Again, without realizing, I was experiencing first-hand the cost of cognitive overhead - things that cost me cognitive bandwidth but do not contribute anything useful.

Without techniques to expand my cognitive bandwith further without too much effort, I had to face this hard fact: there was a limit, and I had reached it. Every time I’d want to add something new/different to think about on a given day, even if I was taking something else out of my day to make time, I still couldn’t perform as well on everything because the increased diversity of tasks was reducing my cognitive bandwidth by adding cognitive overhead.


The formula is simple really:

cognitive throughput + cognitive overhead < cognitive bandwidth
  • the bandwidth is fixed at some level which I can barely change.
  • the throughput depends on the complexity of the stuff I am working on.
  • overhead is the cost of doing stuff where I have to care about organization, task switching, etc, and is paid against the bandwidth limit.

From that point on, I’ve been playing a game of balance.

On one side, I’ve been more and more aware of this limit and accepting that there are things I find interesting and for which I do have enough time available each day but which are so different from the other things I already fit in my day that I do not have enough cognitive bandwidth for them because either I’m hitting the bandwidth with my useful throughput, or the overheads are too high. So there will be days when I will simply not do much in terms of hours, so as to preserve cognitive bandwidth for one or a few activities that are relatively complex.

On the other side of the balance, I am also working on little ways here and there to increase my cognitive bandwidth and reduce cognitive overhead.

Notes, as I already found out many years ago now, work well. With years I have learned to make my notes more effective: shorter so they don’t cost me much to make, yet dense and explicit so that they enable me to pick up where I was, later, without having to think about them in the meantime. Subjectively, it feels like notes bought me a solid 20% increase in bandwidth.

For a couple of years now, I have eliminated time management entirely from my head and delegated it fully to a couple digital agendas. This was initially only a small improvement because my agenda was very sparsely allocated. Now that I juggle between more than 20 appointments per week, having to keep track of them myself would penalize me cognitively at least 15%, so that’s how much overhead I understand it buys away from me, and that’s probably bound to improve further.

Keeping track of expenses, financials, taxes etc. is now delegated to various documents and an accountant, and so I still only spend an hour or two every three months, just like I did when I started paying attention fifteen years ago. It may seem a lot, but considering that the growth in the number and complexity of the items and transactions has been at least geometric, if not exponential, I am pretty happy about the reduction in overhead.

Meanwhile, there are lifestyle choices and changes that have had beneficial influences on my cognitive bandwidth or overhead beyond the reasons why I made them.

The absence of a television, a car, facebook, newspaper subscriptions and a couple of other similar things from my life are time savings and emotional boons but also a significant reduction in cognitive overhead.

Healthy and regular food have increased my health and thereby my cognitive bandwidth, I feel by a solid 10% so far. Not spending time for groceries and cooking other than in social situations has reduced my cognitive overhead, I feel by a solid 10% too.

Having to translate back and forth between the domain of ideas and concepts and human language form a definite cognitive overhead. I place it at around 20-30% every day. Mine is probably higher than most folk because using multiple languages on the same day costs extra cognitive bandwidth, and my technical activities require both nuance and semantic breadth which require a large “paged in” set of linguistic reflexes, forming an overhead of its own. I have made this lifestyle choice actively and consciously and I am still OK with it because of its positive consequences, but I am painfully aware of its cognitive cost. Meanwhile, other linguistic abilities had to give way to liberate me from the associated overheads—I probably have lost my ability to express myself formally in anything else than English, and I have dropped from my actively maintained skills my ability to remember and properly use the gender of common nouns in Dutch and French.

Other activities and topics of interest are accumulating with age and life experiences and come with their own overheads, which I must learn to offset. I have few dependents and no children among them, so thankfully none of the associated cognitive costs to bear. However, I do have a growing social circle of friends and acquaintances that I think about and interact with regularly; in addition, since they come with a relatively diverse set of cultures, social-economic backgrounds, languages and interests, I cannot reuse interactive reflexes readily, and so the complexity to navigate these various interactions come with their own significant overhead.

This costs me much, probably the largest overhead by far, and more, I reckon, than the average folk due to my handicap; but again, the emotional payoff justifies the cognitive expense.

The really challenging bit, here, is that this expense is not only large, it is also growing because I am investing in those aspects of my life more lately, and so it became to compete for cognitive bandwidth with productive (work-related) cognitive throughput.

Last year, despite optimizing overheads away and maximizing bandwidth as much as I could, I saw the two collide unpleasantly—times where I didn’t feel comfortable spending sufficient time at work, not because I was physically tired or bored or challenged, but because I felt I would be otherwise cognitively too committed (“my head would be too full”) to properly interact with important people after work hours. And other times where I had to say to friends “no, we’re not meeting today” realizing that I was also thinking “…even though I have time and I want to see you, because our conversation is too stimulating and if we talk I won’t be able to focus on this important project at work afterwards.”

This balance the hardest adulthood problem I have had to solve so far.