Neural Limitations of Social Relationships

Neural Limitations of Social Relationships

As our vast matrix of information and understanding about the universe grows, it seems that our significance is shrinking. Science and Philosophy have greatly pushed forward our cosmic insignificance, at times even arguing whether we truly exist at all. Our long gone ancestors did not seem to be preoccupied by such notions; they emerged as the king of the animal kingdom, be it in the savannah, waging war with the big five, or in the frozen tundra, where only fossils remain of the majestic mammoth. Indeed, what makes us so successful as a species is our social dexterity and communication capabilities. It wouldn’t be ludicrous to say that friendship is the strongest proof of our own existence; in some sense, friendships enable us to cement our essence in others and extend our being to frames of reference external to our own. Perhaps, the combined meshwork of the personal and extra-personal construct of who we are is what truly defines and validates our existence.

What is friendship? I will keep this question as broad as possible so as to avoid contradictions to arise later on. According to the Stanford’s Encyclopedia of Philosophy, friendship “is a distinctively personal relationship that is grounded in a concern on the part of each friend for the welfare of the other, for the other’s sake, and that involves some degree of intimacy.” This broad definition enables each of us to form our own personal interpretation of what friendship is. Another question to be asked is the following:  are we limited in any sense when it comes to how many stable social relationships we can maintain? It would appear so. Often we need to be reminded of some people’s names, birthdays and so on, a result of our finite memory. As such, could the root of our social limitations arise from biological, and more precisely, neuronal  restrictions? Could the cognitive burden of the extensive socialisation we impose on ourselves in this new age of social media consist of yet another form of overload?

Which Friend are you?

Let us study a conventional sociological and philosophical topic with the merciless and objective hammer of science. We’ll break the academic ice with a review of two “social numbers” : the Dunbar’s number and the Bernard-Kilworth’s number. Both studies present a theoretical number as an upper bound to the number of stable and continuous social relationships humans are “designed” to cultivate. What follows is a purely personal yet scientifically based account that doesn’t attempt to debunk these social statistics, but merely attempts to explain how our brains deal with relationships past the saturation level — for those interested in popular science rather than actual science, skip to the last section.



Dunbar’s number (1992): “Neocortex size as a constraint on group size in primates” [1]

Before reviewing the now classing paper “Neocortex size as a constraint on group size in primates“, it is important to understand who the man behind the paper is and why he was driven to the study of social relationships. He has dedicated his life to the study of primates and their social habits in order to better understand the roots of the complex social interactions characteristic of humans.

Let me begin by introducing the key finding of his paper : an elegant equation which has stood the test of time and vitriol of academia.


Here, Dunbar correlates the mean social group size of monkeys N to the mean neocortex ratio CR (neocortex volume / rest of brain volume), and surprisingly (or not, if you believe we are simply sophisticated monkeys) the expression seems to hold well for humans. What led Dunbar to make the link between neocortex volume — the most evolutionarily recent part of the brain, strongly associated with cognition and planning — and social group size, was his extensive study of behavioural habits and lifestyles of primates and sharp intuition.

At the time, two general theories had been advanced to explain the larger brain sizes of primates compared to the rest of mammals. The first posits that the large group sizes and complex relationships (dominance, affection, grooming, rivalry, adultery…) primates maintain is very demanding from a cortical point of view. Each relationship demands cognitive space and hence explains the larger brains observed in monkeys. The other suggests that the increase in cortical volume is due to ecological and lifestyle reasons. In other words, the nomadic lifestyle of primates requires them to keep a highly detailed cognitive map to navigate its way from one food source or refuge to another, while simultaneously keeping track of the states of previously visited ones.

Dunbar demonstrated that while both hold an element of truth the first is the real deal. Therefore, let us focus on the social restrictions imposed by brain’s limited processing capabilities. Bear in mind that the brain is the most needy of organs; at 2% of our body weight, it consumes roughly 20% of the body’s total energy consumption. On the other hand, biological evolution is perhaps the most conservative and wise strategist known to man, so why would it insist to develop such a demanding organ? In other words, natural selection doesn’t know the meaning of waste, why then does it appear as if the brain is such an energy glutton?

Dunbar’s working hypothesis to describe the neural space occupied by social relationships and consequently, the limit it imposes on the size of a primate’s social circle can be broken down into three main components :

  1. The number of individuals that an animal has to know and maintain relationships with.
  2. The necessity to keep track of third party relationships — in modern terms, gossiping.
  3. The nature of relationships themselves.

Accordingly, Dunbar decided to analyse information on a large number of primate species — humans included — and focused on the ratio of the neocortex  in comparison to the rest of the brain, mean social group size, and lifestyle / ecological information. The raw data can be found below.


The most important graph can be seen below: it is from this graph that the above-mentioned equation was derived. This is also where the now notorious Dunbar number of 150 for humans was first introduced.

Like any research paper, he concludes with a discussion of his findings; I will not report them here. Instead, I will mention some interesting arguments Dunbar brings forward in his book, “Grooming, Gossip, and the Evolution of Language” [2]. The first validation of the Dunbar number came comically enough at the special time of year when friendship is key — Christmas. Like a good scientist, Dunbar asked a number of his friends to keep track of the number of holiday cards they sent, because at the time it required a decent amount of effort to show commitment to your friends. Once sorting out the data, his very own special gift popped out : the average number of cards sent was 153.5.

Digging deeper, he studied Western military history, or more specifically, the size of companies — the smallest military unit. From the Roman’s centurion commanding a troop of about 80-120 men, to platoons of similar size in Napoleon’s imperial legions and so on, the number 150 was ubiquitous. Self-governing communities such as the Amish, Mennonites and Hutterites traditionally split up once their populations exceed 150. Similarly, many companies have begun to adopt and recognize the benefits of a strategy in which once a given branch exceeds 150 employees, a new office is built and the employees are split up into smaller offices.

When this number is exceeded, problems arise. In primates, violence is often observed as testosterone builds up. Members of the group may face brutal resolutions such as demotion, exile, and of course death.. For humans, we observe a similar build up of tension in mutinies in the military, professional backstabbings in large offices, uprisings in small communities, etc…

Do you still think it is pure coincidence that a simple equation aligns so well with the wisdom accumulated over thousands of years regarding the ideal social group size? Oh well, there is merit in trying at least, right?

dunbar graph


Bernard-Kilworth’s number (2001): Comparing Two Methods for Estimating Network Size[3]


In the interest of your time, I will only briefly mention the motivation behind this paper and its results. About a decade or so after the initial buzz around Dunbar’s number, Peter Killworth’s interest in “uncountable populations” —the HIV positive community, rape victims, heroin addicts and so on — led him to devise new methods (with the contribution of Russel Bernard) to formulate a number similar to Dunbar’s resulting in an alternative dubbed the Bernard-Kilworth’s number.

Through an extensive set of detailed surveys and mathematical blackmagic, two separate methods yielded a remarkably similar social number of 290. You may consult the paper if you are interested in the derivation of the Bernard–Killworth’s number; I’m an aspiring engineer, not a mathematician, so I don’t find this derivation aesthetically pleasing as Dunbar’s proved to be. Below is illustrated the key figure of the paper.

kilworth graph

The implications of this number is that it indeed turned out to play a vital role in accurately measuring the key socio-demographic issues (the “uncountable populations”) that concerned Killworth. It enabled governments to reliably estimate the scope of problems for which obtaining reliable data proved tricky. The question of whether governments subsequently addressed these issues is a tale for another day.


Validation/refutation of Dunbar’s number:


  1. Modelling User’s Activity on Twitter Networks : Validation of Dunbar’s Number.” [4]
  2. Dunbar’s Number: Group Size and Brain Physiology in Humans Reexamined.” [5]


In the first paper, it is demonstrated through the analysis of user’s activity on Twitter that while number of “friends” we may have on social networks far exceed Dunbar’s number, users only entertain a range of 100-200 stable relationships. On the other hand, the second paper consists of a less scientifically intense review on Dunbar’s number and generally, the disapproval of the authors to reduce the complex phenomenon of social relationships to the size of the walnut in your head.


Personal Interpretation on our seeming capability to far exceed these social numbers in today’s overloaded society:


A note to the reader, we will now detract from a scientifically based argument and discuss a scientifically-grounded yet unfounded interpretation of these results.

First of all, we all have a lot of friends on Facebook these days. I am fairly certain that more than 80% of you have more than 150 Facebook “friends” (“friends” is a term used lightly here).I’m also fairly certain you can develop sufficiently elaborate archives / memories of people you interact with on a regular basis, thereby exceeding this Dunbar’s number — at least for the more extroverted amongst you.

So how is this possible? Well, this is where I believe the brain starts to use tricks to efficiently classify what quickly becomes an informational overload on the brain’s processing capabilities. Again, the brain is the most intense property market in existence. If a neuron or even a neural network isn’t used extensively or deemed worthy of existence by neural regulatory mechanisms, the individual neuron or system will be pruned in cold blood ( Donald Trump’s “You’re fired” ain’t so bad now is it?!). This unforgiving competitiveness means that the brain has, over the course of evolution, constructed parsimonious solutions to problems it can’t conceive, at least not in the way we do.

While memories are certainly not as fully understood as some popular science articles would lead you to believe, memories are certainly plastic. They can be modified, duplicated, and cropped in a fashion that is reminiscent of Photoshop. Therefore, it is conceivable that for any friends past the theoretical limit defined by Dunbar, cognitive tricks might be used to “compress” the information of very large social networks, much like a zipped file helps reduce the required memory to store large amounts of information.

Imagine that your 150 closest friends are given a neuronal pedestal, and indeed occupy meaningful cognitive space unique to the individual, what about the rest? It is not absurd to suggest that the brain copes with this oversaturation of information by classifying the less meaningful relationships into categories. For example, you will have the cognitive label of geeks, party animals, bombshells, awkwards, Frenchies, friendzoneds, and so on. Consciously or not, we have an inherent tendency to label people who share a similar pattern of features into a common bin.


To socialize, or not to socialize? That is the question…

How might this be an efficient way of storing information? Simply put, it would be as if you used the same broad collective memory label and apply it to the individual in question. Of course, it’s not this simple as the face or voice of an individual (a unique stimulus owing to the uniqueness of a person’s features) would have to somehow trigger this palette of vague memories and add a few layers of transient information, transient because it is often forgotten (hence why we often need to remind ourselves about general facts about the people we rarely interact with).

Before I conclude, I must seize an opportunity too golden not to exploit, what of the lavish socialites inhabiting their castles high up in the skies? While many of the stars and idols of modern society are undoubtedly geniuses in their own rights, a considerable amount are not — at least in my books. Paris Hilton or the Kardashians — the list is not very extensive as I do not wish to spend much time studying these anomalies — are well known to have massive social circles, and in my opinion probably have an off the charts Dunbar’s number. But fear not, this is no praise of their social nimbleness. I believe the brain is highly malleable. Accordingly, its owner has the somewhat conscious ability to choose how much cognitive space is dedicated to certain pursuits or skills. In the case of the above mentioned socialites, have you ever heard them say anything intelligent or inspiring? But I digress, I believe some willingly sacrifice intellectual pursuits in order to metamorphose into a social butterfly — with an IQ similar to the insect’s.

The brain, in its ruthless pursuit of efficiency, simply doesn’t feel the need to maintain such an accurate representation of an extensive set of individuals, and rightfully so. Perhaps what makes friendships, the real ones, so meaningful is their inherent scarcity. Just as our own mortality forces us to find meaning during our ephemeral passage on Earth, limitations are critical to bring out the real worth of various phenomena.

Besides, in this new age of information, and constant scrutiny of our peers via social networks, the idiom ‘less is more’ finds ever greater significance.

Written by Bernie Mak

[1] Dunbar, R. I. (1992). Neocortex size as a constraint on group size in primates. Journal of Human Evolution, 22(6), 469-493.

[2] Dunbar, R. I. (1996). Grooming, gossip, and the evolution of language. Cambridge, MA: Harvard University Press.

[3] McCarty, C., Killworth, P. D., Bernard, H. R., Johnsen, E. C., & Shelley, G. A. (2001). Comparing two methods for estimating network size. Human organization, 60(1), 28-39.

[4] Gonçalves, B., Perra, N., & Vespignani, A. (2011). Modeling users’ activity on twitter networks: Validation of dunbar’s number. PloS one, 6(8), e22656.

[5] De Ruiter, J., Weston, G., & Lyon, S. M. (2011). Dunbar’s number: Group size and brain physiology in humans reexamined. American anthropologist, 113(4), 557-568.


Bernard Mak is a fourth year Mechanical Engineer minoring in Neuroscience at McGill. His interest in science and technology has only grown over the years, with past contributions with McGill Robotics and having helped create Graphite's Sciences section. His goal is to share his scientific knowledge and passion to the rest of the community through Graphite.