How did social science help us with understanding the brain’s connectivity pattern?

By Dr. Kayson Fakhar.

Let’s start with ‘networks’. They build useful intuitions for otherwise complex interactions all around us. To unpack, a network is simply a set of interconnected ‘nodes’. This broad definition allows us to make networks from almost anything, for example, social networks in which nodes are people and connections are their relationships, gene interaction networks where nodes are genes, brain networks modelling anatomical connections among brain regions, or even geopolitical networks that capture how countries interact. The connections between pairs of nodes usually come with different flavours too. Some are binary, meaning that they either exist or not: In a simple friendship network, we are either friends or we don’t know each other. Some are weighted: we can put a number on our friendship, from strangers to best friends forever. Some are signed: we can be enemies too, encoded in negative values. They can also be directed: I follow you on social media but you don’t, as opposed to symmetric where we both have to accept a friendship request to be able to interact. 

Making networks and studying them is not new, in fact, social science can be seen as a pioneer of ‘network science’ as some seminal works were done to find the structure of human societies. For instance, a series of works done by Stanley Milgram led to the idea of ‘small-worldness’ as they showed that very few intermediate hops are needed to bridge two strangers. Specifically, in one work they randomly selected about 300 individuals in Nebraska and Boston. These people were instructed to make a chain of people to a target person in Massachusetts. Interestingly, around 65 chains managed to reach the target person and the average number of intermediate people was a little smaller than 6. You can play around this concept using the website oracleofbacon.org in which you write down two names and see how they can be linked via intermediate Hollywood stars. According to the website, for example, Barack Obama could reach Elon Musk via only one hop, i.e., Liam Neeson since Obama and Neeson both appeared in the movie ‘Battleship’ and Neeson was in ‘Men in Black’ with Musk! That aside and even more interestingly, about 50% of chains made in Milgram’s experiment passed through the same three persons before reaching the target. Put differently, these three well-connected and ‘central’ people were essential for routing information in that experiment. Without them, possibly the chain would have been broken or would end up far longer than it was with them. 

Long chains might not be a problem in this experiment but it surely is in the brain. It should be no surprise by now that many think brain networks have a similar pattern of connectivity (i.e., small-world connectivity), which allows them to exchange information efficiently. What is surprising though is the fact that the average number of intermediate nodes from any brain region to another seems to also be around 6. This means, on average, information can reach from any brain region to another by just hopping over 6 regions in between! To put this number into perspective, the population of the US back when Milgram conducted the study was about 200 millions. Routing information only via 6 people is certainly impressive, however, there are about 86 billion neurons in the human brain, yet the number is still 6, which to me is a much larger achievement. Think of the potential signal degradation that could happen if this number was larger. As impressive as the brain is, we should remember that it’s a wet, messy, and expensive organ. We tend to compare modern AI models with the human brain, reminding everyone how efficient the brain is. What we often overlook is that the brain consumes a lot more energy compared to other organs with a big chunk of it going directly to sending signals from a region to another. In fact, some researchers estimate that communication among neurons consumes 35 times more energy than computation, so no wonder it did everything it could to bring this cost down. 

That said, why 6 intermediate hops and not any other numbers? This number is already leaked to our collective vocabulary as the ‘six degrees of separation’. A possible answer to this question came from a recent theoretical work, with an elegant explanation: six degrees of separation naturally emerges from the desire of every node to be central, given the cost of maintaining its centrality! What this means is that, both in the brain and in human society, most nodes tend to maximise their connectedness. However, keeping up with those connections comes with a cost (not everyone can handle a large group of friends, I’m sure many of you can relate!) and balancing these two leads to having an average of six hops from any node to another in the network. 

There are so many more examples of measures that were designed for analysing social networks but ended up also useful for studying the brain. However, I think this one is particularly interesting because it hints at core principles shared by the human brain and its users. Because of that and specifically nowadays, science needs to be evermore integrative and multidisciplinary. Networks give us an intuitive way forward in this direction. Who knows, maybe one day we find even more similarities between human societies and brain networks. Maybe we find that we are, collectively, one planetary brain all along. I’m sure our brain cells are as clueless about what role they play in the brain as we are on earth.