Our social network defines our lives much more than we think. It shapes our views, drives our actions, and influences our behaviour. But most importantly it enable us to discover ourselves through our relationships with others.
Through analysing social networks we can reveal patterns of relationships that define our influence and role in relationship to others. Social network analysis has become a key field in sociology since the 1960’s, long before the proliferation of social media.
Tell me who your friends are and I will tell you who you are
Social Network Analysis uses graph theory to represent people or actors as nodes and their relationships as ties or links. It also uses network modelling techniques to investigate the complex social dynamics within networks, and how people interact with each other. Governments use it to understand criminal gangs, how information diffuses across a population, and how communities work together.
On an organisational level, a common mistake is to confuse organisational structure with social and communication networks. Many business leaders waste a lot of money and effort in creating particular organisational structures, thinking they can control communications. But this is rarely the case as informal social networks tend to circumvent artificial silos and hierarchical structures, and reestablish damaged connections.
Key characteristics of a social network
A social network is visualised with graphs. The nodes of the graph represent the actors, which can be individuals, communities, companies, countries or any group of people. The links or edges represent the social relationships between the actors.
The social connections between actors represent friendship, influence, affection, trust, working together or the opposite, conflict and dislike. There can also be multimode relationships between different types of actors, such as between individuals and teams or corporations.
Relationship symmetry
Social relationships can be either symmetric or asymmetric. In symmetric relationships there is an equal relationship between two actors. For instance, I am friends with Tim and he is friends with me. Another example is I am connected with you on LinkedIn and you are connected with me as well. Symmetric relationships are represented with an undirected line and are usually much more stable.
On the contrary, in asymmetric relationships there is a directionality in the relationship. For instance, the relationship between a teacher and her student is “directed” and asymmetrical. This is also the case in friendships or romantic relationships where there is lack of mutual affection. Asymmetric relationships are represented with a directed line and are usually less stable.
Relationship strength
Another key characteristic is the strength of the social relationships or connections.
According to Tsvetovat, M, & Kouznetsov, the frequency of communication between two people reflects the emotional content and amount of influence in their relationship. When two people communicate often, this indicates a stronger tie than when they communicate infrequently.
Of course, there are exceptions to this. But the authors found that asking how frequently two people actually talk to each other reveals more about the strength of their relationship than asking how much they like each other.
Based on the strength of the relationship, there are two ways we connect with each other, strong or weak ties. Strong ties are usually the connection we have with close family, colleagues, teammates and friends. While weak ties are usually the connection we have with distant relatives, acquaintances, and friends of friends.
Mark Granovetter, Professor of Sociology in Stanford, has written one of the most widely cited articles in Sociology, called the the strength of weak ties. In his article he describes weak ties as social connections between people that require little or no emotional attachment, do not communicate often and require little or no personal time and energy to maintain. People that connect with weak ties can often be quite different and strongly disagree. However, they still maintain a basic common ground and a level of curiosity that prevents them from engaging in conflict. Granovetter pointed out eloquently how weak ties are key to innovation and for passing novel information across people.
On the other hand people with who we are already very close with do not offer many new opportunities compared to weak ties. This is why many new discoveries in science are interdisciplinary in nature. Hence, the paradoxical strength of weak ties.
Triad as the basic building block
Another key characteristic of social networks is the tendency of actors (people, groups) to form triads. This is the smallest group of people that we can call a team, as many experts argue that dyads are not a group. In a dyad each person remains an individual and there is no concept of group. But when a third party is added, then group dynamics emerge. The third person acts as a source of balance, providing a different view,and acting as a feedback loop and mediator. In triads, apart from the direct relationship of two members there is indirect relationship through the third party.
As Tsvetovat, M, & Kouznetsov say in their book, triads over time create their own local culture by developing artefacts, such as jargon, shared rules and stories, that are private to the triad. Furthermore, it seems that triadic structures are the most stable over time. They also require less effort to maintain compared to bigger or smaller structures. As such, they seem to be better building blocks for larger networks than dyads.
Types of triads and how they affect the network structure
Here I will explore the two main types of triads, open and closed, based on whether all three parties connect with each other or not. However, when we take into consideration relationship asymmetries there can be up to 16 different combinations of triads.
Closed Triads
Tsvetovat, M, & Kouznetsov in their book researched terrorist cells. They discovered that most cells comprise mostly of 6 people with a dense triadic structure. This means that everyone in the cell is in triads with everyone else, connected by strong ties of loyalty, friendship, trust and a powerful sense of belonging. This type of triads that have equivalently strong ties and fully connect a group are called closed triads. In turn, several overlapping closed triads can create a clique, inheriting many of the culture-generating properties of closed triads.
Open Triads
On the other hand, there is the open triad, which has two ends that do not communicate directly. This is also called the forbidden triad or structural hole, as two of the three members are not allowed to communicate with each other. An example of this are love triangles or middlemen that benefit from the asymmetry of information.
Open triads can create high levels of stress for the person who is in the middle of the triangle. But also create a significant higher rate of success in a competitive marketplace due to its asymmetry.
Overall, when there is a higher ratio of forbidden triads in a network this indicates a more hierarchical network structure. On the contrary, when there is a higher ratio of closed triads in a network this indicates an egalitarian structure.
Becoming Viral – Transformation of Triads
The relationships that exist within a triad and how they transform over time are key in understanding the formation of larger networks. It all starts with a triad as the basic building block.
As relationships start forming in a network, open triads become closed. Every closure sparks even more connections, creating new open triads. Of course, the creation of new connections depends on the novelty-seeking or novelty avoidance tendencies of individuals. There is a spectrum within individual are curious enough to talk to each other and establish a connection.
When they are too similar or extremely different the probability of a connection is less likely. Forming many new connections has a knock-on effect inside the network. The density of the network grows exponentially until every node is connected to every other node. However, through the growth described above, the probability of conflict also increases and spreads through the network.
Earlier, I talked about strong ties. However, a strong tie may indicate either friendship or conflict, depending on the circumstances. In addition, this relationship between two actors is not static, and can change dynamically over time. For instance, someone who is friends with two others who are in conflict may have to pick sides. When this happens, the conflict increases. Alternatively the third party may act as an intermediator and resolve the conflict. Therefore there is a constant evolution of relationships over time that propagates within a network.
Eventually the density of the network (closed triads) reaches a critical mass which is constrained by the emerging conflict. The network never becomes fully homogenised as different opinions, within an acceptable range, coexist in a dynamic equilibrium. This kind of behaviour is called self-organised criticality.
Scaling up your social network
As I discussed in my previous article, our cognitive limitations influence the number of ties we can maintain. But they also define the thresholds over which scaling up creates significant communication complexities.
British anthropologist Robin Dunbar suggested that there is a cognitive limit of around 150 people, with whom we can have an active social tie. In some cases this number may be a bit higher or lower. But the reality is that our brain’s prefrontal cortex limits the number of people we can interact with. After this cognitive limit a qualitative shift happens.
Nevertheless, this is not an absolute limit. The development of social media have allowed us to extend our social networks to hundreds or thousands of people. But if I would ask you whether you actively keep in touch with all of your LinkedIn connections, your answer would probably be no. It is just impossible to keep actively in touch with so many people.
Another interesting concept is the horizon of observability. According to Noah Friedkin, horizon of observability is:
“The distance in a communication network beyond which persons are unlikely to be aware of the role performance of other persons”.
Noah Friedkin
As Tsvetovat, M, & Kouznetsov mention we are good at knowing who our friends are (around 70%). We are less good in knowing who the friends of our friends are (only 30%). While we almost have no knowledge of friends-of-friends-of-friends. In other words, there is a limit of how far we can see inside our social network as we tend to be restricted to our direct links or those with whom we have at least one common friend.
Despite the above, there are interesting connection patterns that enable us to reach far away within our networks.
Types of Networks
As social networks form, expand, and dynamically adjust, specific types of networks tend to emerge.
Paradoxically, the networks that dominate the social network landscape are small world networks. These are dense communities that keep a local neighbourhood structure but which also allow a small number of ties to reach far away through individuals who act as connectors. In small world networks someone is never too far away from each other and everyone can reach each other through a short sequence of acquaintances. Even more interestingly, we are good at finding shortcuts in the network and reaching far away without having to scale up the network structure.
In popular culture this became known as six degrees of separation and was based on an experiment in the 1960’s by the social psychologist Stanley Milgram. Milgram showed that we are all no more than 6 social connections away from each other, although this has been challenged by other scientists.
The Ego and Star Networks
The Ego Network is the social network that surrounds us. In other words it is a network that shows all the connections going out of a single node, us. As discussed earlier, our ability to see inside our network is about 2 levels deep, based on our horizon of observability. This means we know friends-of-friends but almost have no knowledge of friends-of-friends-of-friends.
There are two types of ego networks, high dense and low dense networks. In high dense ego networks, your friends are also friends witch each other and have high mutual trust (high clustering coefficient). This resembles the overlapping closed triads (cliques) mentioned previously. It also shows that there is a high degree of homophily, our tendency to bond with similar others.
On the other hand, in low dense networks, the people connected to the central node do not know each other (low clustering coefficient). We call this type of Ego Network as a Star Network. An example of this is when a celebrity broadcasts to their followers on Twitter, who are mostly strangers to each other.
Social network connection patterns
There are many interesting connection patterns in social networks, based on how individuals and groups are connected, how close they are together, and how influential they are.
Below, I briefly present the key metrics that measure the importance of an individual inside a complex network (centrality).
Degree Centrality
Measures the number of links a single actor (node) has. The higher the degree the more central the node is in the social network.
Closeness Centrality
Measures the distance or closeness to others. The closer the distance of a node to others the more easy it easy for that person or actor to spread information across the social network, influencing the perception of the world inside the network. A more common word for people that have high closeness centrality is gossipmongers.
Betweenness Centrality
Measures the influence of an individual over the flow of information across people or communities who otherwise would not be able to interact with each other. This person acts as a bridge or a bottleneck depending on the role they want to play in the network. They are also called boundary spanners or brokers and diplomats, bridging structural holes.
For instance, UN negotiators are key in making peace agreements between communities that are in conflict.
Eigenvector Centrality
Measures the level of influence of a node within a complex network. This metric reflects your influence which is defined by the importance of the people in the network you are closely connected with. You can be very close with a large number of individuals. But if they have low influence, then this means that you cannot be too influential either.
Google’s PageRank algorithm is a variation of this, as it defines the importance and authority of a webpage based on the number of links it gets from others.
Final Thoughts
I always used to hate professional networking events. I always thought that it was extremely awkward having to randomly talk to people I didn’t know anything about. However, my outlook on this changed completely when I decided to start salsa dance classes, after my trip to Cuba in 2017, where I fell in love with its music and culture. I was fascinated by the Cuban style of dancing, which was very social, fun and inclusive, and was the catalysts that was bringing communities together on the island.
Through dancing, I gradually overcame my shyness and realised the hidden power of connecting with people I had never met before. From being shy to talk to strangers I was now comfortable to dance with them! Gradually, I saw my social circle significantly expanding. A whole new world of people, who had nothing else in common apart of their love for Cuban Salsa, opened up in front of me. It was one of the most exhilarating moments of my life and made me realise the endless possibilities that come from socialising with people we don’t know. I was experiencing the strength of weak ties, as Granovetter described in his seminal article.
We tend to focus on close friends and relationships and usually not pay too much attention outside our ego network, thinking that our social world is largely static. However, in reality the world is extremely dynamic. We need to be open and curious to new acquaintances, especially those that are not like us, as this has the power to enrich our lives, bring unexpected opportunities and in the process help us discover who we really are.
References
Tsvetovat, M, & Kouznetsov, A 2011, Social Network Analysis for Startups : Finding Connections on the Social Web, O’Reilly Media, Incorporated, Sebastopol. Available from: ProQuest Ebook Central. [3 May 2021].
