This phrase, which we have probably heard more than once, takes on a special meaning today when we look at it through social psychology and the study of human relationship networks. It becomes especially revealing when we try to understand how balances are built, maintained, or broken within a group.

Balance, coherence, and social relationships

Fritz Heider, an Austrian psychologist and a key figure in the Gestalt school, is widely recognized for his contribution to social psychology through attribution theory. He was also the author of the balance theory, which argues that people tend to maintain coherence among their attitudes, affects, and relationships.

According to this theory, it is common for a person to agree with someone they like and have a positive bond with, just as they tend to disagree with someone they find unpleasant. This behavior is not arbitrary, but a way of seeking stability and coherence both in our relationships and in the way we perceive our environment.

Other influential social psychologists, such as Theodore Newcomb and Harold Kelley, later expanded Heider’s ideas and applied them to the field of communication. Their work deepened the study of attribution, interaction, and, more broadly, the understanding of interpersonal relationships.

Within balance theory itself, Heider also noted that not all relational configurations are equally stable. Some social structures tend naturally to persist over time, while others generate tensions that lead to changes or breakups.

When theory is observed in real networks

These ideas were reinforced decades later by a study by Michael Szell, Renaud Lambiotte, and Stefan Thurner, disseminated by Imperial College London, titled Multirelational organization of large-scale social networks in an online world. In it, the authors showed that positive relations exhibit a stronger tendency toward reciprocity than negative relations.

The study analyzed interactions among more than 300,000 users in an online game called Pardus. Six types of interactions were identified. Communication, friendship, and trade were considered positive interactions, while hostility, punishment, and aggression were classified as negative interactions. Each interaction formed its own network, and the combination of all of them produced a large social network.

Some conclusions were intuitive, yet had never been observed at this scale. For example, there was a clear overlap between communication and friendship networks, which is logical if we consider that friends tend to communicate with each other. By contrast, there was no overlap between networks such as hostility and trade.

Similarly, when one player requested friendship from another, it was usually reciprocated. However, when a player declared another as an enemy, the action was rarely reciprocal, which weakened the relationship and affected the overall stability of the network.

These interaction patterns show that individual attributions and the positive or negative nature of ties are key factors in the stability of a social network. It is these ties that strengthen or weaken it over time.

Why we seek balance

At the core, we seek balance because it feels more comfortable. These situations arise when we agree with a friend or when we disagree with someone we dislike. That is why it is more common to assume that the enemy of my enemy is also my enemy, and less common to consider a friend’s friend as an enemy.

We seek balance because it feels more comfortable.

The need for coherence and balance not only explains many of our individual decisions, but also helps us understand how social networks are organized and evolve, both in the physical world and in digital environments.