Social network analysis
Social network analysis (SNA) is the interdisciplinary study of social structures through the lens of network theory. It treats individuals, organizations, or other social entities as nodes, and their relationships — friendships, collaborations, conflicts, communications — as edges. The fundamental premise of SNA is that social structure is not merely a backdrop for individual action but an active force that constrains and enables behavior. Who you know, and how they are connected to each other, shapes what you can know, what you believe, and what you can achieve.
The field draws on sociology, anthropology, psychology, mathematics, and computer science. Its mathematical foundations — graph theory, matrix algebra, and statistics — provide tools for measuring properties such as centrality, clustering, assortativity, and community structure. But SNA is not reducible to its methods. The interpretation of network measures requires domain knowledge: a high centrality score means something different for a political lobbyist than for a protein in a metabolic network.
SNA has produced some of the most influential findings in the social sciences. Mark Granovetter's 1973 paper on the strength of weak ties argued that infrequent, bridging connections between social clusters are more valuable for information diffusion and job searches than dense connections within clusters. Ronald Burt's work on structural holes showed that individuals who bridge disconnected groups accrue disproportionate power and reward. Duncan Watts and Steven Strogatz's small-world model demonstrated that even highly clustered social networks can have surprisingly short path lengths, explaining the famous six degrees of separation phenomenon.
The modern practice of SNA is inseparable from computational methods and large-scale data. Social media platforms, mobile phone records, email archives, and bibliographic databases provide unprecedented resolution for mapping social structure. But this abundance brings new challenges: privacy, sampling bias, and the confounding of online and offline relationships. A Twitter follow is not a friendship; a citation is not an endorsement.
Social network analysis is often criticized for being mere description — for mapping structures without explaining why they form or what they mean. This criticism misses the point. The description IS the explanation, at one level of analysis. To show that a job market operates through clustered, homophilous networks is to explain why meritocratic outcomes are structurally impossible. To show that misinformation spreads through low-clustering, high-centrality pathways is to explain why fact-checking fails. SNA does not replace causal explanation; it reveals the structural conditions under which certain causes become effective. The network is not a metaphor. It is a mechanism.