This paper presents an approach to understanding the relationship between a network of interpersonal influences and the content of individuals' opinions. The authors argue that social influence models should focus on the process of opinion formation rather than social equilibrium. They derive several existing models of social influence as special cases of their approach and explore implications for theories of social conflict and conformity. The authors propose a network paradigm for opinion formation, which includes inputs (exogenous conditions), outputs (settled opinions), and the process linking them. The process is divided into time periods, with opinions initially determined by exogenous variables and later influenced by previous opinions. The model is described by equations that show how opinions evolve over time based on the influence of previous opinions and exogenous variables. The authors discuss the implications of the model for social conformity and conflict, arguing that the distinction between conflict and conformity approaches is artificial. They show that the model can account for both types of processes and that the effects of social influence can be understood in terms of the relative weight of exogenous and endogenous factors. The paper also discusses various special cases of the model, including the linear discrepancy model, the peer effects model, and group consensus models. These models are shown to be consistent with the broader framework and provide insights into how opinions are formed and influenced by social networks. The authors conclude that the model has broad implications for understanding social influence and that it can be applied to both situations of social conflict and consensus. They argue that social conflict and conformity behaviors coexist in any group and that the relative importance of each depends on the initial opinions of group members and the nature of interpersonal influences. The paper emphasizes the need for further theoretical development and empirical testing of the model.This paper presents an approach to understanding the relationship between a network of interpersonal influences and the content of individuals' opinions. The authors argue that social influence models should focus on the process of opinion formation rather than social equilibrium. They derive several existing models of social influence as special cases of their approach and explore implications for theories of social conflict and conformity. The authors propose a network paradigm for opinion formation, which includes inputs (exogenous conditions), outputs (settled opinions), and the process linking them. The process is divided into time periods, with opinions initially determined by exogenous variables and later influenced by previous opinions. The model is described by equations that show how opinions evolve over time based on the influence of previous opinions and exogenous variables. The authors discuss the implications of the model for social conformity and conflict, arguing that the distinction between conflict and conformity approaches is artificial. They show that the model can account for both types of processes and that the effects of social influence can be understood in terms of the relative weight of exogenous and endogenous factors. The paper also discusses various special cases of the model, including the linear discrepancy model, the peer effects model, and group consensus models. These models are shown to be consistent with the broader framework and provide insights into how opinions are formed and influenced by social networks. The authors conclude that the model has broad implications for understanding social influence and that it can be applied to both situations of social conflict and consensus. They argue that social conflict and conformity behaviors coexist in any group and that the relative importance of each depends on the initial opinions of group members and the nature of interpersonal influences. The paper emphasizes the need for further theoretical development and empirical testing of the model.