Reviewed by
Juliette Rouchier
Groupement de Recherche en Economie Quantitative d' Aix-Marseille
This book is a collection of chapters that look at social norms from different methods, e.g., surveys, observations, experiments, modelling and simulations. Given that it is difficult to study social norms comprehensively from a unique perspective, this book helps us to realise the benefits of diversity of approaches, as well as that of the type of social norms that can be studied.Norms are often conceptualised in social simulation as an emerging phenomenon induced by the imitation between agents at a micro-level. In this book, the definition of social norm is more subtle in that it is linked to (1) the macro-feedback on agents who understand that the observed regularity is a norm, (2) the essential interactive nature of the norm formation and stabilization which relies on individual expectations about possible behaviour of other agents and what others expect an individual does, as well.
The introduction of the book, by the two editors, situates social norm research in a large set of recent studies and approaches. The first part includes chapters that provide empirical evidence and logical explanations about the complexity of norms. Chapter 2 demonstrates that, given that norms are not always discussed openly within a group, misunderstanding and focalisation on norms that actually are followed only by a minority can occur. This can be shown by surveys about «negative norms» (e.g., alcohol abuse). The discrepancy between reality and expectations can be high, either because people rely on (self-censored) reported behaviour or due to observation bias. The effect of this pluralistic ignorance can be reduced through discussion; Chapter 3 details situations where deliberation can help to make norms evolve to attain more desirable habits and expectations. In Chapter 4, the way an individual updates his/her beliefs on a norm is described within the cognitive framework of enaction. It shows the importance of looking at how individuals interact with the environment to perceive, possibly understand or revise a norm. Here, the complex question that relates micro and macro levels of social phenomena - how norms emerge and are recognised and internalised by agents - is discussed. This is also treated in Chapter 5, but referring to a multi-agent theory of cognition, BDI. Here, it is shown how we-intentionality is necessary to integrate norms as a cognitive emergence for agents. Chapter 6 deals with the complexity of norm enforcement at a group level. It shows that expectations about other’s expectations about sanctioning might create a self-fulfilling prophecy in that a norm can be abandoned because nobody thinks others wants to maintain it and protect it.
The second part of the book presents examples of applications, where norms are analysed in different contexts. Chapter 7 deals with norms in the case of the implementation of a tax system, which had to be iteratively implemented and then corrected, ranging from abstract principles to concrete record techniques for establishing a level of contribution. Chapter 8 uses simulations to show that cognitively elaborated agents, who can understand that a regularity is a norm, provide more robust norms than simpler, rational agents, thus stressing that cognition is an important part of normative behaviour. Chapter 9 describes a simulated work on an abstract norm, where the impact of peer-observation is crucial. Here, little observation can lead to norm disappearance, whereas large diffusion of information stabilizes a norm more efficiently: the relative perceived weigh of cheaters vs. followers has a considerable impact on subjective tendency to cheat, which in turn may influence behaviour of others. In the last part, Chapter 10 is positioned as a commentary of the whole book and goes back to sociological preoccupations, stressing the importance of social and reward structure (including reputational elements) for the spreading of the norm. It articulates well with the preceding chapter: «knowing that I can be seen» is as important as «seeing other» in the norm compliance.
To conclude, this book offers a very complete survey of different approaches to the understanding of social norms. Each chapter has been built scrupulously, with detailed references to recent literature. The multi-level definition of norms, in their interactive and dynamical nature (also stressed in the concluding chapter) provides interesting ideas to any modeller, especially on implications of the analytical scale to be considered. The last element, which is presented as central in this last chapter, i.e., context dependence, can be difficult to integrate straightforwardly in a model, but is certainly essential to decide why and what to model and how to interpret simulation results. In short, I believe this is a must read for anyone interested to have a comprehensive overview of social norms.
Return to Contents
of this issue