© Copyright JASSS

JASSS logo ------

Diversity and Complexity (Primers in Complex Systems)

Page, Scott
Princeton University Press: Princeton, NJ, 2011
ISBN 9780691137674 (pb)

Order this book

Reviewed by Cesar Garcia-Diaz
Faculty of Applied Economics, University of Antwerp

Cover of book Writing a book with an interdisciplinary approach is a daunting undertaking. Trying to convey different concepts that might look familiar across disciplines puts an author at risk of inconsistencies, but Scott Page has succeeded in carrying out the remarkable task of setting out a set of common knowledge pieces to different audiences while dissecting concepts that are difficult to deal with like diversity, complexity and their interplay. The book draws heavily on biology but also refers to many examples in social systems (economic, political, organizational, innovation systems).

Without falling into details of specific disciplines, or reviews of empirical works, the author speaks about the fundamentals of diversity and introduces them through a number of thought experiments. As the author himself has uttered earlier (Page 2007, 87), "by boiling down causes and effects to their spare fundamentals, they enable us to understand the hows and whys; they tell us where to look and where not to look for evidence. They also help us to identify conditions that are necessary and/or sufficient for past choices and outcomes to influence the present". These words also relate to his objective in this book. The author gets a closer look to concepts that have somehow obeyed to vague descriptions in the past, presenting a logical approach that seeks to materialize in a set of sharp definitions and mathematical representations of what we usually call "diversity".

The book basically elaborates on a set of different notions of diversity, its measurement mechanisms, and its effects on systems with limited interactions and synergies as well as on systems characterized by sets of heterogeneous, interdependent and networked adapting agents. The book is divided into nine chapters. The first chapter talks about preliminary ideas like variation within and across types, variation in configurations among such types, the meaning of complexity and its distinctiveness from ordered and random systems, the relationship between diversity and complexity, how diversity can produce complexity, and the difficulties of assessing the effects of diversity on complex systems. Chapter 2 gives an introduction to different approaches to measure diversity -a good complement to this chapter is the article of Stirling (2007). Chapter 3 and 4 refer to the causes of variation within and across different types, as well as to the constraints that different types of diversity might face, respectively. Chapter 5 explains how diversity enhances robustness in complex systems, and how interactions (that produce either negative or positive feedbacks) affect stability in different ways. Chapter 6 and 7 speak about the benefits of diversity (averaging and diminishing returns to scale), while Chapter 8 emphasizes the role of diversity on the performance of complex systems (through, e.g. responsiveness, synergies and the construction of collective knowledge, just to name a few mechanisms). Chapter 9 draws on main conclusions and final ideas for future exploration.

The author warns the reader regarding many apparent misconceptions, and offers examples to clarify them. One of the conceptual clarifications goes to explain that robustness and stability are different concepts (Chapter 5). Also, for example, the "more diversity implies more robustness" notion is weakened with at least two examples (pages 165 and 236). For instance, variation of beliefs in a power game might introduce incentives to one of the players to attack his opponent, thus generating a non-robust system. Also, increase in connectedness diversity of a set of nodes with a preferential attachment rule might prove to be more robust against random attacks with respect to a random connected network, but might lose robustness if such attacks are strategic: in the case of networks that behave according to a power law, the set of most connected nodes is more susceptible to be spotted. Among others, the author also offers interesting insights on how diversity increases accuracy in prediction (pages 224-227), the redundancy-diversity tradeoff (pages 227-240), and provides also examples of how coordination mechanisms can either reduce or increase diversity depending whether the focus is within a community or across communities (pages 109 and 139).

The book falls short in providing a more extensive account on the downside of diversity, as it does with the benefits of it. Such problematic aspects are just briefly mentioned in several parts of the book. For instance, in Chapter 8, the book explains how diversity, through crosscutting cleavages, can enhance robustness. Although it is said that in total absence of crosscutting cleavages agents can easily segregate at no cost, it is not emphasized that crosscutting cleavages are just a way of blocking negative diversity effects that, through demographic faultlines, might trigger group fragmentation (Lau and Murnighan 2005). In the light of organizational behavior literature, diverse teams may be less cohesive than homogeneous teams, and the lack of cohesiveness may hamper the performance benefits of diversity (Flache and Mäs 2008).

In summary, this book may constitute an important resource for the social simulation community for several reasons. First, it provides a common vocabulary and an inspiring vision on how to approach effects of diversity at different levels. Second, as mentioned in the book, diversity can create complexity and has minimal effect on noncomplex systems. Thus, effects of diversity are important because they do not operate in a vacuum, but through many interdependent, adaptive entities. However, the fact that they operate in such systems makes it difficult to statistically test its effects (page 46). As Levinthal (1997) proceeded when he studied joint and separate effects of adaptation and selection in industry evolution, computer models can build up on the arguments developed in the book and more precisely point at aimed impacts that can be isolated, so that such models can consequently inform empirical research. Last, under the light of social system design, understanding the benefits and consequences of diversity at multiple levels helps to conceive systems able to cope with the required variety we face in the contemporary world (Conant and Ashby 1970). All in all, I recommend this book.

* References

CONANT RC, Ashby WR (1970) Every regulator of a system must be a model of that system. International Journal of Systems Science 1(2): 89-97

FLACHE A, Mäs M (2008) How to get the timing right: a computational model of the effects of the timing of contacts on team cohesion in demographically diverse teams. Computational and Mathematical Organization Theory 14(1): 23-51

LAU DC, Murnighan JK (2005) Interactions within groups and subgroups: the effects of demographic faultlines. Academy of Management Journal 48(4): 645-659

LEVINTHAL D (1997) Adaptation on rugged landscapes. Management Science 43(7): 934-950

PAGE SE (2006) Essay: path dependence. Quarterly Journal of Political Science 1(1): 87-115

STIRLING A (2007) A general framework for analysing diversity in science, technology and society. Journal of the Royal Society Interface 4(15): 707-719


ButtonReturn to Contents of this issue