Reviewed by
Roberto Serra
Modena and Reggio Emilia University and European Centre for Living Technology (Venice)
This book provides a highly readable account of some of the major concepts and results which stem from recent studies of collective behaviour. The writing is lively and non technical, and most of the cases which are considered regard human systems (e.g., crowds), with a few exceptions dealing with biological systems (e.g., locusts). Therefore the book does not require a particular scientific or cultural background, and it should be of interest for the general reader, and in particular for managers.Given the popular scope of the work, the fact that several important studies in complex systems science are not taken into account should not be considered a drawback. Nor is it the fact that a historical account of the development of the field is lacking. However, an important concern is related to the fact that, while the author touches upon many important aspects, he does not deepen their analysis. Rather, he aims at deriving from these studies some practical advices, and this makes the book resemble in part a collection of "recipes" to deal with several aspects of everyday life or business. Indeed, the final chapter is a synthetic list of the various recipes which have been introduced in the previous chapters. Some look like good but fairly obvious wisdom (e.g., "look for cues that might help you to make a choice between options, and choose the option that has the excess of positive over negative cues", at the end of page 169). On the other hand, other advices are very specific and highly questionable, like e.g. rule 4 (page 168): "If you are with a crowd in a dangerous situation, use a mixed strategy for escape: follow the crowd 60 percent of the time, and spend the other 40 percent searching escape routes on your own". If you look at where this recipe comes from, it turns out that this is the outcome of some computer simulations - and while there are good reasons, presented in the text, to believe that a mixed strategy outperforms the extreme cases, it is highly likely that the relative proportion of the two (60-40) depends upon the features of the particular model which is used. And these aspects are not critically discussed. Similar remarks apply to other recipes which are presented.
In more general terms, I think that a major result of applying complexity science to social systems can be the appreciation of the high degree of unpredictability, and of the fallacy of "general rules" which promise success. A specific action can lead to very different outcomes depending upon several factors, and complexity provides theoretical concepts to take uncertainty into account - it provides proper lenses to see what is happening, without however overcoming the radical ("ontological") unpredictability of many situations. Instead, The perfect swarm adopts a style which is similar to that of several popular books for managers, and it prefers to focus on simple rules, instead of stimulating the reader to deal with the most intriguing aspects of complexity.
However, if one filters out the author's tendency to give simplistic advices, the book presents several interesting pointers to important concepts and research. The introduction to self-organization is a very effective way to present this concept to newcomers. The chapters on crowd dynamics (ch.4), on quorum consensus (ch.6) and on networks (ch.7) surely deserve reading. Moreover, the notes (which occupy more than 70 pages, versus the 170 of the text) provide interesting complementary information and anecdotes.
That said, it is worth stressing that the topic which is dealt with is certainly important and that such an attempt to bring it to the general public definitely deserves praise. Moreover, the publication of the book is timely and might stimulate careful readers to deepen their knowledge of the field.
Return to Contents
of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2010