Bigger than Chaos: Understanding Complexity through Probability

by Michael Strevens

Many complex systems--from immensely complicated ecosystems to minute assemblages of molecules--surprise us with their simple behavior. Consider, for instance, the snowflake, in which a great number of water molecules arrange themselves in patterns with six-way symmetry. How is it that molecules moving seemingly at random become organized according to the simple, six-fold rule? How do the comings, goings, meetings, and eatings of individual animals add up to the simple dynamics of ecosystem populations? More generally, how does complex and seemingly capricious microbehavior generate stable, predictable macrobehavior?

In this book, Michael Strevens aims to explain how simplicity can coexist with, indeed be caused by, the tangled interconnections between a complex system's many parts. At the center of Strevens's explanation is the notion of probability and, more particularly, probabilistic independence. By examining the foundations of statistical reasoning about complex systems such as gases, ecosystems, and certain social systems, Strevens provides an understanding of how simplicity emerges from complexity. Along the way, he draws lessons concerning the low-level explanation of high-level phenomena and the basis for introducing probabilistic concepts into physical theory.

Genres: Philosophy, Science

430 pages, Paperback

First published June 1, 2003

Reviews

[Joshua Stein · Rating 4 out of 5]

Strevens' Bigger than Chaos is different than much of the existing work on the role of probability in scientific methodology, partly because it focuses less on physics and more on fields like ecology and economics. This allows for Strevens to show that the methods under discussion in some areas of philosophy of science and mathematics can be generalized out to discuss domains of science that are not in vogue. If for no other reason, even a decade after its publication, this book is rather different than much of the contemporary philosophy of science.

This is a particularly technical book, partly because Strevens is working with a new vocabulary and spends large portions of the first two sections defining his terms, and large parts of the second and third sections giving a more formal, mathematical framework for the discussion. These discussions are important for the later sections, and so it is important not to neglect them, despite being very challenging and dense.

Sections four (in particular) and five are where the major payoff for the work done in the early sections, where Strevens fleshes out his account and really gets into the substantive philosophical baggage of the role of probability in scientific explanation; the text does a great job at discussing probabilistic characterizations of causal independence, and the role of causal independence of micro-level conditions bearing on the macro-level phenomena that he's interested in offering an explanation of. Strevens acknowledges throughout these sections that, for any particular domain, you have to demonstrate certain features in order for probabilistic analysis to function the way that he wants it to, but he does a good job of offering a way to perform such a demonstration.

Overall, the book is well organized and thorough. Sometimes that makes a philosophical work particularly challenging, and this is such a case. Strevens is not really able to open up his discussion and play with the ideas until very late in the game, but the payoff is well worth it.

A quick disclaimer: Michael Strevens is my current thesis advisor; tied into this is my respect for him as a person and philosopher. I don't think that has colored this review in any important way, but it is important to have noted.