Social emergence: societies as complex systems

Social emergence: societies as complex systems
R. Keith Sawyer

This book is awesome because it has Sim City 4 on the cover... well more seriously this book is exactly what my list this semester is about. One of the things I'm primarily interested in with society is the macro-micro causation. A macro-level structure is created from the behavior of micro-level behavior, but the structure can constrain micro-level behavior. I tend to agree with the central aim of this book, methodological individualism misses a lot of the causal power of including system level causes.

Emergence is a very tricky subject. Definitions and understandings of it vary wildly. It is used a number of different ways. Sometimes it's difficult to distinguish between self-organization and emergence and many people accept them as the same thing. But it's not just self-organization, a recent paper by Alex Ryan on Mobius strips demonstrates that. A bunch of triangles topologically have two sides, but if combined into a Mobius strip the strip will have only one side (*poof* emergence!). He also suggests that emergence relates to scope and resolution instead of scale, per se', which I totally agree with*.

Before I read this book, I read a page on emergent properties on Cosma Shalizi's notebooks. The idea Shalizi puts forward is that a theory of emergence should be useful if anything at all. Many believe that emergent properties intrinsicly cannot be predicted from knowledge of the individual agents (a property Sawyer referred to as multiple realizability, many different micro-state can produce the same emergent property), but this almost completely eradicates the utility of a theory of emergent properties. If however, emergent properties can be modeled and predicted, then it is possible to make a model that makes predictions of behavior not yet observed and we can go out and search for them (*poof* science!).

This book provides a nice history of emergentism and the thinkers involved in cooking it up. It discusses the major disagreements, such as the one I hinted at in the previous paragraph. Supervenience was a key term I was introduced to. It means that if two higher level properties A and B have the exact same properties then the lower-level properties that created them are the same as well. In addition, if the lower level properties that create B ever occur again, then B is sure to emerge again. That term is the most easy to agree with as a materialist who rejects vitalism**.

The book moves from there into a discussion of emergentism and psychology, but I didn't spend much time in that chapter, maybe some other time. I focused on the chapters about emergentism and sociology, Durkheim's emergentism, and modeling using agent-based models. It seemed the point that needed the most justification was the feedback from emergent structure back on individual behavior (downward causation), and that is what much of the section on emergentism in social theory focused on. To me emergentism is useful since it suggests that individuals create and perpetuate a social structure that has a causal influence on the behavior of individuals.

I don't feel like the book was argumentative at all. Sawyer seemed more intent on describing emergence and its importance in social science than tearing down other theories (kudos, let it stand on its own). He seems to present a straw-man of methodological individualism (MI), like most theorists do of theories they criticize. Most people I've met who subscribe to MI aren't unreceptive to emergentism or it's implications. I feel it's just a Kuhnian thing, that much of sociology, psychology, economics, etc. has been in practice for at least a hundred years doing what they've been doing. And it's not like MI is incorrect, it just misses the forest in favor of the trees.

I don't think I'd ever give this to family or friends. This book is made of serious theory and it does not have the feel of being popularized at all.
If all my friends and family were social scientists though, I might recommend it. I would recommend 1,000 Years of Nonlinear History by DeLanda first, then this second.

* - I was playing with cinnamon bears where I could bite the heads off and use the stickiness to attach them to each other. Then I thought, "could I make a larger bear from small bears?" Then I thought, "at what size and resolution of small bears could we say a large bear can be created?" That is, I could take 5 bears and make it vaguely resemble a larger bear, but if I used 2,000 I could make a very high-fidelity larger bear. What what scope and resolution can a new larger bear be made?

** - vitalism is a form of dualism, which suggests that life requires some 'vital' force in addition to physical forces in order to exist and is not supervenient on lower-level interactions.