Yes, it’s a dry title. It captures the main point of my argument, however, and so it will stand.
I am a historian, but I also have a fair background in computer technology and the physical sciences. This is, it seems, a rather unusual mixture. That’s a messed-up situation, because my experience is that awareness of both “fields” (as it were) enables you to approach things in a much more powerful way. It also allows you to see some of the glaring deficiencies in approach that are so horribly common. And no, despite what the hard-core physics nerds tell you, the cause of science would not be served by universalising their approach. We will not gain an understanding of all the important phenomena around us through reductionistic mathematical models with single variables that become easy in the right degenerate cases.
On the other hand, idle speculation disconnected from both formal models (or plausibly simulable ones, at the very least) and rigorous examination of repeatable empirical evidence leads to pretty certain doom. Freud’s psychology fell into this trap to a remarkable extent. It is, bluntly, astounding that anyone still believes Freudian theory can be used to explain fragments of human behaviour. They do, though; I’ve talked to modern-day university students who are taught Freudian psychology in the context of studying culture.
If the choice were only between Skinner and Freud, we’d be in a pretty bad state. Fortunately, that’s not where we find ourselves. Around the middle of the twentieth century, people starting seriously studying really complicated systems for real. There were a number of avenues, whose contemporary offspring include information theory, chaos theory and systems theory. One interesting episode in the development story of those fields is the history of the discipline of cybernetics, popularised by Norbert Wiener as the generalised science of control and communication. Cybernetics held that a sharply interdisciplinary approach was the most fruitful, and that all fields dealing with large mechanisms would have something to contribute and something to gain from a unified study of information flows.
Unlike most of the other mathematically-inspired approaches to understanding complex systems, cybernetics was conceived of as interdisciplinary and with significant relevance to the non-mathematical social sciences.1 This means two things. Firstly, that cybernetics can be seen as providing a very powerful toolbox to students of social science, be it economics, sociology, political science or even history. Secondly, that many of the things already done in those fields may be useful in improving the general model provided by cybernetics. We don’t have to be stuck in a place where some people know the canon of great literature and others know the laws of thermodynamics and neither can understand the other.
Instead, we find a place where human society is a subject matter for information-theoretic research; a large-scale communications network we can, to varying extents, model, analyse and simulate. Social scientists find their work supported not only by basic statistical work but increasingly by formal mathematical models capable of backing up not only the quantitative results but also the logical structure of the qualitative conclusions. The more mathematically inclined in turn gain access to a vast collection of systems to be modeled, generalised from — and, most helpfully, improved.
Interestingly, but probably as a result of the heavily mathematicised nature of modern economics, they’ve long since jumped on the chaos theory bandwagon.[return]