Programming vs. Understanding

In the US we can see two broad approaches to strategic thought after World War II.

First, the notion of strategy as a form of programming. This form of strategic thought deliberately blurred the lines between human and machine, simulation and reality, and operator and artifact. It aimed to understand the mechanical aspects of decisionmaking and how decisionmaking could be mechanized. There are many sources of this, but Herbert Simon’s studies of economic reasoning, chess-playing, and bureaucratic organization are some of the best and most fruitful examples of it. Very few, however, have comprehensively assessed this as an intellectual program.

The second, is strategy as a way of understanding. Andrew Marshall’s corpus is less a direct model of strategy as much as a epistemic program for how to strategize. Ditto to John Boyd’s OODA, which is not a literal model of decisionmaking as much as a design for epistemic growth and change over time. Although, one may say that this is not a purely US distinction to begin with. Antoine-Henri Jomini’s notion of strategy as the art of “making war on the map” vs. the Clausewitzian notion of “critical analysis” in tracing effects back to their causes parallels this. And Marshall was also inspired by Simon’s notions of bounded rationality as well.

When it comes to computational social science, I see both as being useful.

1. Strategy as programming is useful because, as reductive as it is, social science can only simulate strategy from a mechanistic point of view. In other words, we should be honest that machines can only crudely approximate intangibles, so the art of simulation is the act of mechanizing decision in a way as to illustrate some limited subset of the strategic whole.
2. Strategy as understanding is useful because we can use simulation as a creative tool for showing how different methods of problem framing and analysis may produce differing results by playing a given strategic interaction either forward or backwards. We are using the computer as a creative tool, an mechanistic aid to the human process of critical analysis.