Reductionism

I wrote the following in response to an online discussion about critiques of E.O. Wilson as reductionistic. The question that someone posed was "What's wrong with reductionism?"

Scientific reduction is a matter of modeling and simplification -- precisely the potent and limiting characteristics of law (nomos) in Statesman -- it is a set of purpose-built tools for specific phases of the intellectual process that is science (or any systematic search for knowledge). Reduction is probably indispensable (one is tempted to say irreducible!) for any such inquiry. To critique something (such as Wilson’s Sociobiology, or Consilience) as reductionistic is to level the charge that it has in one way or another misused the tool; it need not, and had better not be, to reject wholesale the tool as such.

By analogy, one really does need a hammer to build a house, but we would rightly reject as egregious “hammerism” any attempt to use the hammer as a paintbrush, to clean the windows with it, or generally to imagine that once the hammer’s work was done the house was complete.

One of the reasons reductionism, on the model of hammerism, is a fair criticism of some attempts to extend scientific findings into socio-cultural or moral principles is that the reduction model used for the (generally analytic) objects of scientific study ignores the emergence of properties in complex systems that are neither predictable from nor reducible to the properties of the components of those systems. Life emerges from combinations of water and minerals, obviously, and studying those components is vital, but an organism is more (because of its self-replicating organization of them, for example) than the sum of its component elements. The concept of irreducibly emergent properties is not unrelated to Marx’s employment of Hegel’s notion of dialectic.

Reductionism remains a permanent trap for intellectual pursuits, precisely because we need the tool of reduction to understand things, and so always risk imbuing its more striking conclusions (such as Wilson’s) with more weight than they can bear, and applying them unmodified to a level of complexity within a system for which they are ill-suited.