Friday, June 28, 2019

Racism is a framework, not a theory — Andrew Gelman


I am posting this for its relevance to philosophy of logic and philosophy of science rather than specifics. It draws a useful distinction between frameworks that generate theories and which are themselves not testable (hence falsifiable) and the theories a framework is used to generate.

This is obviously relevant to philosophy of science, but why philosophy of logic? In his later work, Ludwig Wittgenstein sought to show that the overarching frameworks of a culture as a way of life are deeply embedded in the structure and function of ordinary language.

Such frameworks are "world pictures (Weltbilden) that function as a world view. Although many of the propositions they generate appear to be descriptive, many are normative in operation. For example, the fundamentals of a world views are criteria for valuation and judgment, hence, they are stipulations that cannot be falsified from within that world view. For example, in doing science methodological naturalism is fundamental.

As a result many make the illogical jump from a methodological assumption to a metaphysical assertion (materialism) in neoclassical economics methodological individualism, microfoundations, rational maximization, market forces and equilibrium are key fundamentals that are assumed as criteria. This is a reason that neoclassical economists reject "heterodox economics" out of hand.

World views and ideologies are similar and need to be distinguished. A world view is a way of seeing the world that is embedded in ordinary language based on interpretation of context. They are subconscious and very difficult to articulate since they are the basis for using a language to communicate. This applies even to formalizations to the degree that assumptions are stipulated that link that symbols to life. (Pure math says nothing about the world. It is about how the rules for sign-use work in a particular syntactical system.)

Conversely, ideology is articulated at least in outline and people within a single world view can recognized differences of expression based on competing ideologies. Politics functions in this way, for instance. American liberals and conservatives degree but they function within the same framework.

So the application of the point that statistics professor Andrew Gelman is making goes far beyond the context of racism.

Statistical Modeling, Causal Inference, and Social Science
Racism is a framework, not a theory
Andrew Gelman | Professor of Statistics and Political Science and Director of the Applied Statistics Center, Columbia University

No comments: