Ramanan — On Models

Ramanan has a short piece on models and modeling that hits on some key issues. Everyone uses conceptual models. Models may be more or less precise. Formal models, however simplistic, reveal their assumptions. Informal models are generally much less clear about assumptions. So whatever their limitations, formal models are useful in attaining greater conceptual clarity.

I would agree with that with qualifications. All thinking takes place within language in which semantic meaning is context-dependent. Ordinary language is based on a worldview that one learns in acquiring the language. Most assumptions are hidden in that they comprise the infrastructure of the model. 

Not only is this never articulated in any degree of completeness, there is no way of knowing from within the model whether the articulation is complete. Moreover, criteria that the model uses are established by the construction of the model. There are no independent or absolute criteria that are model-independent in a worldview defines "reality" instead of reality defining worldviews as is generally assumed illogically.

Social sciences other than economics recognize the difficulty if not impossibility of constructing a comprehensive model of a society or even limited models that explain complex social phenomena. Only economics assumes that this is possible and that there is a correct model, that is, a model verifiable as true based on independent criteria. But that is a logical impossibility without either a vicious circle, infinite regress, or some postulated foundation.

Complicating this is the observable fact that different explanations (models) can account for the same data by constructing information differently. Another complicating factor is that society or economy size models are difficult to impossible to test rigorously using laboratory method. Establishing causality is therefore often impossible to do without assuming the model.

Social scientists recognize such difficulties. Many economists, not so much, making claims that exceed what is warranted.

This is philosophy of social science and philosophy of economics. It is a branch of logic, i.e., examination of how an information system functions. Keynes was well aware of this, for instance, and he was critical of misuse of econometric modeling.

Formal models are be useful in social science, including economics, if the limitations of the model are recognized. However, informal models are imprecise in another way. Moreover, humans are trapped within the limitations of language and therefor of conceptualization. Therefore, skepticism of knowledge claims is in order generally.

The Case For Concerted Action

On Models

Ramanan