Paradigm Change in Science and Economics

In this post, I discussed normal science, a term used by Thomas Kuhn in The Structure of Scientific Revolutions to describe the day to day work of scientists, focusing on the example of my brother’s work on transmission of pain in the body. In normal science, Kuhn explains, people expect the puzzles they choose to work on will have solutions that can be worked out using the paradigm, and if the first try doesn’t get the solution, scientists just keep plugging away, sharpening their instruments, their theories, their rules of engagement and trying to eliminate prejudices until they get a solution. And mostly, they do. That’s a good description of my brother’s work.

If not, generally they assume they failed, not that the answer doesn’t have a solution inside the paradigm’s limits. They put that problem to the side, and work on a related problem or maybe just move on to something different. Frequently the problem disappears as more and better techniques are created, measurements become better, theories evolve and prejudices are conquered. But if unsolved puzzles accumulate, there is growing pressure on the paradigm, and growing unease among the scientists working in the area. Kuhn gives examples:

The state of Ptolemaic astronomy was a scandal before Copernicus’ announcement. Galileo’s contributions to the study of motion depended closely upon difficulties discovered in Aristotle’s theory by scholastic critics. Newton’s new theory of light and color originated in the discovery that none of the existing pre-paradigm theories would account for the length of the spectrum, and the wave theory that replaced Newton’s was announced in the midst of growing concern about anomalies in the relation of diffraction and polarization effects to Newton’s theory. P. 67, fn omitted.

This is the crisis state. It is a necessary, but not sufficient, condition for a change in the paradigm. Kuhn analogizes the situation to political revolutions:

Political revolutions are inaugurated by a growing sense, often restricted to a segment of the political community, that existing institutions have ceased adequately to meet the problems posed by an environment that they have in part created. In much the same way, scientific revolutions are inaugurated by a growing sense, again often restricted to a narrow subdivision of the scientific community, that an existing paradigm has ceased to function adequately in the exploration of an aspect of nature to which that paradigm itself had previously led the way. P. 92

Another necessary condition for a paradigm shift is the existence of a new paradigm. Scientists cannot work without a paradigm, so until a new one obtains a concensus, they struggle on under the old one. New paradigms are suggested and tested, but Kuhn points out that there isn’t any way to prove that one is better than the other, because proofs only exist inside paradigms. The new paradigm has to satisfy the relevant scientific community that it will solve the old problems, and open the way to new problems. But this is a matter of persuasion, not of scientific proof, because the standards of proof are connected to a paradigm; they do not exist in some Platonic state above it all.

One final point. Kuhn says that in scientific revolutions, the new paradigm completely replaces the old one, and he gives plenty of examples.

There’s more to be said about the process of paradigm change, but this will suffice for this post. In the wake of Kuhn’s work, several papers were published trying to identify paradigm shifts on the order of the Copernican Revolution in the history of economics. One such is The “Structure of Revolutions” in Economic Thought, a 1971 article by Martin Bronfenbrenner. He thinks the history of economics is more like the Hegelian dialectic, thesis, antithesis and synthesis, than the catastrophic destruction of the previous paradigm.

Bronfenbrenner identifies three revolutions in economics as

1. The classical school, based on Adam Smith’s Wealth of Nations and David Hume’s Political Discourses.
2. The marginal utility revolution, dating to about 1870, led by John Stuart Mill and David Ricardo.
3. The Keynesian revolution, about 1936.

He adds the response of the Chicago school as a possible fourth, and time has proved his suggestion correct.

It should be obvious that none of these revolutions destroyed the older view. Instead, they sit side-by-side, if uneasily and with some overlap. Bronfenbrenner doesn’t see a problem with the survival of the natural law as a partial explanation of 20st Century capitalism, and assumes that the future will include some of those ideas as well. This is clear from his approval of Paul Samuelson’s textbook. I point out the problems with that view in several posts here and at Naked Capitalism, including this one.

Like others, Bronfenbrenner points out that Kuhn’s definition of the term “paradigm” is loose at best. For purposes of this post, it’s sufficient to regard it as the entire set of theories, understandings, prejudices, instruments, and interpretations of the measurements of instruments that guide the scientist in the course of normal science. It is, however, important to note that neither Bronfenbrenner nor any of the other writers I’ve seen so far try to explain the sense in which the Classical School, the Marginal Utility School, the Keynesians or the Chicago School, or, for that matter, any of the other schools, constitute a paradigm in a way similar to the way General Relativity acts as a paradigm for physicists and astronomers.

That offers two more or less neutral explanations of why economists aren’t all freaked out by the failure of their theories demonstrated by the Great Crash. First, they may well assume that events like the Great Crash are just anomalies that future work will solve. That would explain the response of Gary Becker, “You need a theory to beat a theory.” Link here. Becker couldn’t imagine an alternative theory, so he just continued to work inside his old one, as if his Chicago School were a paradigm.

Second, Bronfenbrenner is right that old economic theories never die. They cannot die. Instead, in his view, they will be assumed into the heaven of some synthesis, hopefully with the favorite views of each economist on top.

As a road map for the rest of this series, what does all this say about the claims of authority of economists?

7 replies
  1. Alan says:

    Are there many Marxists, Smithians, etc.? Not in economics departments or those hired by governments and financial institutions. Neoclassical economics dominates economic thinking. Yes, you’ll find a very different economics practiced in anthropology and elsewhere but alternate economic thinking in other disciplines is ignored by the discipline of Economics. 
    I’m not sure the discussion of Kuhn’s use of the paradigm in the context of the natural sciences advances the discussion of economics and surveillance. Unless one brings it up to merely dismiss the pretensions of economists that neoclassical economics is a science like the natural sciences. Economics can never be like a natural science because the topics of the human sciences, unlike the natural sciences, exist inside the social matrix that makes them possible and, in fact, the human sciences create and recreate the matrix.
    Foucault is a better guide to the human sciences as he looks as how man becomes a topic of investigation and the relation of the sciences of man–the creation of ‘man’ through measurement, recording, accounting, classification, surveillance, etc.–and structures of power. Yes, there are models, metaphors (the panopticon, confessional, etc.), that define sets of practices/discourses, what is and what is not, forms of self, and forms of governmentality. But none of this is usefully thought of in terms of paradigms in quite the sense Kuhn uses the notion in the context of the natural sciences. Economics wants to tell you that it has discovered the objective scientific rules, the essence, of economic behaviour (read all social behavior in neoliberalism). But there is no essence. It’s a fabrication of a matrix of knowledge and social relationships.

  2. Lawrence Hudetz says:

    Responding to Alan.

    Karl Popper, in his book, Open Universe, used the concept of falsifiability to falsify scientific determinism, while indicating that determinism itself cannot be falsified because it cannot be tested. My sense is this extends to economics or any of the social sciences.

    They cannot be tested because there is nothing to test, from the physical point of view.

    • Alan says:

      The Open Universe? I think you may mean The Open Society or maybe The Logic of Scientific Discovery?
      I don’t find the whole notion of falsification very convincing. What I find of value in Kuhn is that science involves communities that are the repository for sets of practices and skills-based knowledge. Falsification doesn’t really come into it.
      When people do an experiment there are always many dimensions so it is always possible to rationalize why something ‘didn’t work’ or you got a particular result. What makes a falsification a falsification rather than a failure of skill, equipment or failure to notice the significance of some variable? So, whether one is dealing with science or the social sciences there is always has the problem, when is a test really a test? When a relevant community decides it is.
      Normal science is a slow steady analogical extension of existing practices. Dramatic innovation often only happens because there are flows of personnel that bring different skills to bear on a problem e.g. as in my earlier example of molecular biology where a lot of the innovation happened when physics, communications theory, cybernetics, etc. were brought to bear on biological problems. In the latter case sources and patterns of funding played a key role in the changes.
      Kuhn’s work had sociological elements. His work was taken up to create a sociology of scientific knowledge. The sociologists (actually a group of sociologists, historians and philosophers, some of whom had been trained as scientists) jettisoned a lot of the ideas in Revolutions that didn’t work very well e.g. the vague notions of paradigms and revolutions. The sociologists emphasized Wittgensteinian thinking in Revolutions, the role of community, tacit knowledge, etc. Kuhn himself wasn’t very taken with what his work inspired. If you want to read reworking see T.S Kuhn and Social Science.

      • Lawrence Hudetz says:

        Dang! big long response just got trashed when I looked up a reference!

        Popper’s “Open Universe” relates to The Logic of Scientific Discovery.

        My computer seem to f up when I ht a wrong key. I’m going to write it out in Word first so I don’t lose everything!


        • bmaz says:

          For what it is worth, I looked in all our backend, and cannot find any semblance of it, sorry.

  3. person1597 says:

    Finite Element Analysis is a modern tool for simulation of complex physical systems. Used extensively to model thermodynamic processes, the technique instantiates local domains with individual characteristics and functions to be bonded together towards an approximation of an unknown solution. Sounds handy, eh?

    Adding adaptivity to each domain — as the solution unfolds — brings behavioral realism to the simulation model. Perturbations, fat tails, and exogenous nonsense are all welcomed variables in a giant partition of unity. The Big Tent!

    The idea that individual elements can be endowed with active properties that inform the macro construct entices the econometrician to embrace modern tools for modeling complex systems.

    Of course, this flies in the face of the leadership principle underlying market stability so we couldn’t have that. What would quant wrought?

  4. person1597 says:

    Fed’s ahead of everybody… From 1993:

    “Solving the Stochastic Growth Model With a Finite Element Method.”

    Surprisingly, no related research googles up. Hmmm…

Comments are closed.