Problems with Popper’s Rationality Principle

The bottom line of this argument (yet to be composed) is that the problems with Popper’s RP in the context of his Situational Analysis (SA) can be resolved quite easily, if:

1. We accept that plans and intentions count as causes of human action.

2. We think we can find general laws (or at least laws of tendency or propensity) in the human sciences. There is a tension on Popper’s work because when he first wrote The Poverty of Historicism he advocated the search for universal laws in sociology and economics but later on he decided that such laws could not be found. I thought he told me that in a letter but it must have been in conversation on one of the handful of occasions when we met. Also in a casual social conversation a man in Cambridge told me  that he worked with Popper on a project to find sociological laws but they gave  up because all the ones they formulated had exceptions.

This is the kind of thing that needs to be sorted out while the thinker is alive. I don’t recall that he gave up on sociological laws in print, though in the discussion of  history he wrote that the kind of laws that historians might invoke are quite trivial. He didn’t go on to say this but it would seem that they not the kind of laws that the generalizing sciences would try to test and refine. Like when two armies of similar size and equipment engage, the one with very much larger numbers will win. To which he added something like (“or usually”!).

Two Vienna papers.

At the 2002 conference there were two papers on Popper’s RP and both can be found in the published proceedings, Maurice Lagueux on ‘Popper and the Rationality Principle’ and Boudewijn de Bruin on ‘Popper’s Conception of the Rationality Principle in the Social Sciences’. It is not possible to cut and paste out of google books so I will have to do some typing!

The paper that treats the RP had a strange history, starting as a lecture at Harvard in 1963. Some parts were worked up into a paper that was published in French in 1967. This appeared in English in A Pocket Popper (1983). Two larger manuscripts (one larger than the other) circulated among friends and relations until one of them became public in 1994 in a collection titled The Myth of the Framework edited by Mark Notturno.

Lageux notes that the basic formulation of the RP is “agents always act in a manner appropriate to the situation in which they find themselves”.

He wonders whether Popper is moving in the direction of von Mises with an apriori theory of rationality whereby human actions are rational by definition.  Popper insisted that his principle was not apriori, indeed it was quite capable of being wrong.

Replying to critical comments in footnotes to the published version of the lecture Popper noted that the referred to two versions of the principle and for good measure submitted a third referring to the situation “as the agent could (within the objective situation) have seen it”.

Lagueux argues in favour of Popper’s apparently cavalier attitude to the truth or falsehood of the RP, it is to be regarded as a part of the Situation and it  is necessary to animate the model of the Situation.

I am perplexed by the need for a universal law to animate the model, as Newton’s laws animate models of the solar system. Human actions are animated by people, not universal laws, for my money the laws impose limits but they don’t determine a sequence of events.

In a section headed ‘Why immunise the RP?’ Lagueux suggests that Popper is using rationality in a particular sense in the RP which prompts the thought, not for the first time, that rationality is one of the words along with induction, realism, liberal, conservative and rightwing that could usefully be given up as far as possible due to their multiple meanings. The Lagueux story improves when he brings up the idea of using the RP as a methodological principle, bearing in mind Jarvie’s account of the layers of rules in Popperian method, starting with the supreme rule that all the other rules should support the testing of theories and hold at bay the various strategies that are used to immunise theories from testing. So according to Lagueux the RP is allowed to be dispensible so that the full weight of testing falls on the other components of the model. He writes “Far from protecting scientific statements against falsification, this decision, according to Popper, allows us to falsify the other statements o f the model, the only ones that are meaningfully falsifiable”.

At this point I am screaming for an example of SA and the RP in live use, working on a real or substantive problem. Much discussion of methods in economics does not relate to concrete problems. A shining exception is the work by Wong on Samuelson’s theory of preferences which in effect shredded Samuelson’s position. The field reacted to this achievement by ignoring it, as described by Mirowski (with a dig at Popper on the way).  

There is another way to  look at the RP, as a  hueristic device, not itself a univeral animating force but a prompot to ask questions like – how did the agent perceive the situation, what was his or her aim and plan at the time, and is there any record of the decision procedure that they used before they acted?

Lagueux concludes that the RP, treated in a relaxed way, is good enough because it is near enough, most of the time and he cites a previous publication of his own where he argued that the explantory theories of classical, Marxist and Austrian economists are clearly based on such a principle.

What is Popperian SA supposed to explain?

The example that Popper used in his paper was Richard Gombrich (son of Ernst) crossing the road. What has the model of a single action got to do with the things that economists most want to explain, things which are the unintended outcomes of many actions,  like prices, booms and busts, inflation and unemployment?

To be continued…I think people don’t like long blog posts so I  will come back another day and do another post to continue the discussion.