Is rationality ‘rationalizing’?


In the context of the INSOSCI Witten group, Lea Diederichsen is working on various case studies, one of which is herding on financial markets. She explores an interesting approach that transfers the ‘social intuitionist model’ in moral psychology to this field of research. In my view, it ideally converges with the framework of social psychology as developed by G.H. Mead and which I promote as a potential substitute for dual systems models in behavioural economics (see my recent blog posts, such as of November 24, 2017).

The social intuitionist approach was seminally developed by Haidt and starts out from a dual systems view, too. Yet, it goes beyond the standard approaches in making the role of social interaction central. Whereas standard dual systems views approach the reflective system as being centred on internal processing of information and decision-making, Haidt suggests that the main task of the reflective system is to give reasons for actions to others. That means, the reflective system is not just ‘rational’, but ‘rationalizing’. Interestingly, Haidt also refers to Gazzaniga’s conception of the ‘interpreter’ that might even be localized in the left hemisphere, which would converge with G.H. Mead’s ideas that the self is divided into an ‘I’ and a ‘me’, with the ‘me’ continuously making sense of ‘I’s actions relying on interpretive means that are fundamentally social in nature. Thus, the primary function of reflection is producing reasons that can be communicated to others, and this includes internalization, which means, that we can also present reasons to ourselves. Accordingly, in moral judgment, intuition comes first, and reason after.

Haidt concedes that intuition can be dysfunctional, but also emphasizes the literature that argued (writing in 2001) that often or even mostly intuition leads to good results, as shown by Gigerenzer in his work on ‘rules of thumb’ or Damasio in his work on emotions and rationality. These results are seen as well-established also today, I think. In other words, if reasons produced post hoc are socially recognized as good reasons, they just show the power of intuition. Haidt also emphasizes that there are many feedback loops from reasoning to intuition, especially when considering social interaction. That means, among other effects, reasons communicated post hoc by one individual can influence the intuitions of others, thereby improving their judgment, if reasons were good ones. I think that this is an extremely interesting idea and is another case in point for the questions of boundaries raised in my previous blog: Intuitions are shaped by social interaction and communication, and that means that rational discourse may also shape the way how the ‘fast’ system operates.

Lea Diederichsen thinks that this model could bridge the two approaches to herding that are available: The rational economic models and the psychological models. The problem is that the economic models are far too general and abstract to result into convincing and empirically testable causal explanations of herding. On the other hand, the psychological approaches are fragmented and without theory. The social-intuitionist model may be a powerful midway solution. I interpret it as a mechanistic model, and indeed, this is what Haidt did and Lea is doing too, in the best tradition of mechanistic approaches: Deciphering a complex causal dynamics and starting out from graphical presentations of the interactions.

In simplest words, this would mean that financial actors mainly start out from their intuition when joining or staying out of the herd, and that they would continuously conduct conversations about this, from ‘silent speech’ to factually chatting with others. Herding is not just a one-spot decision, as they can always change their mind. In this approach, the distinction between ‘private’ and ‘public information’ loses much bite, which is so essential for standard economic models. What is ‘rational’, emerges in the process, and cannot be treated as a precondition for explaining the observations.

I think that Lea’s work is an excellent example for the fecundity of the mechanistic framework in philosophy of science for dealing with real-world phenomena in the financial sector: Adopting the methodological standards of mechanistic explanations enforces a conceptual discipline on what might otherwise be just a collection of effects and possible causes. What we get are models of mechanisms which can claim to be falsifiable, hence turn out to be genuine hypotheses. At the same time, these models are much richer than standard economic models, thus becoming genuinely realistic.

Haidt, Jonathan (2001) „The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment.“ Psychological Review 108, 4: 814–34.


Funded by: