Dear Colleagues,
I want to build on what I have voiced yesterday when speaking about the need for a next enlightenment. As I have indicated, I think that we need to help people educate themselves about how to deal with large scale uncertainties.
There are several issues I would like to highlight:
– I have consciously phrased above “help people educate themselves”. This refers to Humboldt’s idea that it should not be academics who educate scholars, but human beings who educate themselves. It also refers to Humboldt’s idea that education is not identical to training people in specific professional skills, but means to grow as individual and become a person who is able to use her reason without external assistance. Clearly, this refers to Kant’s concept of enlightenment.
– I think that one of the biggest challenges for policy- and decision-makers is dealing with large scale uncertainties. Just developing some fancy tools that build on today’s risk tools will not do the job as today’s risk tools are severely limited. Many of them build on the frequentist notion of probability. They do a reasonable job when a sufficient amount of data is available, and when one can expect that the system that generated this data will not undergo substantial changes.
– For many global systems, we expect them to undergo substantial changes in the future. These changes will “devalue” the data from the past and subsequently devalue risk tools that need past data for projecting future risks. Instead, we need concepts and tools that help us build expertise about global systems, their future dynamics, and the uncertainties involved.
– Quick fixes will not do the job. I expect that we will need a further “round” of the enlightenment, which I like to denote as enlightenment 2.0, and which I expect to last for another 250 years. If, in 2250, we will have accomplished a decent job in dealing with global systems, more challenges will await us. Enlightenment 2.0 is only the next step. But as the Chinese say: The longest journey begins with the first step. I suggest to make the second one.
I very much agree with Armin that we need a fundamental change in the way we think about uncertainties and risks, and that this will have to come about collectively, with “educators” (=people educating themselves) interacting with other people also educating themselves (but without that lofty title).
From the very long term time perspective that is mine as an archaeologist, I would argue that recent centuries have seen an emphasis on ‘learning from the past’ rather than ‘learning for the future’. An emphasis on ex-post ‘explanation’ of observed phenomena, rather than ‘ex ante’ observation of the emergence of new phenomena. Our empirical science has been part of that development, as it is (a) reductionist and (b) linking the present to the past in formulating explanations, rather than also linking the past and the present to the future.
Some of the root cause is in our current education system, which predominantly socializes young children by telling them ‘truths’ (narratives that exclude other points of view), rather than stimulating them to think in terms of alternatives. I would therefore argue that the “Enlightenment 2.0” that Armin refers to has to include such thinking in terms of alternatives, and the unintended consequences that each choice may trigger in the (more or less distant) future. GSS is in my opinion as much about such thinking about alternatives as it is about looking at phenomena and challenges in a (much) wider context.
Thinking in alternatives is indeed an important issue. But then we should go a step further and attach probabilities to these alternatives. Why? Because we do it anyway, but most often unconsciously. If we tell “the truth”, we basically attach a probability of unity to what we announce as truth. But if we do it anyway, we should do it consciously, and openly.
Together with some colleagues, I have developed an approach that we call Bayesian Risk Management.
The central building block for this approach is the subjectivist Bayesian notion of probability. For subjectivist Bayesians, a probability is a subjective measure of uncertainty. The more daring Bayesians even define a probability as a willingness to bet. Interestingly, this is the notion of probability used by Immanuel Kant.
I really mean it when I speak about an enlightenment 2.0. It is a straightforward endeavour that can and should build on enlightenment 1.0.
Bayesian Risk Management comes with concepts and tools that help deal with competing scientific hypothesis in a specific field in a way that is much more sophisticated as a “consensus” approach. It can, for example, deal with the rather frequent situation when minority scientific positions develop into the new mainstream opinion. Alfred Wegener and his plate tectonics may serve as a paradigmatic example.
There is, however, a substantial complication with all this. There is a reason why people most often prefer to cling to “the truth” — they are looking for safe grounds they can stand on. From personal experience, I can tell that I encountered quite many human fellows who become very uneasy when they no longer sense firm ground below their feet. Enlightenment 2.0 must therefore address an existential dimension and help people tackle a condition humaine in which typically, firm ground is out of reach. This, however, is a big challenge.