Are Climate Models Well Confirmed? And Why Would We Want to Know?

Gregor Betz (Karlsruher Institute for Technology, Institute of Philosophy)  video conference
Abstract: At first glance, understanding the extent to which different climate models are empirically confirmed seems to be essential both for practical as well as scientific purposes. At second glance, however, things are much less clear. I will explore two points. First is the idea that what is at best confirmed are not climate models themselves but hypotheses that refer to climate models which specify how they resemble a certain target system (i.e., structural similarity hypotheses and adequacy for purpose hypotheses). There are potentially infinitely many such hypotheses which can be articulated in regard to one and the same model. It’s not trivial to understand which of these hypotheses gets confirmed when, for example, an empirical implication of a model turns out to be true. Taking for granted that we can specify the extent to which different hypotheses are confirmed, the second point examines whether this information is really useful. My doubts stem from analogies to historical scientific contro – versies, such as the so-called Great Devonian Controversy.