Global warming is real and in large part due to human activities. Increasingly, climate models are used to make high-resolution forecasts about future changes in our immediate environment. Do such forecasts provide policy-relevant information? Given the systematic errors in current climate models, treating model outputs as a basis for decision-making can be misleading. We have to relinquish unreasonable demands for detailed long-term projections and instead think about policy making under conditions of uncertainty.
Statistical mechanics is a reductionist project: it describes the behaviour of macroscopic systems in terms of the laws governing the dynamics of the system’s microscopic constituents and probabilistic assumptions. The dynamics of a large class of systems is epsilon-ergodic, which explains why they behave thermodynamic-like. The theory’s probabilities ought to be interpreted as Humean objective chances, determinism notwithstanding, and a generalised version of Nagelian reduction is the project’s ‘background philosophy’.
Modelling and Representation
Models are of central importance in many scientific contexts. We study models and thereby discover features of the phenomena they stand for. What are models and how do they represent their target systems? Models are fictional objects of the same kind as, say, Tolkien's Middle Earth, and they represent their target systems by both denoting them and being equipped with translation-keys. The fiction view of modelling is an alternative to structuralist approaches, which are too austere to capture important aspects of scientific modelling.