Measurement, information and uncertainty (Q1091965)

The issues discussed in this paper are probably best classified under the heading of epistemology, in that they concern the differences in the limitations of what can be known (observed) in the physical, social and biological sciences. The smooth, continuous processes described by Gaussian distributions in the physical sciences are contrasted with the clustered discontinuous processes described by inverse power-law distributions in the social and biological sciences. The structural and dynamic properties of the underlying process implied by the observed power-law distributions are discussed using a maximum entropy formalism, stochastic differential equations and a renormalization group relation.