ABSTRACT
This chapter explains how the scaling hypothesis follows from the presence of a diverging correlation length. It shows the relationship between coupling constants defined at different length scales is more complicated than assumed. K. G. Wilson elaborated and completed L. P. Kadanoff’s argument, showing how the relationship between coupling constants at different length scales could be explicitly computed, at least approximately. Wilson’s theory — the renormalisation group (RG) — is capable of estimating the critical exponents. The RG also provides a natural framework in which to understand universality. The chapter provides the argument in two parts, the first being the thermodynamic scaling laws, and the second being the scaling laws for the two-point correlation function. Kadanoff’s block spin argument successfully motivates the functional form of the scaling relations. The conceptual importance of the argument is that it suggests how fruitful it might be to get away from the conventional statistical mechanical approach of treating all the degrees of freedom at once.
