Abstract
The far infrared absorption in polar materials has often been attributed to the presence of dipole-dipole interaction between the molecules of the material. In particular the short time behaviour of the microscopic dipole autocorrelation function is most affected by this dipole-dipole interaction. In this paper, the assumptions involved in predicting the behaviour of the microscopic dipole autocorrelation functions from macroscopic spectral data are reviewed. It is shown that considerable care must be exercised in drawing any conclusions about the microscopic behaviour from the experimental data.