The processes that damp solar pulsations also influence pulsation frequencies. This creates deviations from frequencies computed assuming adiabatic motion without turbulent stress and is one possible reason why the observed frequencies of the five-minute oscillations do not agree with theoretical eigenfrequencies. Another possible cause is that the commonly used local mixing-length theory fails to reproduce the mean structure of the Sun in the turbulent surface boundary layer. The computational machinery developed for linear stability analysis in an earlier paper is used to gauge the degree of error incurred in employing these two standard approximations. The results indicate that the assumption of adiabatic stress-free motion is less severe than errors likely to occur in modelling the solar superadiabatic boundary layer by conventional formulations of mixing-length theory.