Some assumptions widely adopted in clump theory are reviewed. It is shown that: (1) The incoherency assumption of (fcfcl)=0 is incorrect. (2) Clumps do not supply any drag term in f's equation, it only modifies the diffusion coefficient. (3) The main effect of clumps is in formation of steady turbulence. For practical purposes, the more reasonable simplification is starting from given turbulence (estimated from experiment) instead of from linear modes. (4) For given turbulence, the upper and lower bounds of clumps effect can be estimated from QL and RQL respectively. Though the clump part of (ff) is dominant its macroscopic effect is small.