Marmur has claimed that large values of activity coefficients for nonelectrolytes, particularly in the context of hydrophobic interactions between solutes in aqueous solution at ambient temperature and pressure, cannot be accounted for by thermodynamics, and has suggested that association (self-assembly) of solute molecules in solution solves this dilemma. We show that the analysis of Marmur is incorrect, specifically because the equilibrium in solution between monomeric solute molecules and associated solute molecules is entirely ignored. We show further that activity coefficients such as that for nitromethane solute in hexane solvent, 39.7, and that for solute hexane in solvent water, 4.48 x 10(5), can be calculated as 31.9 and 4.71 x 10(5), respectively, by methods based on well-known molecule-molecule interactions. No assumption of self-assembly is required.