DavidHolmes comments on Categorial preferences and utility functions