A contribution to a theory of modality meanings and forms, on the basis of an analysis of complements. The questions addressed are: Can one find a small, finite set of meanings which systematically underlies the enormous variety of meanings found in complements? And can one make any predictions from this set about the variety of forms they take?