A categorically postcategorical interpretation of the modality effect: a reply to Nilsson.