It is puzzling that some anatomical systems in the brains of monkeys and men appear to function quite differently although there is no established basis for this either in their fine-grained anatomical organizations or in the inherent behavioural capacities of the two species. It is suggested that in some instances the discrepancies may arise because inappropriate tests have been used with the animals, and examples are given of positive evidence for cross-modal perception and a possible experimental basis for hemispheric specialization in the monkey. In other instances, they may derive from inadequate or inappropriate methods of testing human subjects, and attention is focused on two major examples: memory disorders associated with medial temporal lobe lesions and blindness associated with occipital lesions in man. In both examples it has been generally concluded that the human deficits are far more severe and even qualitatively different from those studied in the monkey. Evidence suggests that the discrepancies may be resolved if human subjects are tested not by means of 'commentary' questions (e.g 'do you see this? or 'do you recognize this?') but by methods that depend upon forced-choice or identification procedures that are more closely related to those used with animal subjects. It is argued that the study of dissociations between a capacity and its acknowledgement by a human subject may suggest a type of brain organization that is consistent both with the engineering approach that Craik would have fostered and also with one that places older and newer brain structures in the single evolutionary framework.