The work that has formed the backbone of our research over the last ten years has involved the development of a novel theory of the organization of cognition in the medial temporal lobe and ventral visual stream. The main idea behind this “representational-hierarchical” view is that rather than confining ourselves to the prevailing paradigm of assuming modularity of cognitive processes within anatomically-defined cognitive boxes in the brain, we would do better to consider the representations maintained in differing brain circuits, which can subserve a variety of cognitive functions. These organizational principles have been instantiated in computational models, which have in turn been used to drive empirical work. Initial experiments tested the implications of this framework for memory in the healthy brain; however, over the past five years we have been applying it to the understanding of cognitive function in the dysfunctional brain, in conditions such as amnesia and dementia. Recent work, for example, has explored the prediction of this model that memory impairment in cases of medial temporal lobe amnesia and Alzheimer’s Disease may be due in large part to abnormal susceptibility to perceptual interference (McTighe et al., 2010, Science; Romberg et al., 2012, Brain, in press). We believe that novel thinking can lead to novel therapies and indeed, in these experiments treatments suggested by the predictions of the model were found to ameliorate memory impairments resulting from brain damage and Alzheimer’s pathology.