ABSTRACT
There is a long history of attempts to throw light on the mechanisms of normal cognition based on its breakdown in brain-damaged patients. In the 19th century, a number of German neurologists attempted to create models that would explain the deficits shown by their patients, linking behaviour to the neuroanatomy of the brain. This approach had much in common with the information-processing “box and arrow” models that were beginning to influence cognitive psychology in the 1960s, particularly in the UK, where patient-based data relevant to cognitive models were increasingly presented at experimental psychology meetings. Such individual cases can be highly intriguing, as subsequently demonstrated by Oliver Sacks in the various books describing his patients. My own initial view, however, was that they were likely to be too complex to provide useful theoretical guidance. Hence, when asked by Elizabeth Warrington, who I knew from student days at University College, if I would like to work with amnesic patients I replied that it seemed unlikely that they would be obliging enough to have their lesions in theoretically interesting locations and declined. I did, however, agree to go along and demonstrate a technique I had developed that Elizabeth thought might prove theoretically interesting when applied to patients.
