This issue could almost lay claim to being a special issue about all those things that aren't supposed to matter in learning but do. Although emotions and interpersonal skills have been recognized as an essential component of competence for any health professional, they are rarely integrated into an overall perspective on the what and how of learning. At admissions, we speak of assessing ''non-cognitive'' (un-thinking?) skills. We hold separate courses in communication skills, frequently taught by professionals like social workers and clinical psychologists, whose professional qualifications are unassailable, but who reinforce the notion that this is a separate domain, detached from the mainstream. We speak of the ''hidden curriculum'', which all too often reinforces a paternalistic and aloof role for the professional.Our research endeavours reinforce this dissociation. Cronbach (1957) first noted two solitudes within psychology, where one group-the correlationists-studying clinical, developmental, intelligence, focused their research on individual differences and the other, the experimentalists, eschewed any hint of individual attributes, having only recently graduated from dogs, mice and pigeons to undergraduate psychology students. In health sciences education, there is an analogous dichotomy. People like me do psychometrics on tests, and while we may look for indices of individual differences, they lie exclusively in the ''cognitive'' domain. The new movement in ''Science of Learning, Science of Instruction,'' championed by cognitive and educational psychologists like Bjork, Roediger, Dunlosky, who have moved from cognitive psychology into mainstream education and health professions education, speak of human learning as if we're all pretty well wired the same, with more or less the same working memory and long-term memory, and certainly nothing of relevance outside the prefrontal cortex. It's as if we think about learning of emotions, or learning or emotions, but never learning and emotions.