Human variability in temperament allows a unique natural experiment where reactivity, self-regulation, and experience combine in complex ways to produce an individual personality. Personality disorders may result from changes in the way past memories filter new information in situations of emotional involvement with others. According to this view, disorders are specific to their initiating circumstances rather than a general difficulty that might extend to classes of information processing remote from triggers for the disorder. A different view suggests a more general deficit in attentional control mechanisms that might extend to a wide range of situations far from those related to the core abnormality. This paper outlines methods for examining these views and presents data from the study of borderline personality disorder, arguing in favor of high negative emotionality being combined with a deficit in an executive attentional control network. Because this attentional network has already been well described in terms of anatomy, the cognitive operations involved, development, chemical modulators, and effects of lesions and candidate genes, these findings may have implications for understanding the disorder and its treatment. We consider these implications in terms of a general approach to the study of personality development and its disorders.