Clinicians have a convention whereby symptoms are subjective statements 'as told by' patients, whereas signs are outwardly observable facts. Yet both first-person reports and third-person observations are theory laden and can bias conclusions. Two aspects of the oft-mentioned unreliability of reports are the subject's interpretation of them and the experimenter's assumptions when translating introspective reports into scientifically useful characterizations. Meticulous training of introspectors can address their mischief, whereas investigators can become more attentive to their own theory-laden biases. In the case (...) of hallucinations, for example, ignoring some customary third-person constructs and focusing on the visual experience itself has led to fresh explanations of visual symptoms based on cortical physiology rather than conceptual categories. Constructs that historically have ignored the subject's state of mind are also problematic; an example is the so-called resting state during metabolic brain imaging, long believed to reflect a blank mental slate. Introspective reports, not accepted literally but properly interpreted and revised by investigators as necessary, are legitimate sources of data. (shrink)