Contemporary communication theory and research increasingly reflects a cognitive orientation to human behavior. Two key questions underlying any such approach are: Are humans cognitively active? And, if so, can they accurately report on their cognitive activity? Langer and her associates answer the first question by asserting that human behavior is typically mindless, and Nisbett and Wilson respond to the second by claiming that verbal reports are generally untrustworthy. This essay evaluates the research adduced by Langer and Nisbett and Wilson in support of their claims and presents additional data. For a variety of reasons (e.g., faulty design, alternative explanations, atypical phenomena) their claims are inadequately supported. On the other hand, insufficient evidence is available to support contrary claims (i.e., that humans are typically cognitively active; verbal reports are usually accurate). Rather than attempt to defend either of these dialectical contraries, we ought to do two things. First, we must eschew overly simplistic dichotomies (i.e., either mindless or mindful, either accurate or inaccurate), recognizing that both cognitive activity and accuracy of verbal reports exist on a continuum. Second, and most importantly, we should focus theory and research on discovering the factors and situations which predispose humans to be more or less cognitively active, more or less accurate in verbal reports. In the conclusion, this essay sketches a direction for theory and research on these issues for communication. © 1986 Taylor & Francis Group, LLC.