Hot Cogs and Divine Proportions Advances in Cognitive-Behavioral Research and Therapy (Vol. 5). P. C. Kendall (Ed.). Orlando, FL: Academic Press, 1986. (285 pp.) This is the fifth volume of an annual series devoted to reviewing cognitivebehavioral topics that may be of interest to research workers and clinical therapists. It contains a diverse collection of seven contributions, whose value may be best conveyed and appreciated by a brief description of the content of each. The first article, by Robert Schwartz and Gregory Garamoni, suggests that psychological well-being may be a function of holding both positive and negative cognitions in the divine proportion (the so-called "golden section hypothesis" of Pythagoras) of roughly three to two, that is, three positive cognitions to every two negative ones. They argue that this cognitive asymmetry enables one to feel generally positive while at the same time being prepared to cope with any negative events that may arise, and they present evidence that seems to support this rather intriguing idea. They go on to develop a "states of mind model" which proposes that certain psychological problems are the result of deviations from this proportion. For example, hypomania and mania are the consequences of holding too many positive self-statements. Although the implications of this model are not drawn out for the practicing clinician, it suggests that the more astute therapist may be able to monitor the progress of therapy by paying careful attention to the evaluative content of the client's comments. In the following contribution, Timothy Smith and Kenneth Allred provide a useful and detailed critical review of the appropriate empirical evidence on the original version of Ellis's rational-emotive theory. This simple and attractive theory suggests that inappropriate emotional distress arises from holding irrational beliefs. Despite its appeal, there is at present little methodologically sound support for it. Most of the relevant research is cross-sectional in design, assessing the presence of emotional distress and irrational beliefs at one point in time, which means that it is not possible to infer the causal nature of any obtained relationship. In addition, the two most frequently used measures of irrational beliefs, the Irrational Beliefs Test and the Rational Behavior Inventory, do not appear to possess strong discriminant validity. This may be partly attributable to the affective content of some of the items on these tests, which overlaps with that of the psychological distress measures. Consequently, the observed relationship between emotional distress and irrational beliefs may be to some extent a function of overlap in content of general dissatisfaction. Although not reported here, there are now measures of irrational beliefs that should overcome this problem. The third article, by John Marzillier, is most probably of greatest general interest to the clinical practitioner. In it, he begins by outlining Beck's cognitive theory of the accompanying therapy for depression. He suggests that it may be more appropriate for the cognitive therapist to view the dysfunctional beliefs that are associated with the patient's depression as reflecting an alternative set of values rather than being illogical in themselves. For example, the dysfunctional statement, "If I make a mistake, it means I am a complete failure as a person," may not be illogical or untrue as such but may embody a very high value being placed on perfection as a way of guiding one's life. He backs up this point of view by describing Rowe's personal construct therapy. She makes a distinction between rational beliefs and metaphysical beliefs, and argues that it is metaphysical beliefs (such as, "No matter how good or nice I appear, I am really bad") that are central to depression. Unlike rational beliefs, these cannot be changed by reasoned argument or empirical validation. What the therapist can do, however, is to discuss the advantages and disadvantages of holding such a philosophy and to suggest alternative and less harmful sets of values that may be held instead. …