1968, 11, 641-649
PRINCIPLES OF TEXTBOOK ANALYSIS A Review of J. R. Millenson's Principles of Behavioral Analysis.' T. VERHAVE2 AND J. GILMOUR SHERMAN ARIZONA STATE UNIVERSITY
Most textbooks in psychology, and particularly the eclectic survey, are probably best placed in the second category because they do not provide any new organization of a coherent body of information. Some introductory texts, however, take a systematic approach, and belong to the third category. Such is J. R. Millenson's Principles of Behavioral Analysis, which provides an updated version of a previously successful systematic treatment (Keller and Schoenfeld, 1950). The principal task for an introductory course in any science must be as clear and dedicated a presentation of fundamental concepts and theories as an author can muster. Even if a field is still relatively immature and undeveloped, a firm knowledge of fundamentals must be the basis for whatever is to be accomplished, either in subsequent courses, or ). (3) It can provide a new organization of ex- by later generations.inIt a is neither surprising more or less unexisting information (for example, Keller nor alarming that, plored domain, there should be different and Schoenfeld's Principles of Psycholview, vocabularies, and schools of ogy  or Galanter's Textbook of points ofNo thought. doubt, the relative maturity of a Elementary Psychology ). science is indicated by the extent to which (4) It can present conventional material in a new mode, as did the programmed such heterogeneity of treatment exists. However, by not presenting an encyclopedic text by Holland and Skinner (1961). account, whether superficial or detailed, but (5) Finally, there are the subsidiary books, such as collections of readings, labora- by concentrating instead on the power, meantory manuals, and brief primers, all of ing, and scope of a limited number of basic concepts which are primarily intended to be used reader a and ideas, one may hope to give the view of psychology as an interconin conjunction with a more substantial nected, even rational, subject matter. The partext. ticular set of concepts and topics selected by The criteria of success are different in each an author will always be open to argument case, and a book intended to fulfill one func- and controversy. Which ones are selected, tion should not be faulted for failing to meet however, is a matter of less importance than the criteria of another. the honest attempt to present them in a coherent and systematic framework, whether this be 'New York: Macmillan, 1967. Pp. xxii + 488, $7.95. Titchenerian, Tolmanian, behavioristic, men2Reprints may be obtained from Thom Verhave, talistic, phenomenological, or whatever. Nor are systematic books written in a manDepartment of Psychology, Queens College of the City ner which talks down to their audience. They University of New York, Flushing, New York 11367. 641
In evaluating any new book of a scholarly or technical nature, one has to consider the purpose for which it was written, for it is all too easy to attack it for not being something else. While other categories are possible, it seems that a book in psychology can be thought of as serving one or more of five functions: (1) It can provide new information (for exple, Hull's Hypnosis and Suggestibility  or Skinner's The Behavior of Organisms ). (2) It can attempt an extensive or definitive summary and digest of an available body of knowledge (for example, Cofer and Appley's Motivation: Theory and Research  or Munn's Handbook of Psychological Research on the Rat
T. VERHAVE and J. GILMOUR SHERMAN
are likely to be more difficult to read and teach from. They are the kind of book that Alfred North Whitehead (1929) evidently had in mind when he observed: "Whenever a textbook is written of real educational worth, you may be quite...