This research paper will appear in substantially its present form in G. David Hughes, Michael L. Ray and George Haines (Eds.) Consumer Information Processing, Chapel Hill, N. C.: U. of North Carolina Press, 1974. The book is based on a conference supported by the Association for Consumer Research and the American Marketing Association at the University of Chicago in early November 1972. The conference had three parts: information search, initial processing and central processing. This paper constitutes the introductory chapter for the “initial processing” section of the book. Initial processing is here defined as communication and persuasion, related most closely to the applied area of marketing, communication. Dependent variables for research are communication goals of marketing communication. Independent variables are the message strategy and message distribution aspects of the applied area. Five developments-controversies are discussed: 1. One is whether sales or communication goals are the appropriate objective of a communication campaign. This has been resolved within a planning framework. Sales have relevance for setting the tentative budget mix, communication goals for the implementation of expenditures. As such, the latter goal types and measures are essential for studying initial processing. 2. Another controversy involves the order of the stages of communication effect. Originally there was one “hierarchy” of effect proposed. Now there are three that have been researched and applied in the field. 3. Should initial processing research be problem- or theory-oriented? Both types of orientation are being taken, but there is a movement toward research on problems of the middle range instead of theories the middle range. 4. A fourth controversy is whether initial processing research should be carried out in the lab, the field or in computer simulation. A recent development is the trend toward using these alternative sites for research together in a systematic way to attack particular problems. 5. Finally there is the ubiquitous quandry of whether to con- centrate research on measures or on process. Although measure validation is needed in initial processing research, there is more emphasis recently being put on small, well defined processes.