The building block theory of language structure, an information processing approach, is applied to the development of a model of the reading process. Specifically, the model is concerned with the amount of time an individual will spend reading and the amount of content he will decode which is determined by a series of cognitive processes involving some aspects of sampling, matching, switching, analysis of subjective probabilities, and analysis of syntax. According to the model, a person intending to read a communication message begins by evaluating the message on the basis of likelihood that it might be comprehensible, sensational, opinionated, etc. and uses that basis to make several hypotheses about the quality of the message. He then gathers information from the syntax of the message with which to evaluate his hypotheses. Once he has decided on the merits of an hypothesis, he continues reading until he has enough information to begin the process again. Several possible applications of this model for reading research are given. One suggests that since people often read only beginning and final paragraphs, the repositioning of important data might require shifting of reading emphasis. Another application might be in training individuals to look for higher levels of abstraction when they read. Research on the model is urged. References are included. (MS)