The initial means for arriving at a dynamic model of reading were suggested in the form of "behaviormetric" research. A review of valid reading models noted those of Smith and Carrigan, Delacato, and Holmes as eminent, and it distinguished between models based on concrete evidence and metaphors of the reading process which are basically speculative and theoretical. It was contended that difficulties in measuring what is in progress during reading make developing a dynamic or operational model--as opposed to a static one--very difficult. It was suggested that the way to approach such an achievement is by making no assumptions but by using the latest devices available to measure and record all reading activities which can be observed. Eye-movement data, various physiological measurements, electronic brain monitoring, and measurement of blink rate were examples listed. Such external data should indicate, it was suggested, means to structure a model of internal reading behavior. Other problems that were considered were the need to develop a means of ratioing the relationship of the variables that play a role in reading, the tendency to overrely on static measurements and thus to fog the examination of the operational process, and the need to determine comprehension reliably. References are included. (BT)