The structure and dynamics of goal oriented activity were considered with respect to particular task situations. An algebraic method was developed for the detection of the phenomenal parts or the sub-goals units within the activity structure. Further, the same method was used for the quantification of the extent to which the path to the goal was organized. These considerations led to the development of two hypotheses:​ one on the correspondence of the algebraic method to subjective experience, and another on preference for paths that were structurally similar to the task and were 'good' gestalts.

A cognitive model of locomotion was developed to account for the dynamics of a person's movement within the activity structure. The model was based on a numerical representation of the person's experience of locational changes in the course of the activity. The correspondence of the model to subjective experience was expressed in several hypotheses under a central theme of a preference for the paths in which the differences between locomotional variations were maximized.

All hypotheses were tested in the context of four experiments. The first experiment was designed to test the structural hypotheses, and the other three experiments were conducted to examine the various aspects of the locomotional model. The results of these experiments supported the relevance of both the algebraic method as well as the model of the relation of locomotion to activity .