Abstract
The field of movement behavior has been strongly led by a research tradition in which manipulations are performed in order to search for differences in dependent variables. In this article, I argue for an alternative viewpoint, in which experimental manipulations are provided so that invariances in dependent variables can be seen. If, in the face of changes in nearly countless dependent variables, a particular simple or derived value remains constant, strong suggestions are provided about the underlying control of the motor system. This method is illustrated with examples of invariances in motor behavior, the most important of which is the tendency for the temporal events in a movement to expand or contract nearly proportionally with movement time so that relative timing is invariant. A model following from these invariances, and criticisms of it, are discussed.

This publication has 18 references indexed in Scilit: