Abstract
The term Fixed Action Pattern (FAP) was designed to classify behaviour patterns that are species-specific and constant in form. A major problem is to classify objectively patterns that have been identified subjectively. Many such analyses have been attempted but few have resulted in criteria that can truly convey the essence of the constancy that is intuitively evident. I argue that the unsatisfactory nature of these analyses is a methodological one: behaviour patterns are fragmented into pieces, and fragmentation results in a series of frozen frames which cannot adquately detect constancy of movement. Rather, it is detected in the acutal performance of the movement. Therefore, methods, that measure variables continuously, and thus trace movement in all its spatiotemporal richness are necessary. The Eshkol-Wachman Movement Notation (EWMN) is one such method. EWMN provides a framework which enables several frames of reference to be recorded simultaneously, namely, the path of the movements in space, in relation to some aspect of the environment (e.g. social partner) and in relation to the performer''s own body. When the source of the invariance for a given FAP has been identified, one can ask such questions as how can variability of components lead to form constancy and to what, if anything, is this variability related. Several examples are given.