In my work I combine two main aims: describing the behavior of organisms from the presumed point of view of the organism (the organism as an engineer, not a machine), and uncovering behavioral homologies across the animal kingdom. Both aims are accomplished by using a phenomenological approach.
I have been guided by the following principles:
1. Behavior is an extension of anatomy, therefore it has a structure.
2. Much like in comparative anatomy, it should be possible to identify behavioral homologies across the animal kingdom at the level of geometry and kinematics, low level representations that allow comparisons across distant systematic groups, highlighting universal structures.
3. EW Movement notation is a symbolic language of representation fulfilling the above requirements. It has the necessary features for reflecting the interaction between the body and the environment in terms of the organism's own presumed experience ( in reference to gravity, to absolute and body-related coordinate systems, and to the body part(s) supporting the body ).
Other principles that guide my analysis of behavior:
4. Identifying the origins used by the organism itself and measuring behavior in reference to that origin (e.g., Contact with weight in locomotor behavior, immobility in egocentric space, home base in absolute space, maintained relationship of opposition with another organism in partner-related space).
5. Conceiving of the animal's operational world (Umwelt) as the sum total of the (measured) quantities actively managed by it.
6. Studying the developmental dynamics of behavior in situations and preparations in which the behavior is allowed to unfold gradually, from simple to complex and vice versa, yielding a kind of electrophoresis of the behavior to its intrinsic constituent processes.
The second half of my career is marked by my cooperation with Yoav Benjamini, with whom I joined forces, both consolidating the above principles and adding new ones:
7. Extensive preparation of the data for analysis: this includes elaborate smoothing, segmentation, customization and stratification of the behavior into its constituents (SEE).
8. Identification of the elementary units of behavior on the basis of intrinsic statistical and geometrical constraints. These units are particulate processes rather than discrete inert blocks. They exhibit specific developmental dynamics.
9. Quantifying the extent and complexity of behavioral growth and decay processes.
10.. Using the degree of replicability of behavioral results across laboratories as a touchstone for the quality of the quantified measures.