NONLINEAR DYNAMICAL PATTERNS AS PERSONALITY THEORY FOR NEUROBIOLOGY AND PSYCHIATRY

ADVANCES in the theory of nonlinear differential equations and their s tatistical representations have yielded a powerful, qualitatively desc riptive yet quantitative language that captures characteristic pattern s of behavior (what the psychoanalyst Roy Schafer calls ''continuity, coherence, and consistency of action'') that has begun to influence st udies of complex systems in motion as diverse in specifics as signator y patterns of discharge of neurochemically defined single neurons and the dynamical structures characteristic of a particular composer's mus ic. What might be called personality theories of neurobiological dynam ics have arisen to replace neurobiological theories of personality. It is in this way that rigorously proven and powerful general mathematic al insights have changed the face of determinism in research in brain and behavior, Two examples: (1) Very complicated looking behavior of n eurobiological forced-dissipative (expanding and contracting) systems over time take place on low dimensional abstract surfaces on which onl y a few underlying abstract parameters control the action. (2) Indepen dent of specific details (chemical, electrical, and/or behavioral), th ere exist a relatively few fundamental categories of behavior in time and transitions, among them a property called universality. Results fr om this new theoretical, in contrast with experimental, reductionism y ield analogies with and new approaches to historically important dynam ic ideas about personality and character patterns that are equally rel evant to micro- and macrocomplex systems such as neural membrane recep tor proteins and individual personality styles. Research findings achi eved over the past decade and a half in our laboratory and others in n eurochemistry, neurophysiology, and animal and human behavior, as well as the results of a new demonstration experiment involving the predic tion of dynamical category membership from abstract expressive motion in humans, are used to exemplify this use of a quantitative dynamic ca tegory theory across disciplinary levels in brain and behavior. Multip le measures of complexity adapted from current research in the statist ical properties of chaos on unobtrusively observed and reconstructed o rbits on the computer screen made by non-premorbid subjects executing content-free, computer-game-like tasks with a mouse, were used to reli ably differentiate the ''signatures'' of two Axis II diagnoses as esta blished using SCID-II criteria. Whereas the techniques of nonlinear sy stems have achieved some success in quantifying and stimulating the dy namical styles of relatively local phenomena such as the spontaneous b ehavior of neuronal membrane conductances, single neurons, neural netw orks, and field electrical events, we think that the real power of the se techniques lies in their quantitative description and statistical p rediction of global patterns of behavior of entire systems. For exampl e, since the late 1970s our work has shown that these measures could b e used to discriminate categories of drug action and dose when applied to patterns of rat exploratory behavior in space and time. The combin ation of abstract generality and quantitative precision of these metho ds suggests their usefulness as a cross-disciplinary language for fiel ds like psychiatry that deal with complicated behavior of both neurobi ological elements and ''the whole person.''