ABSTRACT
Attneave (1971) and Lehky (1988) pointed out the similarity of cognitive bistability and electronic multivibrator circuits and suggested analogous neural structures with locking into alternative schemata and exhibiting fatigue. In contrast to the microscopic neural approach I propose to model the dynamics of the macroscopic behavioral variables perception and attention by a phase feedback equivalent circuit which is related to the mean field theory of temporal binding (Schuster & Wagner 1990). Based on a previously outlined nonlinear dynamics model (Fürstenau, 2003), I show that the spontaneous reversals between two perception states with an ambiguous stimulus, e.g. the perspective switching of the Necker-cube, can be explained as self – oscillation between chaotic attractors in attention – perception phase space. The perception state variable is represented by the phase v of a recursive cosinuidal mapping function with feedback delay time T and attention control parameter G. G is proportional to feedback gain g of a corresponding equivalent circuit representing the dynamics of behavioral variables v, G. According to Hillyard, Vogel & Luck (1999) a difference between bias and gain control mechanisms of attention is observed. Like in the multistability model of Ditzinger & Haken (e.g. Haken (1996)) the perception variable is treated as order parameter within the formal framework of Synergetics, with the slowly time varying attention parameter G(t) exhibiting satuation or adaptation. The important aspect of the present approach is the neurophysiologically motivated delay T, giving rise to the chaotic attractors in agreement with Freeman et.al. (e.g. Freeman 2000). The statistical analysis of simulated time series predicts gamma distributions of the perceptual reversal times with mean values and variance in reasonable agreement with experimental results of Borsellino et.al. (1972).
