An open-loop (or entrainment) method is used to control a neuron to a
given goal-dynamic. By specifying a goal-dynamic for a neuron in diffe
rent states, we realised the following controlling processes: chaos --
> order; order --> chaos; chaos --> chaos; order --> order. By changin
g the parameter of a goal-generator, the dynamic behaviour of an entra
ined neuron can be modulated systematically. A possible mechanism unde
rlying the self-organisation of a neural network based on entrainment
is discussed. Copyright (C) 1996 Elsevier Science Ltd.