The What and Why of Binding: Review The Modeler's Perspective

our parlance, mechanisms of organization). Neurosci-ence and neural modeling, on the other hand, have the ambition to find general answers. It is this commitment Christoph von der Malsburg* Institut fü r Neuroinformatik Ruhr-Universitä t Bochum to generality that results in the binding problem being 44780 Bochum a fundamental feature of the neural code. Federal Republic of Germany The Psychological Moment University of Southern California Before discussing the answers to the above questions Los Angeles, California 90089 that are postulated by classical neural networks, it is important to introduce an important parameter. Although not often made explicit, it is important to fix a In attempts to formulate a computational understanding temporal scale T, which we will refer to as the " psychological moment " (in the sense of " short period "). At times of brain function, one of the fundamental concerns is shorter than T, one speaks of mental state or brain state, the data structure by which the brain represents infor-whereas at times greater than T one sees a succession mation. For many decades, a conceptual framework of states or a " state history. " Whereas state history is has dominated the thinking of both brain modelers and subject to conscious scrutiny (that is, it potentially gets neurobiologists. That framework is referred to here as reflected in all modalities—memory, language, etc.), no " classical neural networks. " It is well supported by ex-such conscious analysis is possible below T. State his-perimental data, although it may be incomplete. A char-tory is ignored by most or all models, and the conceptual acterization of this framework will be offered in the next disagreement about the binding issue focuses exclu-section. sively on the definition of state, that is, on times below Difficulties in modeling important functional aspects T. It is difficult to pin down a definite value for T, but a of the brain on the basis of classical neural networks plausible range may be from 50 to 200 msec. Regardless alone have led to the recognition that another, general of the exact value, it is important to realize that the arena mechanism must be invoked to explain brain function. for the discussion of the binding problem is at a time That mechanism I call " binding. " Binding by neural signal scale less than T. synchrony had been mentioned several times in the liter-before it was fully of Brain Function formulated as …