Facial emotion recognition by adaptive processing of tree structures
Proceedings of the 2006 ACM symposium on Applied computing
| Hi-index | 0.00 |
The controversy between the theory-theory (TT) and simulation-theory (ST) has evolved so that it is often hard to tell exactly what the difference is between a simulation and a theory. I believe that this distinction was originally inspired, and can be freshly reconceived, as the distinction between verbal abstractions and concrete pictures. I argue that the multi-dimensional spaces described by connectionist neuroscience are best understood as pictures of a special sort. These multi-dimensional pictures do not have the limitations of ordinary three-dimensional pictures, and are capable of performing many of the cognitive functions that were traditionally thought to be the exclusive domain of abstract linguistic concepts. Consequently, there is a real possibility that a pure simulation theory could actually explain some sophisticated kinds of social cognition, without having to rely on a hybrid that combines simulations and theories. Paradoxically, such a pure simulation theory would not actually use simulations in the strictest sense of that word, because something can be a simulation only if it is verbally labeled as a copy of something else. Rather this kind of social cognition would establish vector transformations between perception and behavior without requiring any verbal labels at all. This would mean that the emotions caused by perceptions of other people would not be simulations of other people's emotions, but rather the same emotion transferred by a kind of emotional contagion.