Neural assemblies and circuits involve computational operations which are at the same, low level of computational abstraction. The general properties of neural biocomputers are determined, and their contribution to cognitive tasks assessed. 

 

One of the main reasons for the success of the TDE project was the way the TDE-R cognitive architecture fitted the known epistemological overview of the human mind. However, several barriers to that goal stood in the way, the most serious of which was semantic grounding. In simple terms, how do we (or any animal) know what we are looking at? Although this seems to be a low-level perceptual (image processing) issue, it seems not dissimilar to the way we understand language. Each perceptual symbol, whether processed concurrently within a sensory field, or processed sequentially, as a stream within a sensory channel, depends on the interaction between two abstract parameters- context (external code) and content (internal code).

 

This problem (and others, such as SLAM*) are solved in the TDE system by a neural circuit element called a Pose Cell Array (PCA).