Search Results

This chapter argues that brain systems must meet several 'hard' requirements to qualify as conscious, while other requirements are 'soft' in that they are important for sustaining normal, daily-life awareness but not strictly necessary for having the most basic form of conscious experience. The hard requirements include, first, the ability to interpret (or reconstruct) sensory inputs as having particular qualities or content, within a rich repertoire of modalities or (sub)modalities, such as visual motion, shape, depth and color. Second, this process of attributing sensory “feel” or meaning to inputs occurs in a dynamic or stable state, depending on the constancy of variables governing the sensory flux. Projection of interpreted sensory inputs into an external, perspectival space (vision) or body map (somatosensation) is seen as a relatively basic process, but patient studies indicate that core consciousness does not strictly depend on this ability, as applies as well to normal requirements on the grouping of similar features and binding of different submodalities into objects. Also the “unity” of consciousness and self-awareness are not classified as an essential feature but rather as a constantly maintained “illusion” of the healthy brain empowered by proper multimodal and motor alignment.

What are neural network models, what kind of cognitive processes can they perform, and what do they teach us about representations and consciousness? First, this chapter explains the functioning of reduced neuron models. We construct neural networks using these building blocks and explore how they accomplish memory, categorization and other tasks. Computational advantages of parallel-distributed networks are considered, and we explore their emergent properties, such as in pattern completion. Artificial neural networks appear instructive for understanding consciousness, as they illustrate how stable representations can be achieved in dynamic systems. More importantly, they show how low-level processes result in high-level phenomena such as memory retrieval. However, an essential remaining problem is that neural networks do not possess a mechanism specifying what kind of information (e.g. sensory modality) they process. Going back to the classic labeled-lines hypothesis, it is argued that this hypothesis does not offer a solution to the question how the brain differentiates the various sensory inputs it receives into distinct modalities. The brain is observed to live in a "Cuneiform room" by which it only receives and emits spike messages: these are the only source materials by which it can construct modally differentiated experiences.