Patterns In Mind (Preprint)

The leading theories of consciousness offer productive models for understanding consciousness, but they tend to stop short of detailed operational definitions. Predictive processing does not suggest how phenomenality arises. Global workspace theory assumes consciousness follows ‘ignition’ but leaves qualitative aspects unexplained. Higher-Order Thought (HOT) theories suggest why mental states are conscious but not how higher-order thoughts are constructed. In this paper I present a pattern-oriented approach to thinking about the content of mind that has broad capabilities and a unifying potential for existing theories. Under a core axiom of pattern handling, principles which construct and process hierarchical patterns are used in a thought experiment to explore the development of consciousness in an imaginary creature step by step. The exploration shows consciousness has a high likelihood of occurrence in any creature with sufficient pattern handling capability. Insights gained include operational-style descriptions for the origin of awareness, building higher order thoughts, a mechanism for attention and a candidate approach to the meta-problem. The same principles are subsequently used to describe a candidate approach for phenomenal consciousness (which reclassifies rather than closes the explanatory gap). They are also effective in describing mechanisms for other features of mind including: the threshold of introspection; fast and slow thinking; the origin of language; causal reasoning; and the flaws found in intuitive problem solving. The model is shown to be complementary to the leading theories of consciousness and provides a way to discuss cognitive content. The mechanisms described offer possible ways to assess claims of AI consciousness.