Research into the patterns of neural connectivity using graph theory, connectomics, suggests that the brain has a small-world architecture, in which most connections between nodes are local, with relatively few long-ranging connections. Activity in long-range networks has been correlated with consciousness.
I consider what other features of cognition might be associated with activity in these long-range networks as contrasted with features of cognition that correlate with activity in local networks.
I argue that these findings offer new insights for the challenge to dual-process theories as to why cognitive processes have particular features that tend to cluster, but importantly can diverge. I then consider how this overall architecture underpins concept possession, which offers insights into language-thought relations.