Discussion of "Signatures of Consciousness", Stanislaus Dehaene

Edge discussion and original talk here.

Stan Dehaene always deals with interesting, deep problems in a novel and efficient manner. His last two publications deal with the evolution of a number sense in primates and the evolution of reading ability (a proposed volutionary change based on a cultural phenomenon), and now Dehaene attacks the problem of consciousness in terms of a global neuronal workspace. The quick and dirty version of his argument is that, based on fMRI and temporal EEG measures, it seems the consciousness exists as a kind of neural workspace in the mind; consciousness is something that parallels (or is?) our working memory and exists independently of neural processes and responses, following behind the initial response to stimuli by an average of a quarter second.
I have great respect and admiration for Dehaene, but there's a major flaw in his discussion (and since his talk is speculative, that's acceptable): He explains well the purpose of consciousness, and perhaps even what we might define as conscious, but he fails in any meaningful manner to demonstrate what causes it.
Sam Harris has a critique of Dehaene's points in terms quantum computers, which I won't even attempt to address.