Dreaming is a conscious experience in its own right: proponents of non-cognitive and non-executive theories of dreaming suffer from a retrospective illusion of their waking extended self.

To many influential dream researchers, dreaming consciousness is not of the same kind as waking. In its most radical and paradoxical form, this theoretical stance consists in maintaining that dream is a case of conscious experience lacking cognitive access. In a more moderate and common form, dreamers have cognitive access to their oneiric experience but lack any executive function: they have no conscious control over their thoughts and actions within the dream. Consideration of dreaming consciousness, in other words, would imply the loss of self-regulation. Neither of these two theories holds. First, because the very reason showing dreams are consciously experienced, i.e. the fact we can recollect them on awakening, implies they are access conscious in the minimal sense that the dreamer noticed them. Second, because, consistent with this first evidence, dream reports also indicate dreamers are able to rationally assess their situation within the dream and self-regulate their dream behavior as a result. I argue, however, that dreamers have reduced, if altered, extended consciousness with limited access to their waking autobiographical self, and that this could explain why many researchers have the retrospective illusion that the dream ego has no rational control over its thoughts and actions in the dream. Indeed, it is not the same autobiographical self that regulates and recollects the dream. Finally, the fact that a dream takes place in the particular conditions of a sleeping brain should not prevent us from recognizing that it is a conscious experience in its own right.