Any animal that needs to sleep seems likely to possess consciousness. This would imply that consciousness has been around for a long time in evolutionary terms, and so must serve some important purpose. A lot of effort seems to have gone into optimising the milieu for consciousness; we can hardly guess how many unconscious subsystems silently serve it. Taking this further, say the sense of self descends from own-ness, from the distinction inherent in belonging somewhere, in a territory, or with a group of fellows. You recognise your own, like the dog who knows it own smell when it revisits a tree, something inherent in the olfactory joy. That self is mediated by smell would seem appropriate, it being primarily expressive of an immediate and powerful orientation below the level of consciousness. This means that the sense of self which has also been around for a long time is not original with consciousness, but is inherited by it. (You could argue that insect behaviour expresses self prior to consciousness - pure olfaction, information without joy.) There are perhaps reasons to sever the sense of self from its primeval root and place most of it in consciousness, and these would seem to mostly on the side of consciousness, of aiding in the fuller development of its subsystems. As something like territoriality, the self is always in the context of the boundary between what is your own and what is other, where the other is immediately understood as an alternate self. Beyond the boundary of your world lies the world of someone just like you, only possessing motivations which clash with yours, are incompatible with yours. If there is a howling beast on the borderline it is your self - but such an understanding of self can only exist once self has been almost entirely ingested by consciousness. As for the digestion of self, well, that process has hardly begun.