At present, I find most persuasive the philosophical notion of monism, which is the notion that the brain's activity is not just correlated with consciousness, but effectively is consciousness. From what I can tell, there's currently no evidence to support the notion that consciousness is something independent of brain activity, while there's ample evidence from neuroscience that they're one and the same. As much as it's still an open question, I take this stance for the same reasons I take the atheist stance; the other side makes a weaker case that isn't epistemically justified.
You may be wondering what this has to do with the title, but bear with me for a moment. Despite my position, I wonder why it still seems at least intuitively natural to treat the reds, greens, blues, yellows, blacks, and whites of my experience - my qualia, if you insist - as something separable from the hyper-complicated supercomputer circuitry that is my nervous system (and the amazingly sophisticated simulation software it runs). Come to that, how does the brain see the unseeable, and translate a 2D photon scattering on my retinas (or worse still, a series of 1D signals sent down the optical nerve after the fact) into a 3D model that matches the world near-perfectly? According to reverse optics, it's an impossible task.
Now, I don't have the technical solution to this problem, like the design specifications of the brain's visual lobes, but based on a suggestion taken from How the Mind Works by Steven Pinker, at some point the brain uses a cheat sheet. This is basically information already built into the brain that corresponds with the outside world. The only convincing explanation for how it got there is that the mutations arose that resulted in brains that held a variety of ideas about how the world works, and the ones that weren't whittled out by accidental death or failure to breed happened to best match the environment they found themselves in. Natural selection, in other words.
I don't know about you, but I get chills at this point. What this means is that what I see, hear, smell, touch, taste, and all the rest of it, is essentially the outcome of a series of random guesses - note that the alleles produce the brain's design specifications before the result meets the outside world - that were more useful in navigating the world, rather than being necessarily a perfect replica. In a sense, my entire experience is one big guess that just keeps being coincidentally true enough to be useful. Even more interestingly, this applies to our intuitions as well, such as those about plants, animals, tools, everyday problems, society, and other people.
Dualism, and by extension monism, enter into all this in the following way. A human brain, in addition to having the usual tools for deciphering the world via the sense organs, contains built-in intuitions about other minds. Obviously, a brain can't handle every bit of information about even one other brain, never mind about a whole suite of individual brains that it shares its environment with. So it has a cheat sheet - a shorthand program that allows it to capture the minimum number of properties about other brains while still being practical enough to work - and this manifests as the intuitive idea of dualism. In short, the ghost in the machine (chained or not chained to the brain) - the soul, the self, the spirit, consciousness, mind, anima, ghosts, agents, invisible beings, gods, devils, the "we" who make the choices, all that lot - is essentially a simplified stand-in for the mechanistic brain.
(By the way, a notable side-effect of this is that such dualist entities are treated as atomistic and belonging to a magisterium separate from that of physical or chemical entities. By atomistic, I mean that people aren't usually invited to reduce their dualist entities to simpler parts like a machine or a body could be. The very idea is virtually never raised. This could make for an interesting talking point about some religious and moral ideas, but I'll keep on topic for now.)
Of course, this comes at a price. Such a brain finds it easiest to embrace dualism, not because this intuition is dead-on accurate, but because a brain closer to monism - i.e. one that treated brains as hypercomplex machinery and attended to all the more accurate details of neuroscience - would have been more expensive to produce, and rival brain designs that were more economical happened to be selected for. Another bizarre byproduct is that our dualist intuitions would be strangely nebulous, since all the neuroscientific information of the brains being modelled would necessarily be missing. Also, since our evolved ethical emotions would have coevolved with any model of the minds of others, while such an ethical faculty would have been wasted on models for intuitive physics, this even explains to some extent the moralistic associations of dualism, and why epithets such as soulless and mindless evoke such moralistic horror as they do. It also explains why the evolutionist understanding of humans as genetically programmed survival machines is rarely met with a warm welcome.
I don't mean by all this that dualism as a hypothesis is thereby discredited, though it should make its advocates think twice about appeals to intuition. In fact, the implications for the cheat sheet idea strike me as being the take-home message here. The history of science is characterized by bizarre new paradigms wrestling against intuitive theories - heliocentrism versus geocentrism, the classification of humans as either animals or something distinct from animals, evolution, quantum theory - and perhaps this accounts in part for why some people find science strangely alien, if not outright unappealing. Yet, ironically, their down-to-earth intuitions are themselves strangers of the truth - shorthands and cheat sheets installed by an imprecise guess made long ago by genes.
I think this cheat sheet hypothesis is a fascinating idea that needs exploring. What other intuitions could be approximations, and how did they manifest in our theories about the world? Any speculations?