The Ego Tunnel (2009) is a popular treatment of the ideas from Metzinger’s longer, denser book, Being No One. I’ve tried to read the longer book a couple times and bounced off, so I thought I would try the easier one out. Scattered thoughts follow:
Metzinger’s big idea is that the conscious self is an illusion of the “user interface” of something called a “transparent phenomenal self-model (PSM).” We have our consciousness, and experience of selfhood, because our brain contains a representational model of us (the organism) that can’t be inspected from inside — it is “transparent” — only the “contents” of the representation is available, not the form.
The best part of the book it its discussions of the science of the self-model. Through lucid dreaming accounts, and experiments like the famous “rubber hand experiment” and other experiments targeted at out of body experiences (OBEs), Metzinger goes beyond merely saying that we have a phenomenal self-model, and showing us what its parts are, how it works, and how it may be manipulated to act outside of its normal range and expose its workings.
Metzinger is very good on the science (which he has helped motivate and develop, not just reported), and less good on the claimed implications of the science. He does much to illuminate the phenomenology of the sense of self. But he wants to get from here to the further implications that 1) there is no “real” self, it is completely exhausted by these special effects, and 2) this special effect of a self is exactly what phenomenal consciousness is. I am very sympathetic to 1) and puzzled by 2). But in neither case does Metzinger really sell me.
Unlike Dennett, whose bottom-line claims are similar, Metzinger doesn’t really try to tackle the difficult parts of the theory. Who is being fooled by the illusory self? The PSM is “transparent” to us, but how do we cash that out? Perhaps the longer book answers these questions well. In Ego Tunnel there is basically handwaving about the temporal resolutions of the different processes. The PSM and the associated reality tunnel are transparent because internal monitoring processes do not run fast enough to interact with them at the level of implementation, only at the level of represented content. (It’s also not clear what association is being made between the PSM — the person-model — and the world-model. Some important connection is asserted.)
It is also not clear why this particular magic trick is what leads to conscious experience. For example, Metzinger seems to suggest that a non-transparent PSM would not be conscious. Why not? It seems somewhat counter-intuitive that adding more insight into the machine makes it less conscious. It’s also not clear to what degree the PSM / world-model is transparent — we can at least cognitively assimilate the proposition that appearances are not reality, even if we are born “naive realists.”
Some final musings consider consciousness ethics. Mostly too broad for my tastes, but I copied some passages for flavor. His big ethical commitment is that conscious states are what are morally relevant, or at least — conscious states are morally relevant (possibly among other things; he touches on Nozick’s experience machine). Pain and other contents with “negative valence” become moral bads exactly when they are internally attributed to a transparent PSM — then they seem to be “owned” by a conscious being, and from Metzinger’s point of view that makes all the difference. So there is a very sharp line demarcating moral concern.
We should be very cautious about acts that we perform that create conscious states, he says, and especially in the new area of artificial consciousness. (In Being No One he suggests that, whatever our different positive moral intuitions, we share a wide consensus of negative utilitarianism — suffering is bad.) Evolution so far can be viewed as the tuning and expansion of suffering — it may even be that consciousness itself is a moral bad.