The Case Against Reality

This is a great book on the way we perceive the world from the perspective of evolutionary biology, but it drifts off into incoherence in the last chapter. The argument is, essentially, that we have evolved to experience "objective reality" not in terms of truth, but fitness. We see things in a way that help us avoid danger and reproduce, eat, etc. This isn't to say "stuff" doesn't exist, but that humanity's understanding of it is not--and could never be--universal. The way we experience it is merely an interface. Spacetime, motion, and matter are all elements of this interface.

Hoffman echoes a lot of Kant's phenomena/noumena distinction here, and the noumena are intrinsically unknowable and unperceivable.

The place where the book doesn't make sense is the last chapter. There he argues that there is an objective reality, and we can know things about it, even if we cannot directly perceive it. He offers the idea of "conscious realism," made up of "conscious agents" (in contrast to "physicalism," which dominates the sciences today) which construct the universe. This is interesting, but it's so vague as to be meaningful. I guess conscious agents are merely "nodes" of experience, without being things of themselves, and all else arises from it.

It's certainly an interesting thought and worth playing with, but I don't think he goes far enough into the philosophy here. He's pushing at the limits of what knowledge can even be, and this requires falling back on the philosophy of science and mathematics. He's deeply critical of physics from a philosophical standpoint, and he comes at biology to lesser extent. But mathematics seem almost untouchable to him. After all, his understanding of conscious agents seems to come out of mathematical formulae. But, might not mathematics also be invented within this interface? After all, it is a profoundly complex formal language that emerges out of a few axioms that are agreed-upon as being self-evident, but may not necessarily be.

Still, it was nice to see this sort of work coming out of the sciences, trying to push forward the science of consciousness, rather than being stuck in philosophy. I'm not sure that he has been able to escape the Kantian cul-de-sac here, but it also seems that no one has for more than two hundred years now.

Hoffman's book is well worth the read, although it has its limits.

The starting premise here is that evolution has shaped not only our senses but how we interpret the data we gather from our senses. Not a great leap, especially given what's happened in perceptual science and neuroscience over the last few decades, although Hoffman acts as if this premise isn’t a prevailing belief among vision scientists (which I doubt is true). Much of the usual evidence in support of this premise, vis a vis optical illusions and the like, is presented (or, to be a bit more blunt, regurgitated in fitful spasms), and then Hoffman goes further.

Hoffman claims that evolutionary "fitness payoffs" are the basis of all of the sensory information we, or any other living thing, can gather. These fitness payoffs represent a means of preserving and extending our genetic heritage and are neither tied to nor reflective of Objective Reality, which, Hoffman says, is reflected in a mathematical theorem he’s devised, the FBT Theorem. This idea is stated and restated a nauseating number of times, so I understand why other reviewers have complained about the book being repetitive. But it gets worse: Hoffman then takes this idea to such an extreme that he runs his argument off the rails, ditching it in a valley so deep it can never be rescued.

"...the idea that physical objects are just ephemeral data structures that describe fitness payoffs differs sharply from the idea—now standard in vision science—that physical objects are elements of objective reality, and that the goal of vision is to estimate their true shapes and other physical properties."

So it isn't merely that our interpretations of sensory data are "ephemeral data structures" not truly representative of Objective Reality, but that there is no such thing as Objective Reality. According to his FBT Theorem, there's no way for perceptual systems to represent Objective Reality, and therefore, the Objective Reality that we believe in does not exist. He tries to make this point, again and again, by saying that even if we look up at the sky and see the moon, when we look away, the "moon no longer exists." Not in the sense of a baby looking away from something and not understanding object permanence—no: there really is no moon.

If this sounds absurd, it is. Hoffman tries to argue ontologically and philosophically in favor of this absurdity, but there is no actual science (and no proposed experiments or existing experimental data) to back up this conclusion. And then he goes even further, and it gets worse, because he thinks he can lean on quantum physics, that netherworld of neither here nor there, for support of said absurdity.

Now I am not trying to be flippant, but how is it possible that a professor of cognitive science who has had papers published could so fundamentally misunderstand quantum physics? In chapter after chapter, Hoffman tries to suggest that because current interpretations of quantum theory imply that spacetime isn't what underpins Objective Reality, there is no Objective Reality, and his FBT Theorem must be correct. Doesn't Hoffman know what decoherence is? Doesn't he understand that decoherence is one of the reasons I can't be in a superposed state of existence, existing in New York and in the star system of Alpha Centauri simultaneously? Quantum systems decohere not only when they are measured but when they interact with the world around them (Objective Reality). That's why classical physics works the way it does (for the most part) when we zoom out from the quantum world—because of decoherence.

Our senses work in that zoomed out world, not on a quantum level. So it isn't just specious to try to make arguments about whether we are sensing Objective Reality by referring to quantum theory, it's just wrong-headed and, forgive me, stunningly stupid. Our "fitness payoffs" aren't based in the world of the quantum: they are based in the world that arises from the quantum as quantum states decohere. Therefore our "fitness payoffs" reflect our existence in a world where such quantum effects as entanglement are negligible or non-existent.

I'm not suggesting that there aren’t "fitness payoffs" tied to the "ephemeral data structures" in our heads, nor am I claiming that those data structures DO represent Objective Reality. What I'm saying is that a statistical theorem (Hoffman's FBT Theorem) not tied in any respect to physics cannot prove that Objective Reality does NOT exist. The moon is there when we don’t look at it, regardless of what Hoffman wants us to believe.

I wish it ended there, but it doesn't; it gets worse. In the last chapter, Hoffman tries to tell us what reality actually is, since Objective Reality is simply an aggregation of evolutionary fitness payoffs. His answer? Conscious agents, and another (even weaker) mathematical theorem.

That's right: the moon isn't there, because we live in a universe of conscious realism. There is no spacetime; there is only consciousness.

The expression "Come back down to Earth" seems fitting here.