Review: “The Case Against Reality”
This is not a red stop sign:
For one thing, in a ceci n’est pas une pipe way, it’s not a stop sign at all, but a digital representation of a photograph of a stop sign, made visible by a computer monitor or maybe a printer. More subtly, “red” is not a quality of the sign, but of the consciousness that perceives it.[1] Even if you were to generously define “red” to be a specific set of wavelengths of light that could in theory be ascertained without the benefit of consciousness, that still would be a measurement of something that the sign repels rather than is.[2]
We on some level understand these things, but we seem much more prone to think of ourselves as living in a world in which stop signs are red; the redness inheres in the sign out there in the world, rather than the redness being our own reaction to light reflecting from the sign, or rather than the sign and the redness being coincident private figments of perception. And no amount of thinking this through seems to dispel the illusion: it’s just too practically useful to put the color onto the external object to make it seem worthwhile to relearn how to perceive reality differently.
Donald Hoffman, in The Case Against Reality (2019), suggests that our mistake here goes much deeper. Not only are superficial sensory characteristics like color subjective qualities of conscious experience rather than objective qualities of things — but things themselves, as well as the spacetime they seem to inhabit, are too.
All of the stuff we take for granted as making up the world we inhabit — objects, dimensions, qualities, time, causality — are, says Hoffman, not objectively real. They are better thought of as the interface through which we interact with whatever might be objectively real. When we isolate “objects” in “space” and “time” and then ascribe “laws” to the regularities in their interactions, we are not discovering things about reality, but about this interface. “Physical objects in spacetime are simply our icons in the desktop.”
Hoffman compares this to the kind of thing that happens when you “drag” “a file” on your “desktop” to “the trash.” All of these things are meaningful and have practical effects in the interface, but cannot be taken seriously as descriptions of what actually happens under the hood. Similarly, Hoffman says, you need to take what you see in your “real world” interface seriously, but not literally. Your interface does not depict reality, but provides you with a comprehensible way to get by.
This, he calls the “interface theory of perception.”
This wouldn’t be such a big detail if our interfaces were just low-resolution versions of reality, approximations to it, or imperfect mirrors of it. But according to Hoffman, it’s much worse than this. There’s really just no resemblance between the interface and the underlying reality at all.
He believes that we have come to have such an interface rather than a reality-tracking worldview because “fitness beats truth” in natural selection. I think his FBT (“fitness beats truth”) theorem is something like this:
Any biological mechanism of perception is costly: it requires resources to build, maintain, and operate.
More sophisticated mechanisms (those that can extract more information) are, generally speaking, more costly.
At some constant level of complexity/cost, some mechanism of perception can extract some certain amount of actionable information from what it perceives. There is then the question of which bits of information to extract.
Such bits of information can be more or less veridical (information that closely tracks what is true about what is perceived) or practical (information that closely tracks whatever is relevant to the fitness of the perceiver).
These do not necessarily track one another. In other words, one bit of practical information (X is/is not relevant to my fitness) may not correspond to one bit of real information (X is/is not quite like so). Hoffman believes that this is typically the case: that practical perceptions do not represent actual facts, but are jumbled up functions of facts about what is perceived, the context in which it is perceived, and the state of the perceiver. We cannot easily disentangle this and go backwards from our perceptions to try to recover just-the-facts.
Natural selection will favor perceptive apparatuses that extract practical information; it will not operate to preserve equivalently-costly ones that extract merely veridical information. For this reason, the latter don’t stand a chance.
Ergo, everything we perceive is natural selection’s attempt to give us handles we can manipulate to increase our fitness. We have no reason to expect that such handles represent what “is there” or exists outside of our own interfaces. (You may say your interface has the same icon on it that mine does, and that it behaves in the same way in your interface as in mine, but this does not prove that we both have true representations of some underlying fact, but merely that we both have similarly-evolved interfaces. We both say the stop sign is red, but it isn’t.)
What this adds up to is that “spacetime is a communications channel, and physical objects are messages about fitness.” And “the appearance of causal interactions between physical objects in spacetime is a fiction” in the same way that dragging file icons around on your computer desktop is. It is a mistake to believe that spacetime is in fact the reality in which we objectively live.
Hoffman illustrates the limitations of our perception with some optical illusions and other such data, which help to show the limitations of our perceptive apparatus even when expressed in terms of the interface it attempts to represent. And he gives some examples of how people (and animals) manipulate how their icons appear in the interfaces of predators and potential mates, to a similar end.
He also spends some time looking at the usual baffling conundrums of modern physics (double-slit experiment variations, collapsing wave functions, observer-disagreements about simultaneity or the passage of time, etc.) through the lens of his theory. He concludes that the reason why spacetime & why particles behave so oddly at the extremes of observation is because there is no spacetime, there are no particles; those things are just our handles in the interface, and natural selection did not select for handles that remain intuitive at the scales we’re now investigating them in. Is light a wave or a particle? No. Waves and particles are icons on our interface; whatever “light” “really” “is”, it’s not that.
So what is reality composed of? Hoffman’s bet is that it’s consciousness all the way down. This “conscious realism” theory begins by defining conscious agents as things that respond to perceptions from the world with actions which then influence that world, and the cycle repeats. So “consciousness,” here, can be a pretty simple thing and shouldn’t be confused with self-consciousness / awareness. Hoffman believes that very simple (maybe even one-bit) consciousnesses can combine into more complex ones, and those into more complex ones yet. Our own personal consciousnesses are examples of these. But all consciousnesses ultimately interface not with a material world, but with one another. The “world” they perceive and that their actions influence is nothing more than the world influenced and perceived by other conscious agents. It is not some underlying material world that exists independently from those agents.
As you might guess, this is a difficult set of propositions to argue for. For one thing, you can hardly string a sentence together without implicitly asserting the reality of things, causality, time, space, and the like. So Hoffman has to frequently make assertions about the world to back up his theories, where those assertions seem to implicitly undermine those very theories. For example, how do you argue that natural selection shaped our perceptual abilities in such-and-such a way without a spacetime in which such perceptual abilities can be subject to selective pressures?[3]
Hoffman’s book, also, is not the most excellently-structured argument. It can be repetitive about dogmatic points, and breezily fly by more crucial arguments (like the fitness-beats-truth theorem) that would benefit from a more detailed and penetrating analysis.
I am also anxious about an argument of this sort: Philosophy, physics, metaphysics, and so forth are meant as tools that help us to better understand and rise to the challenge of the world we encounter. If you begin to entertain a theory of philosophy, physics, metaphysics, or what have you that seems to be taking you on a path like this one:
There is a spoon on the table.
Is there really a spoon on the table?
“Is” there “really” “a spoon” “on” “the table”?
⋮
I am Elmer J. Fudd. I have a mansion and a yacht.
…then this is perhaps a good sign that you need to put that theory of philosophy, physics, metaphysics, or what have you, down carefully, and to back away from it with all haste.
- ^
Oh, but what is the “sign” if not also an artifact of the consciousness that perceives it… Does a sign exist if it signifies nothing to anybody?
- ^
Sure: if you are reading this on a screen, then the “stop sign” is represented by emitted rather than reflected light, but still...
- ^
If I remember right, his answer to this is that the theory of evolution via natural selection is sufficiently generalizable that it can hold up under a whole wealth of conditions, of which the spacetime of our interfaces is only one. Besides, there is a great deal that we can learn about how the icons that represent our perceptual interfaces behave in the terms of our interfaces, which (again in the terms of our interfaces) was shaped by natural selection in spacetime, that is also consistent with Hoffman’s theory. Or something like that.
I read the book, it was interesting, however a few points.
Rather than making the case, it was more a plea for someone else to make the case. It didn’t replace the conventional theory with its own one, it was far too short and lacking on specifics for that. If you throw away everything, you then need to recreate all our knowledge from your starting point and also explain how what we have still works so well.
He was selective about quantum physics—e.g. if reality is only there when you observe, then the last thing you would expect is a quantum computer to exist and have the power to do things outside what is possible with conventional computers. MWI predicts this much better if you can correlate an almost infinite amt of world lines to do your computation for you. Superposition/entanglement should just be lack of knowledge rather than part of an awesome computation system.
He claims that consciousness is fundamental, but then assumes Maths is also. So which is fundamental? You cant have both.
If we take his viewpoint then we can still derive principles, such as the small/many determine the big. e.g. if you get the small wrong (chemical imbalance in the brain) then the big (macro behavior) is wrong. It doesn’t go the other way—you can’t IQ away your Alzheimer’s. So its just not clear even if you try to fully accept his point of view what how you should even take things.
This, I think, is the most indefensible claim made, if that’s what he is actually saying. OF COURSE there’s a resemblance. If there weren’t, there would have been nothing for natural selection to act on to produce this interface instead of some other. It is an approximate and slight resemblance, very much so, but claiming it is not there is just nonsensical.
It’s not nonsensical. It’s an assertion that can be made sense of with a little effort.
Consider the user interface analogy. On your desktop there is a mouse pointer with which you can drag a file from here to there. In the underlying computer which executes the actions which are represented by this interface, there is nothing that resembles a pointer, a dragging action, or a file. That the interface associates certain activity in the hardware with certain things that appear on the desktop is a useful convention for us, but it is not one that was designed to give us an accurate notion of what is taking place inside the machine. Hoffman suggests that the same thing is true of the interface-reality we perceive and the real-reality underneath. The interface-reality was “designed” by natural selection to be a useful convention for us as we interact with the real-reality which is not apparent to us.
That, to me, reads like a very different statement. One I completely agree with. But different. I maintain that the very fact of the interface’s usefulness demonstrates that there is enough of a resemblance to the underlying reality to create said usefulness—by giving us accurate expectations about selected aspects of the behavior of whatever is going on beneath the surface, while lowering mental overhead by abstracting away the rest.
Edit to add: side note, who says the word red has to be defined based on wavelengths in absolute terms instead of relationally? I can create a less inaccurate but also usually less useful definition based on expected qualia produced in a set of observers in response to ambient light of particular kinds of spectra. Repeat until you get as accurate as desired, with each step eliminating an abstraction that’s normally subsumed in the interface.
Course’ there are: probably 64 bits in memory (for more degree of detail: 64 places of persistent electric charge with two stable states), which change iff pointer moves, and each bit restricts the places pointer can appear at. That resemblance exists certainly; I also agree there’s no resemblance like “small pointer-like thing/charge pattern in RAM module”.
In other words, one has to taboo “resemblance” but it’s not clear if that can be done.
We taboo resemblance all the time for things that refer to other things: Words, for example. The word “mouse” does not resemble a mouse, but we can usefully use the word as a reference. Words that resemble their references are a peculiar and remarkable tiny category (onomatopoeia) that are the exception to the rule.
If you thought your computer interface were an accurate picture of what is going on inside the computer, you might indeed go looking for a microscopic pointer somewhere in the wires. It’s because you don’t think this that you know to look for correspondences and representations instead. Hoffman’s point is that we don’t tend to do this with things like space, time, matter, etc.: we think those things in our interface-with-reality correspond to the same sorts of things in reality-under-the-hood (space, time, matter, etc.). He believes we’re mistaken.
Ok, fair, I agree they do not resemble one another in this kind of way. That isn’t how I was interpreting “resemblence.” I was thinking of it as “correspondence.”
Minor off topic aside: While they are a small minority, I suspect onomatopoeia-like words are probably more common than we usually think. A lot of words resemble, to me, aspects of their referent, even if only in a synesthetic way, or only if you look at older forms of the word. Add that to things like the bouba/kiki effect, and I start to wonder how much we coin new words using some kind of felt, metaphorical sense of meanings corresponding to how sounds feel.