One way to think about perception, probably the most natural way, is to compare it to a window, where your mind simply reads out reality exactly as it is. That chair over in the corner, for instance, is the exact shape, color, and texture that you perceive it to be, and your mind is simply capturing this object exactly as it exists in the world. 

But if you think about it, this cannot be the case; your brain—enclosed as it is within your skull—is encased within a pitch-black environment. It has no direct access to light, sight, sounds, or anything else in the external world. The only thing the brain receives are electrical impulses from its sensory systems that push around chemicals. This flurry of electro-chemical activity somehow produces subjective experience, but this phenomenological reality is nothing more than an approximation, or, as neuroscientist Anil Seth puts it, a “controlled hallucination.”

In Being You: A New Science of Consciousness, Anil Seth, professor of Cognitive and Computational Neuroscience at the University of Sussex, explains the latest science of consciousness and how all evidence points to the “controlled hallucination” view of consciousness. In this view—which in large part validates the transcendental idealism of Immanuel Kant—reality is not what it seems; the mind evolved for action and survival, not to perceive reality in an objectively “true” sense. We label as “reality” the shared world created by our minds, but as for what the world is like “in itself,” we can never really know, because our brains can never transcend the electro-chemical signals processed from within our “bony prisons.” 

One example that drives home the point is the infamous phenomenon of “the dress.” As you’ll remember, the image of this particular dress was seen by some as black and blue (what it was marketed as) and seen by others as white and gold. Those that perceived the dress as white and gold were much more likely to assume the dress was in a shadow, while those that did not make this assumption perceived the dress as black and blue.

So you have two groups of people looking at the same image and perceiving two separate sets of colors. What makes this interesting is that it clearly shows that color is not a property of objects themselves; the dress is not, in itself, one color or the other. Rather, color is a subjective phenomenon that arises from the way the mind perceives the interaction of particular wavelengths of light bouncing off of particular objects. 

What makes this example even more interesting is that it lends support to Seth’s hypothesis that the mind is essentially a prediction machine, creating its world from the inside-out and making corrections based on sensory input from the environment. I mentioned that those that perceived the dress as white and gold assumed that the dress was in a shadow. That means that for this group, their brains were making predictions about the dress, its environment, and its light source, and then “coloring in” the dress based on those predictions. Then, after examining the dress for a while, some members of this group started to see the dress slowly turn to black and blue, as their minds updated the relevant predictions. Color, therefore, is a subjective phenomenon crafted by the brain based on predictions from the inside-out and corrected by environmental cues (this ability to make corrections marks the difference between standard hallucinations and the “controlled” hallucinations of everyday experience).

Where things get trippy is that this doesn’t end with color, but applies to all sense modalities. Seth describes how the mind, by making these kinds of predictions, creates our entire perceptual world, including its associated colors and sounds and even our perceptions of space, time, causation, and our sense of self and free will. We truly live in a virtual reality created by our minds, and as to what lies beyond this inner world, we’re literally in the dark.

Counterintuitively, this recognition actually makes the science of consciousness more manageable. If we take phenomenological reality for what it is, we can then work on explaining, predicting, and controlling it, as has been done in the example of “the dress.” We can work on measuring consciousness and mapping its neural correlates without having to try to understand why consciousness exists in a philosophical sense or how exactly consciousness can arise from purely physical phenomena. 

Which brings us to an important point: this book does not solve, in any way, the “hard problem of consciousness,” and the author makes no claims that it does. Seth does not show, for example, how the electro-chemical processing of light—a purely physical process captured in quantitative terms—can result in the actual phenomenological experience of, for instance, “seeing red.” This problem is, as the philosopher Colin McGinn has stated, beyond the capacity of the human mind to comprehend. As Seth wrote (summarizing McGinn’s views but not necessarily endorsing them):

“Mysterianism is the idea that there may exist a complete physical description of consciousness—a full solution to Chalmers’s hard problem—but that we humans just aren’t clever enough, and never will be clever enough, to discover this solution, or even to recognize a solution if it were presented to us by super-smart aliens. A physical understanding of consciousness exists, but it lies as far beyond us as an understanding of cryptocurrency lies beyond frogs. It is cognitively closed to us by our species-specific mental limitations.”

I tend to agree. Despite all the advances in science—and in neuroscience specifically—we are not any closer to solving the hard problem, and it’s difficult to see how more of the same neuroscience—which considers only correlations between brain states and experiences—could ever get us closer to an answer. Humans, representing one species of primate, are able to sense just a small sliver of total reality, and therefore cannot be expected to solve every mystery presented to us. Consciousness in the “hard problem” sense appears to be one of those mysteries that will remain unsolved. 

However, all hope is not lost. Seth spends a majority of the book intentionally side-stepping the hard problem and focusing entirely on what he labels the “real problem” of consciousness. As Seth writes:

“This is the essence of the real problem approach to consciousness. Accept that consciousness exists, and then ask how the various phenomenological properties of consciousness—which is to say how conscious experiences are structured, what form they take—relate to properties of brains, brains that are embodied in bodies and embedded in worlds.”

This approach dissolves, rather than solves, the hard problem. Elsewhere, Seth writes:

“Dissolving the hard problem is different from solving it outright, or definitively rebutting it, but it is the best way to make progress, far better than either venerating consciousness as a magical mystery or dismissing it as a metaphysically illusory nonproblem.”

By taking this approach, Seth shows how the science of consciousness, far from being a mere philosophical curiosity, has major practical implications, from medical applications (measuring consciousness in nonresponsive patients) to the psychiatric treatment of hallucinations to approaches to the creation of artificial intelligence and even perhaps to consciousness in machines (although it’s questionable whether or not we should even attempt to go down this road).

In terms of creating consciousness in machines, Seth is more skeptical. His entire theory of consciousness incorporaters the biological and physiological realities of brains embedded within bodies and worlds. Brains cannot be conscious without these biological elements; the “controlled hallucinations” of the brain depend on multiple layers of complex interactions with the physical world and with the trillions of cells and chemical signals within the body. If you remove the biology—and replace it all with silicon chips—don’t be surprised if you end up with an intelligent machine that doesn’t experience anything at all (i.e., lacks consciousness). 

As Seth correctly points out, there is no necessary correlation between intelligence and consciousness; they are separate phenomena. Relatively unintelligent animals probably have inner experiences and can feel pain, whereas highly intelligent machines playing games of chess almost certainly are not experiencing any sense of joy or accomplishment from winning a match. 

And so despite being bombarded with sensationalistic predictions of machines taking over the world—and by countless books, movies, and television shows dramatizing how it will happen—the prospects of conscious machines coming into existence is, at least for now, very remote (although non-conscious AI systems that will replace jobs and entire industries are very much a risk worth talking about). 

While consciousness is not necessarily substrate-dependent (it may not require biological material), the process for duplicating it in machines appears unlikely. Consciousness seems inextricably linked to the complexities of the biological world. As Seth wrote:

“All of our perceptions and experiences, whether of the self or of the world, are inside-out controlled and controlling hallucinations that are rooted in the flesh-and-blood predictive machinery that evolved, develops, and operates from moment to moment always in light of a fundamental biological drive to stay alive.” 

Good luck creating this in a machine. 

Overall, I found myself agreeing with much of the content of the book, with the possible exception of the chapter on free will (although I make no claims to have the answers). As with most practicing scientists, Seth dismisses the idea of free will and adopts a physicalist position, without realizing that, if he’s correct, and consciousness arises out of physical processes, then consciousness loses its evolutionary function and ultimately seems to serve no purpose at all. 

The reason is that physicalists are denying reverse causation from consciousness to physical processes (even if they don’t realize it). If everything arises from the movement of atoms, which arrange themselves into chemical compounds and eventually into bodies and brains according to physical laws, and then consciousness arises out of the activity of brains, then consciousness is fully determined by the underlying physical phenomena. Consciousness, since it is dependent on atomic motion as governed by physical laws, therefore cannot change those laws or the underlying atomic motion. If consciousness cannot change physical states—allowing the agent to act otherwise than it did—then in what sense can consciousness have a purpose? In what sense is the illusion of free will useful if the universe can only turn out to be one way (as determinism states)?

These are questions that are only poorly addressed in the book, if they are addressed at all. But we can give Seth a break, since these are deep philosophical questions that, if we’re honest, no one really has a handle on. And that’s the point: Seth is not claiming to be able to solve these harder philosophical questions. And in fact, he doesn’t need to; the science of consciousness can be pursued simply by sticking to the measurement of the neurological correlates of conscious experience. It is here, and only here, that we can make progress. But we must do so with humility and caution. 

Further Reading

Check out the following books to learn more about consciousness, neuroscience, intelligence, and human thought. 

2 thoughts on “Why Your Conscious Experience is Nothing More Than a Controlled Hallucination

  1. Hi Ryan, great review! I have two things I’d like your thoughts on.

    1) In the Ordinary Language Philosophy tradition, there’s a line of argument about hallucinations, illusions, etc., that runs something like this ‘the existence of mistake presupposes some background truth(s) in order to establish said mistake, therefore any conclusion that our entire experience is an illusion is self-defeating’. I realize that’s not quite a formal argument, but I wonder whether views like Anils’ might run afoul of something like it. I understand that our visual systems have particular quirks that can lead to various illusions, and I think it makes sense to talk about prediction in our perceptions, but I have trouble going from that to ‘our perceptions are full blown collective hallucinations and we don’t have access to reality.’ Even if we agree that Anil is in some sense correct about perception, we wouldn’t want to say that our understanding of physics (or sub in any well established area of science) is based on a hallucination and potentially has no relation to reality as it is? I realize the above is a bit of a tangle but I get the sense that Anil’s thesis can be both relatively mundane or taken another way, extremely radical (a form of idealism based on empirical findings, essentially), and I’m not sure how to reconcile the more radical version with what we take ourselves to know about the world.

    2) You write: “As Seth correctly points out, there is no necessary correlation between intelligence and consciousness; they are separate phenomena. Relatively unintelligent animals probably have inner experiences and can feel pain, whereas highly intelligent machines playing games of chess almost certainly are not experiencing any sense of joy or accomplishment from winning a match.”

    I wonder about divorcing intelligence from awareness, or more broadly, from the phenomena of life. Clearly there are programs that beat us at chess and Go, but these systems aren’t autonomous, they don’t choose their goals and pursue them, get frustrated, etc. If we agree that in principle, the difference between a calculator and a program that plays Go is only of degree, then it seems strange that we call the program that excels at Go highly intelligent (in that domain) but we wouldn’t call a calculator intelligent. If we do call a calculator intelligent, then should we not also call an abacus intelligent (very minimally, but still)?

    Liked by 1 person

    1. Thanks for the reply Marc, and great questions.

      1) It is true that an illusion or hallucination can only be defined in relation to what is real. If there is no reference point, how can we claim that someone is hallucinating? Here is how I think we can defend Anil’s claims.

      We know that color, for example, is not a property of objects (how could it be when people can perceive the same object as different colors?). If objects do not have color, yet we perceive them as having color, then this qualifies as a hallucination because our experience of color can be compared to the reference point of the underlying physics which describes an external world that is devoid of this subjective experience. It’s true that this reference point, reality “in itself,” is not directly accessible to us, but, as with the color example above, we can deduce that it differs from our experience and this is enough to label consciousness as a hallucination, albeit a “controlled” hallucination in the sense that our subjective experiences are corrected by sensory input from the environment.

      2) If we define intelligence as the ability to acquire and apply knowledge, then you’re right, in a sense the calculator can be called an intelligent machine. Intelligence is a matter of degree; the calculator contains only minimal intelligence whereas the program playing chess has a higher degree of intelligence. As for the abacus, this is a harder case. Instead of calling it minimally intelligent, I’d consider it an extension of our own intelligence, as we use it as a tool to acquire and apply our own mathematical knowledge (versus the computer program and calculator that are performing their own functions semi-independent of our actions).

      Anil’s point is that, while there are different degrees of intelligence, there is no “threshold” level of intelligence where, once you cross it, you automatically get consciousness. Consciousness seems to be its own phenomenon, also represented in degrees, but not dependent on a certain threshold of intelligence. It is instead dependent on the underlying complexities of biology.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s