Opening — What It Means to See Red
Take a moment and find something red nearby. An apple, a sign, someone's shirt, anything will do. In the instant you look at that red, what is happening inside your head?
Neuroscience can answer in some detail. Light stimulates the cone cells of your retina, a signal travels along the optic nerve to the visual cortex at the back of your brain, and populations of neurons tuned to certain wavelengths fire. Up to here, everything is a measurable physical event.
Yet something is left over. While all that electrochemical signaling goes on, there is for you a particular feeling of red. Not merely the information that light of seven hundred nanometers has been detected, but the vivid texture of that red rising up in your mind. Why does this feeling exist? Neurons fire, but why does it not simply end as a calculation in the dark? Why does it spill over into an experience of feeling something?
This is exactly what the philosopher David Chalmers named, in 1995, the hard problem of consciousness. In this essay we follow that fascinating and maddening question. I will not hand you an answer, because no one yet has one. What I will do is look with you at why this is regarded as one of the deepest puzzles humanity has faced, and at the views that compete to address it.
An Old Puzzle — A History of the Mind-Body Problem
The name hard problem was coined in 1995, but its roots run far deeper. The question of how mind and body are related, the so-called mind-body problem, sat at the very starting point of modern philosophy.
In the seventeenth century the French philosopher Rene Descartes gave it its classic form. He divided the world into two kinds of substance. One is matter, which occupies space, the extended substance. The other is the thinking mind, the thinking substance. The body works like a machine, but the mind is an immaterial substance of an entirely different sort. This is substance dualism.
Descartes two substances
- Matter (extended substance): occupies space, measurable.
- Mind (thinking substance): thinks, doubts, feels.
- The two substances are utterly different in nature.
A troubling question follows at once. How can two substances of utterly different nature affect one another? When I decide in my mind to raise my hand, my physical arm actually rises. How can an immaterial mind push a material body into motion? This is the interaction problem.
The one who pressed this weakness most sharply was Princess Elisabeth of Bohemia. In her correspondence with Descartes she asked, persistently, by what means a mind that has no extension and makes no contact could ever cause motion in a body. Descartes never produced a satisfying answer. The question follows dualism as its heaviest burden to this day.
There is another scene worth savoring. The eighteenth century philosopher Gottfried Leibniz left us the mill thought experiment. Imagine a machine that perceives and thinks, large enough that we could walk inside it like a giant. What we would find inside, he argued, are only parts pushing and meshing against one another, and nowhere among them would we find perception itself.
Leibniz's mill
- Imagine walking inside a thinking machine.
- All you see are parts working upon one another.
- Nowhere is "perception" to be found.
- So can perception be explained by the arrangement of parts alone?
The scene is strikingly modern. Read parts as neurons, and you see that Leibniz, three centuries ago, was already feeling his way toward the intuition behind the hard problem.
The Easy Problems and the Hard Problem
Chalmers insight lay in splitting questions about consciousness into two kinds.
First there are what he called the easy problems. The name does not mean they are actually easy to solve. They are difficult scientific tasks that may take decades. They are called easy only in the sense that, in principle, they can be approached with standard scientific methods.
Examples of easy problems
- How does the brain integrate sensory information
- How does attention focus on a particular object
- What is the difference between being awake and asleep
- How do we control and report our behavior
- How is information stored as memory
These are all questions about function. They ask what mechanism performs what task. And function can be measured, modeled, and simulated.
The hard problem is different. It is the question that stubbornly remains even after every function has been perfectly explained.
> Why is all this information processing accompanied by subjective experience? Why does the processing not proceed in the dark, instead of there being someone inside who feels something?
In the words of the philosopher Thomas Nagel, to be conscious means that there is something it is like to be that creature. There is presumably nothing it is like to be a stone. But there is surely something it is like to be you. Philosophy calls the whole of that something conscious experience.
Qualia — The Vivid Texture of Experience
At the core of the hard problem sits the notion of qualia (the singular is quale). Qualia are the subjective, qualitative characters of experience. The redness of red, the particular quality of the smell of coffee, the sharp pain of a cut on a fingertip.
Let us see why qualia are a problem through a thought experiment. It is the story of Mary's Room, proposed by the philosopher Frank Jackson.
The Mary's Room thought experiment
- Mary is a brilliant scientist.
- She has lived her whole life in a black and white room.
- She knows every physical fact about color.
Wavelengths, retinal action, brain processing, all of it.
- One day Mary leaves the room and sees a red rose for the first time.
Question: does Mary learn something new?
Intuitively, many people answer yes, Mary finally comes to know what red feels like. But Mary already knew every physical fact about color. If so, what she newly learned is something that is not a physical fact, namely the texture of the experience itself.
If this argument is right, the conclusion follows that there is something in the world not captured by physical facts alone. Of course there are several objections to this thought experiment. The best known reply is that Mary gained not a new fact but a new ability, the ability to imagine and recognize red. Which side is correct is still debated. What matters is that this simple story makes the gap between the physical and the experiential vivid.
Looked at more closely, the Mary argument has a set of standard replies that have grown quite refined. Let me lay them out fairly.
The first is the ability hypothesis. What Mary gains on leaving the room is not a new fact but a new ability, the know-how to imagine, recall, and recognize red. Just as knowing how to ride a bicycle is not propositional knowledge, knowing red is not the addition of a fact but the acquisition of a skill.
The second is the old-fact-new-mode reply. The feeling of red Mary comes to know is not a new fact she lacked before, but a physical fact she already knew, now met in a different mode, the first person direct acquaintance. It resembles knowing a person only from photographs and then meeting them in the flesh. You have not learned of a new person, only known the same person in a new way.
The third is Dennett's forceful pushback. With variants like RoboMary he argues that if Mary truly knew every physical fact without exception, she would not be surprised on first seeing color. The reason we intuit that she would be surprised is that our imagination fails to genuinely uphold the premise that she really knows everything.
We cannot conclude that these replies have toppled the Mary argument. Those who favor dualism hold that the force of the intuition survives. But the outcome of this debate is still open, and neither side has fully persuaded the other.
What Is It Like to Be a Bat
In 1974 Thomas Nagel published a short paper titled What is it like to be a bat. It is one of the most cited papers in the history of philosophy.
The bat uses echolocation. It emits ultrasound and perceives the world through the returning echoes. Just as we have vision, the bat has a world woven out of sound waves.
Nagel's question is this. When a bat experiences the world through echolocation, what is that experience like?
Nagel's point
- A bat's brain, behavior, and physiology can be studied objectively.
- But how the world feels from the bat's own standpoint
can never be known from the outside.
- Even imagining that I behave like a bat is only
what it would be like for me to mimic a bat,
not what it is like for the bat itself.
The core of what Nagel highlighted is that consciousness is essentially subjective. Experience always comes attached to a particular standpoint that undergoes it. Yet science pursues objectivity. It deals with facts that yield the same result no matter who measures them. Here a tension arises. How can something inherently first person be fully captured in third person language?
This is why the hard problem is suspected of being not merely unsolved but perhaps difficult in principle.
The Explanatory Gap — A Foretaste Before Chalmers
In 1983, twelve years before Chalmers named the hard problem, the philosopher Joseph Levine introduced the phrase the explanatory gap. The two notions are close but not identical, so it helps to keep them apart.
Levine offers a commonly cited identity statement, the sentence pain is the firing of C fibers. Even if this were factually true, we still cannot at all understand why that firing should be precisely that hurting feeling rather than some other feeling, or none. The identity water is H2O carries no such frustration, because the properties of water follow smoothly from the structure of H2O. In the case of consciousness, however, the path from physical fact to the texture of experience leaves a gap that explanation does not fill.
The explanatory gap and the hard problem
- Levine (1983): physical facts do not explain why
that feeling follows them -> the explanatory gap.
- Chalmers (1995): the gap may be not a mere limit of cognition
but a metaphysical question about the structure of the world.
- The difference: Levine points first at "we fail to understand,"
Chalmers at "physicalism may be insufficient."
In short, Levine's gap is closer to an epistemic remark about the limits of our understanding, while Chalmers' hard problem goes a step further into a metaphysical challenge, asking whether the world is made of the physical alone. Yet both questions point at the same place. However high you stack physical explanations, why feeling should inhabit them does not follow on its own.
Philosophical Zombies — The Person with the Lights Off
Another device Chalmers used to dramatize the hard problem is the philosophical zombie. It is nothing like the zombie of the movies.
A philosophical zombie is a copy of you, identical down to the molecule. Same brain, same behavior, same words. No one could tell the two of you apart. Except for one thing. There is no one inside. The zombie has no subjective experience at all. It sees red but there is no feeling of red, it cuts its hand but there is no feeling of pain. It merely says ouch and pulls its hand away, while inside nothing happens. The lights are off.
The argument runs like this.
Structure of the zombie argument
1. We can imagine a philosophical zombie without contradiction.
2. If it is imaginable, such a being is at least logically possible.
3. Then even if every physical fact is the same,
the presence or absence of consciousness can differ.
4. Therefore consciousness is not explained by physical facts alone.
The argument is powerful but its weaknesses are clear too. The biggest point of contention is step one. Can we really imagine a zombie consistently? Some philosophers reply that we only feel we can imagine it, while in fact we have failed to see a hidden contradiction, just as we may fool ourselves into thinking we can imagine a round square.
The zombie argument does not prove its conclusion. But the question it throws out is alive. Whether it is really contradictory for behavior and function to be wholly the same while experience differs, we are not yet sure.
Going deeper, at the heart of the argument lies a step that crosses from conceivability to possibility. Chalmers holds that if a zombie can be conceived consistently, it is metaphysically possible, and if it is possible, consciousness goes beyond physical facts. The place critics most often aim at is exactly this bridge. Conceivability does not by itself guarantee metaphysical possibility.
From conceivable to possible — is the bridge sound
- Chalmers: we can conceive a zombie without contradiction
-> the zombie is metaphysically possible.
- Critique: feeling able to conceive something
is not the same as its being possible.
- Example: "Hesperus is not Phosphorus" may seem conceivable,
yet both are the planet Venus, so it is in fact impossible.
On this objection, even the moment we take ourselves to be conceiving a zombie, we may merely be sketching a surface picture while failing to grasp a deep link between consciousness and physical process. Of course the Chalmers camp does not sit still. They reply that in the case of consciousness there is no room for such a hidden identity to slip in. This exchange remains one of the most finely honed debates in contemporary philosophy of mind.
The Meta-Problem — Why Do We Think There Is a Hard Problem at All
Recently Chalmers himself raised another intriguing question, the meta-problem of consciousness, which he posed in 2018. The question is this. Why do we feel, in the first place, that there is a hard problem of consciousness, and why do we say so and write about it?
The charm of the meta-problem is that it belongs to the easy problems. Our very behavior of reporting that there is a hard problem is a physical process, so in principle science can explain it.
The meta-problem's two-edged import
- If we could fully explain the mechanism by which
we come to talk of a hard problem,
- illusionist side: that explanation might itself reveal
the hard problem to be an illusion.
- realist side: explaining the mechanism still leaves
the mystery of feeling itself intact.
The meta-problem does not raise the hand of either camp. It is closer to a test that both must face seriously. For the illusionist it is a chance to vindicate the position, and for the realist it is a task to show that a residue remains which mechanism alone cannot fill. Either way, asking why we feel such a puzzle at all is a valuable point of departure in itself.
The Competing Answers — Who Is Right Remains Unknown
On the hard problem, philosophers and scientists divide into several camps. Since none has been proven correct, let me introduce the main camps fairly.
Physicalism — In the End It Is All Matter
Physicalism holds that what exists in the world is ultimately only the physical, and consciousness is no exception. The mind is what the brain does, and once a sufficiently refined neuroscience is complete, consciousness too will be explained.
Some physicalists go a step further and argue that the hard problem itself is an illusion. The philosopher Daniel Dennett is the leading example. He sees what we call qualia as something closer to a kind of user illusion produced by the brain's information processing, much as the folder icon on a computer screen is not an actual folder but a representation made for convenience.
Dualism — Matter Alone Is Not Enough
Dualism holds that mind and matter are fundamentally different kinds of thing. Descartes seventeenth century mind body dualism is the classic. Chalmers himself takes a position he calls naturalistic dualism, which does not reduce consciousness to matter yet does not treat it as supernatural either. Instead he proposes that we accept consciousness as one of the basic elements of the universe, like mass or charge.
The old difficulty for dualism is interaction. How can an immaterial mind move a material brain? It is not easy to answer this question satisfyingly.
Panpsychism — A Little Consciousness Everywhere
Panpsychism is the most provocative position. It holds that a very basic form of consciousness already resides at the fundamental level of matter. Even a single electron has something extremely primitive, and these combine in complex ways to form the rich consciousness of a human being.
It may sound absurd, but the number of philosophers taking it seriously is growing. Galen Strawson and Philip Goff are leading examples. Their attraction to panpsychism rests on a kind of elimination. If it is baffling to say consciousness emerges as if by magic from matter at some moment, and equally hard to accept an illusionism that denies consciousness wholesale, then the remaining path is to place consciousness from the outset as a basic property of matter. Its appeal lies in simplicity. Instead of explaining how consciousness suddenly springs from nothing at some moment, it holds that consciousness was there from the start.
But critics hold their ground for good reason. The largest stumbling block is the combination problem. How do the tiny fragments of experience residing in each electron combine into the single, unified stream of consciousness you enjoy right now?
The combination problem
- Premise: there is micro-experience at the particle level.
- Question: how do micro-experiences combine into
one unified experience of an "I"?
- That a simple sum of small subjects yields a larger subject
is not at all obvious.
Critics also note that panpsychism struggles to offer testable predictions, and that ascribing consciousness everywhere risks being naming rather than explaining. Defenders reply that the combination problem is a homework assignment rather than a fatal blow, and they look for points of contact with scientific frameworks such as integrated information theory. This debate, too, has reached no verdict.
The table below roughly compares the three positions.
| Position | Nature of consciousness | Strength | Main difficulty |
| --- | --- | --- | --- |
| Physicalism | A product of brain processes | Coheres with science | Falls short on the texture of experience |
| Dualism | A substance apart from matter | Honors the specialness of experience | Mind body interaction |
| Panpsychism | A basic property inherent in matter | Avoids the leap of emergence | Combination of the fragments |
Higher-Order Theories and Illusionism — Another Live Option
Within physicalism there are many branches. Among them, higher-order theories hold that a mental state becomes conscious when it is accompanied by a further, higher-order representation of that state. There is a first-order state of perceiving red, and only when a higher-order state is added, the awareness that I am now seeing red, does conscious experience come about. Rather than charging head-on at the hard problem, this view redescribes consciousness as a relation among representations in order to naturalize it.
Illusionism is more radical still. As we saw with Dennett, the irreducible texture we call qualia is in fact only a powerful impression the brain forms about itself, not something that exists at face value. On this view the hard problem changes from a real problem to be solved into a problem of explaining why we are under the illusion that there is one.
Of course the criticism is no less forceful. The most common rejoinder is that while we may doubt everything else, the fact that I now feel something is as clear as anything could be, and to call that clear feeling an illusion is closer to denial than explanation. None of these branches has hardened into settled doctrine, but they are discussed together as serious options for treating the hard problem.
How Far Has Science Come
Apart from the philosophical debate, science is diligently searching for the neural correlates of consciousness (NCC). This is research that measures which brain activity minimally appears together with conscious experience.
The figures who truly launched this research program were Francis Crick, co-discoverer of the DNA double helix, and the neuroscientist Christof Koch. In the 1990s they proposed that, instead of leaving consciousness as a vague and vast metaphysical riddle, we pinpoint concretely the minimal neural mechanisms that correspond to a particular conscious experience. Experiments such as binocular rivalry, where the same picture is at one moment consciously seen and at another not, made a good starting point.
The idea behind NCC research
- The same stimulus is at times conscious, at times not.
- Compare the brain activity in the two cases.
- That difference becomes the candidate most tightly tied to consciousness.
- But finding a correlate is not yet explaining its reason.
Let me introduce two representative theories. Neither is settled doctrine, but both are being actively tested.
Global Workspace Theory likens consciousness to a kind of broadcast. Among the brain's many unconscious processes, when some information rises into a global workspace and is shared across regions, it is consciously experienced. It resembles the way only the part lit by the spotlight on a theater stage is seen by the whole audience.
Integrated Information Theory tries to define consciousness by the amount of information a system integrates. It marks that amount with a value called phi, and the more tightly information is woven together as a whole rather than reducible to parts, the larger the phi and the higher the degree of consciousness. It drew great attention as an ambitious attempt to quantify consciousness mathematically.
Yet the criticism is considerable. Some point out that phi is in practice all but intractable to compute for a real brain. Others worry that, taken at face value, the theory yields the counterintuitive conclusion that even a simple but suitably connected grid of circuits would harbor substantial consciousness. In 2023 a group of researchers publicly criticized the theory as close to pseudoscience, and the dispute spread widely. Supporters pushed back that the criticism was overblown. The point is that the theory is an interesting candidate, but by no means an agreed-upon answer.
That such research is highly valuable is clear. One thing should be noted, however. These theories mostly address under what conditions consciousness appears. They do not yet directly answer the hard problem, namely why feeling arises under those conditions. Finding a correlation and explaining its reason are matters of a different order.
Can AI Be Conscious
This question no longer belongs only to science fiction. As conversational AI becomes part of everyday life, many have begun to ask. Is there someone inside these systems too, feeling something?
First a point to make clear. The fact that today's AI appears to speak fluently and express emotion is not, in itself, evidence of consciousness. Recall the philosophical zombie. From behavior alone you cannot tell whether the lights are on inside. Even if an AI says it is lonely, behavior alone cannot distinguish whether it feels loneliness or was trained to describe loneliness.
Here views diverge.
Two intuitions about AI consciousness
On the affirmative side:
- If consciousness is the result of a certain information processing,
there is no reason silicon cannot carry out that processing.
- Carbon versus silicon may not be the essence.
On the cautious side:
- Consciousness may depend on specific physical properties
of the biological brain.
- Mimicking a computation and actually experiencing
may be different things.
What is interesting is that, so long as the hard problem is unsolved, we cannot in principle judge whether an AI is conscious. We believe other people are conscious. Because they look like us, have brains like ours, and behave like us. It is a kind of analogy. But AI has a structure utterly different from ours. Whether this analogy holds, we do not know.
It helps to draw one distinction clearly here. Intelligence or behavior and experience are not the same thing. That a system solves problems with astonishing cleverness and converses like a person is logically separate from there being feeling inside it. Conversely, the possibility that a very simple-seeming creature has rich experience cannot be ruled out. Degree of capability and presence of experience are different axes.
Another thing that comes to mind is the old problem of other minds. In truth we have never directly looked in and confirmed that other people are even conscious. We merely believe it by a strong analogy. Apply this skepticism to machines and matters grow harder still. A machine shares with us neither body nor evolutionary history nor inner structure, so the foothold of analogy on which we lean for people becomes far weaker.
Why behavioral tests cannot decide
- A system may behave identically whether or not it is conscious,
- and behave identically whether or not it lacks consciousness.
- So matching behavior alone cannot tell experience apart.
- Whether a better test could exist is itself an open question.
So a cautious attitude is needed. To declare hastily that AI is surely conscious, and to insist that a machine could not possibly be conscious, are both claims that go beyond what we know. The honest answer is that we do not yet know. And this not-knowing is no mark of laziness, but an honest acknowledgment of the depth of the problem.
Theories at a Glance
Let me gather into one table what the positions we have met say about the hard problem, qualia, and zombies. It is a rough summary, so please consult the body for the finer differences.
| Position | View of the hard problem | Stance on qualia | Are zombies possible |
| --- | --- | --- | --- |
| Physicalism (reductive) | A bundle of easy problems to be solved | Reduced to brain processes | Impossible |
| Illusionism | Not a real problem but a kind of illusion | Do not exist at face value | A meaningless supposition |
| Naturalistic dualism | A real and deep problem | An irreducible reality | Possible |
| Panpsychism | Real, but eased by relocation | A basic property of matter | Mostly impossible |
A Pause to Think — Mini Quiz
A few short questions to check what you have read. Try answering in your head, then compare with the solutions below.
Question 1. What is the decisive difference between the easy problems and the hard problem as Chalmers describes them?
Question 2. What is the core of what the Mary's Room thought experiment tries to show?
Question 3. Why is it hard to take an AI saying it is sad as evidence of consciousness?
Question 4. What was the core objection Princess Elisabeth pressed on Descartes?
Question 5. How does Levine's explanatory gap differ from Chalmers' hard problem?
Question 6. What is the combination problem that panpsychism faces?
Solution 1. The easy problems concern function and mechanism and can in principle be approached by science. The hard problem is the question that remains even after all those functions are explained, namely why subjective feeling accompanies them.
Solution 2. It shows that even knowing every physical fact, the texture of experience may be something learned separately, that is, the possibility of a gap between physical facts and experience. There are, however, objections to this.
Solution 3. Because behavior and speech can appear the same whether or not there is experience inside. As the zombie argument shows, outward behavior alone cannot decide the presence or absence of subjective experience.
Solution 4. It was the interaction problem, the question of how an immaterial mind with no extension and no contact could cause motion in a material body. Descartes never produced a satisfying answer.
Solution 5. Levine's gap is primarily an epistemic remark that we fail to understand why that feeling follows, while Chalmers' hard problem goes a step further into a metaphysical challenge that physicalism may be insufficient.
Solution 6. It is the difficulty of explaining how the tiny fragments of experience at the particle level combine into a single, unified consciousness. It is reckoned the largest homework panpsychism must do.
Closing — Embracing Not Knowing
The hard problem of consciousness is not yet solved. Perhaps with the conceptual frame we possess it can never be solved. Some philosophers go so far as to say that human cognition has a built in limit, so that this question alone may forever lie beyond our reach. Others are optimistic that one day science will explain consciousness cleanly.
One thing, though, is clear. This question is the puzzle closest to each of us. It is not a distant black hole or the origin of the universe. It is your very experience, feeling something right now as you read these words, that is the puzzle. That the thing we know most surely is at the same time the thing hardest to explain is a strange and beautiful paradox.
I leave a few things to think about.
- If you replaced your brain neuron by neuron with artificial chips, would consciousness vanish at some moment, or would it remain intact?
- When you fall into deep sleep, does consciousness disappear, or is it merely not remembered?
- If in the future an AI pleads that it has experience, on what grounds should we believe or doubt it?
There is no correct answer. But to live carrying such questions may itself be a privilege granted only to a being that is conscious.
References
- Stanford Encyclopedia of Philosophy, "Consciousness": https://plato.stanford.edu/entries/consciousness/
- Stanford Encyclopedia of Philosophy, "Qualia": https://plato.stanford.edu/entries/qualia/
- Stanford Encyclopedia of Philosophy, "Zombies": https://plato.stanford.edu/entries/zombies/
- Stanford Encyclopedia of Philosophy, "Dualism": https://plato.stanford.edu/entries/dualism/
- Stanford Encyclopedia of Philosophy, "Panpsychism": https://plato.stanford.edu/entries/panpsychism/
- Encyclopaedia Britannica, "Philosophy of mind": https://www.britannica.com/topic/philosophy-of-mind
- Nature, "What is consciousness?" (collection): https://www.nature.com/subjects/consciousness
현재 단락 (1/190)
Take a moment and find something red nearby. An apple, a sign, someone's shirt, anything will do. In...