How does the human mind work? A map of the mind

What is this?

The picture above is a map of the human mind. It shows the main parts of the mind and how they work together. It brings together ideas from old philosophers, like Immanuel Kant and Sigmund Freud, and from modern brain science.

Many people think the mind works like a factory line: first you see something, then you think about it, then you act. Our model says this is too simple. The parts of the mind talk to each other all the time, in loops. That is why the picture shows a network, not a line.

Our model talks about three different things, and we try not to mix them up: what the world is like, how the mind is built, and what people experience. We also say clearly when we only describe something, when we explain it, and when we predict something that can be tested.

The world is real

Our model starts with one basic idea. We call it Structured Realism:

• The world exists, even when nobody looks at it.
• Our mind shapes how we experience the world. But it does not create the world.
• Our mind fits the world because it grew in the world, through evolution.

How do we know something is real? We have two answers. First: real is what pushes back. If you walk into a wall, the wall stops you, whether you believe in it or not. Second: real is what stays the same for very different minds.

Here is an example. Long ago, in the time of the dinosaurs, a tree falls in a forest. A dinosaur, a small early mammal and a bird all hear it. Their ears are very different, so the sound is different for each of them. But all three notice the same place, the same moment and the same order of events. The sound belongs to the animal. The wave in the air belongs to the world.

This also answers an old question: does a falling tree make a sound if nobody hears it? Our answer: it makes a wave in the air, yes. But a sound needs someone who hears it. Philosophers call such experienced qualities – the sound itself, the red of red, the hurt of pain – qualia. They belong to the one who hears or sees, while the wave belongs to the world.

Nine parts of the mind

The model has nine parts. We call them domains:

• Perception – taking in the world and your own body through the senses.
• Memory – storing and remembering.
• Understanding – language, logic and concepts: knowing words, ideas and rules, thinking logically.
• Judgment – using what you know on the case in front of you: picking what fits, and seeing how cases are alike.
• Reason – thinking about your own thinking (metacognition), about right and wrong, and about the meaning of life.
• Prospection – imagining what could happen.
• Social cognition – understanding what other people think and want.
• Emotions – feeling, liking and disliking, wanting.
• Will – deciding and controlling yourself.

Things like your conscience, your sense of self or a gut feeling are not extra parts. They happen when several parts work together.

Where is language? We use language for almost everything: we remember, plan and judge in words. We see it mainly as part of understanding – that is why the picture shows it there – and speaking is a kind of action. Two brain areas are famous for language: Broca's area, at the front of the left side of the brain, helps us speak and build sentences; Wernicke's area, further back, helps us understand. But one fact speaks against putting language inside understanding: some people lose their language after a stroke and can still think and plan. So this question is still open.

Thinking about thinking. One ability deserves special mention: metacognition. It means knowing how sure you are, noticing your own mistakes, or knowing that you know a name even when you cannot find it right now. It belongs to reason, and it is what makes the mind a system that turns back on itself. In the picture, it is the first thing written under reason.

The mind always lives in a body – in humans, mostly in the brain. The body is not one more part of the mind. It is what carries all the parts. In the picture, it is the brown band at the very bottom: being awake and being calm or excited come from the body, and they affect all parts at once. The body also adapts to stress over a long time; scientists call this allostasis.

Drives. Something has to get the mind going. That is what drives do. We sort them into four groups: body (eating, drinking, staying warm, staying alive), reproduction (sex, caring for children), social (belonging, rank, territory) and knowledge (exploring, playing, and maybe a need for meaning). Drives are not a tenth part of the mind. They are sources that push many parts at once: curiosity pushes perception and imagination, the wish to belong pushes social cognition, hunger pushes feelings and the will. In the picture, they are the pink band under the field, glowing upwards. Wanting and liking are different: in experiments, rats without dopamine still like sugar, but no longer try to get it. Wanting is the drive; liking is the feeling.

Only humans seem to have a need for meaning – asking what everything is for, or what comes after death. Religions exist in every known culture. Whether this is a drive of its own, we do not know; we treat it as an idea to test. AI language models have no drives in this sense: they have no body that needs anything, and their goals come from people.

Seven axes

Every part of the mind can work in different ways. We describe this with seven axes. You can think of each axis as a scale that goes from one end to the other:

• Awareness: unconscious – almost conscious – fully conscious
• Control: automatic – controlled – carefully thought through
• Time: past – present – future
• Point of view: me – others – the group
• Feeling tone: bad – neutral – good
• Real or not: just possible – real – certain
• Who did it: me – someone else – no one

The real or not axis is about how real something seems to the mind. Sometimes the mind gets this wrong. In a dream, your mind makes everything up, but it feels real. An AI sometimes says made-up things as if they were facts. In both cases, something made inside is taken as real. It can also go the other way: some people who lose their sight see clear pictures, but they know the pictures are not real. The picture is there, but the mind marks it “not real”. This shows that the axis works on its own.

The who did it axis is about who seems to have caused a thought or an action. In a rare condition, one hand moves on its own, and the person feels: “That wasn't me.” Some people with psychosis feel that someone else has put thoughts into their head.

The nine domains say what the mind does. The seven axes say how it does it right now. These seven axes are a proposal. There may be more.

The parts and the axes are two different kinds of thing. The parts are jobs the brain does, and a brain injury can knock one of them out. The axes are not places in the brain. They are more like basic categories, such as time. Their settings are changed by chemical messengers that reach the whole brain at once – for example during sleep, under drugs or under stress. So an axis is never knocked out. It only moves. In the picture, six axes are shown as sliders at the top. Awareness is not a slider: it is the depth of the picture, from the parts at the top down to the unconscious at the bottom.

Four sides of consciousness

The word “consciousness” is used for different things. We split it into four:

• Being awake. This is a state of the whole body and brain, like being asleep or under anesthesia. It is not one of the axes.
• Awareness. Is a thought or a picture available, so that you can talk about it and use it? This is the first axis. In the picture, it is shown as depth.
• Knowing that you are you. This comes from two parts working together: the “me” end of the point-of-view axis, and reason, which can think about yourself.
• Experience. Why does it feel like something to be you? This is the big open question. Our model does not answer it.

In the picture, the unconscious is shown at the bottom, like deep water under all the parts. Just above it is a middle layer: the preconscious. It works like a gatekeeper: it decides what becomes conscious – for example when your name suddenly catches your attention in a noisy room. In the brain, this job is done by a group of brain areas that notices what is important right now.

A circle with the world

The mind does not just wait for information. It always guesses what will happen next and checks the guess. Then you act: your will decides, your muscles move, your action changes the world, and the world sends new information back. It is a circle: world → mind → action → muscles → world.

When the brain is damaged

Brain injuries show which parts of the mind can break on their own. We looked at four examples. You can try them with the lens next to the picture.

• Split brain. When the two halves of the brain are cut apart, the talking half often invents reasons for things the other half did. So our explanations for our own actions are sometimes made up.
• Neglect. After some strokes, people ignore everything on their left, even though their eyes work. They even forget the left side of places they imagine.
• Agnosia. Some people can see an object but cannot say what it is. That is why, in our model, judgment does not get information straight from the eyes. It goes through memory and understanding first.
• Parkinson's disease. Automatic movements like walking become hard. But many people can walk better when they step over lines on the floor on purpose.

Animals show the opposite problem. An elephant that was kept in chains for years may keep swaying from side to side, even years later when it is free. A movement runs on automatically, and the will cannot stop it anymore. Parkinson's disease and this swaying are the two ends of the same axis: in one, the automatic part is missing; in the other, it cannot be stopped.

Character

We say that people have a character: someone is anxious, careful or suspicious. In our model, a mood is a short move on the axes, and a character is where the axes usually rest. An anxious person's feeling tone rests on the bad side. A careful person rests on “controlled” and thinks about the future. A suspicious person quickly feels that someone else is behind things. With pets, the model tells us which words fit. A dog can show something like jealousy, because it has feelings and some social sense. But the famous “guilty look” is not guilt: tests show the dog reacts to being scolded, not to what it did. Guilt needs reason, and dogs do not have that part.

Animals and AI

We compared the model with animals. Instead of one ladder from “simple” to “clever”, we look at four things: which parts of the mind an animal has, how well they are connected, how much it can think about its own thinking, and how far its mind reaches – for example, how far into the future it can plan.

A coral has no brain at all, only a net of nerves. It simply reacts and does not learn much, so it is the zero point of the scale. A fly already learns and predicts its own movements a moment ahead, but it has only a few parts of the mind, in a simple chain. Fish feel more than a simple reflex when they are hurt. Humans have all parts, in many loops. Ravens, octopuses and great apes are in between. Ravens and octopuses show that the same abilities can grow in very different brains. An octopus has many abilities, but much of its control sits in its arms, not in one central brain.

AI language models are the other way around: evolution built from the bottom up, but AIs start at the top. They are strong at language and thinking, but they have no body and no feelings in the usual sense. Their weak spot is the real-or-not axis: sometimes they say made-up things as if they were facts.

We see three steps for AI: a language model that only talks; an agent that uses tools and can fail – for example when a computer program it wrote does not work; and a robot with sensors and a body. We expect fewer made-up facts at each step, because the real world pushes back more often. Whether an AI experiences anything, nobody can say today.

How can we test it?

A good theory must say what would prove it wrong. Our model makes six predictions. For each one, we also say what other explanation could fit the same facts. For example:

• Parts of the mind that are closely linked should break down together. For example, many people who lose their memory also cannot imagine new things. But not all of them – so this is still open.
• Feeling with others and understanding others are two different things. One can be damaged while the other still works. This has been found.

We also suggest a test with the “wiring diagram” of the brain: parts that are linked in our model should also be closely wired in the brain.

Strong objections

We asked: what would other thinkers say against our model? Here are the four hardest objections, in short:

• “You cannot step outside your own mind.” True. But we can compare very different minds. What stays the same for all of them is probably part of the world.
• “Space and time are not real. They are only a screen made by evolution – or we live in a computer simulation.” The first idea defeats itself: the computer models behind it use animals that live, compete and have children over many generations – in a real world with space and time. The simulation idea has four problems. First, a simulation needs a real world with computers in the first place – a world with a long, lucky history. If no asteroid had hit the Earth 66 million years ago, the dinosaurs would still rule, and there would be no humans and no computers. So the real world is still there, just one floor up. Second, the idea can explain away any evidence: “that is simulated too.” Some old ideas work the same way – that the world was made five minutes ago, together with all our memories, or that fossils were put into the ground when the world was made. An idea that nothing can ever push back against is not a claim about the world. And even quantum physics, which says atoms are mostly empty, does not make the wall less hard. It explains why it is hard. Third, the simulation argument trusts our physics to guess how much computers can do – and at the same time says that this physics is probably fake. Fourth: even if we did live in a simulation, its walls would still stop us, and what stays the same for very different minds would still stay the same. For our model, a simulated world would be just as real. Only the “real basement” would be somewhere else.
• “The mind is not a box in the head. It is a living body acting in the world.” We agree that the body matters. That is why the circle with the world and the muscles are part of the model.
• “One simple rule – the brain predicts – explains everything.” That rule is important, and our model uses it. But brain injuries show that different abilities can break separately. So we still need the parts.

What we do not know

The biggest open question is an old one: how do body and mind belong together? Our answer: the body is the carrier, and the mind is what it carries – what the living body does. They are not two things, but one. What stays open is called the hard problem of consciousness: why does it feel like something to be you? Our model can explain when we are conscious. It can also explain why this question feels so mysterious to us: our mind does not show us how it builds our experience, so the experience just seems to be there. But why experience exists at all, our model cannot answer. Nobody can, yet.

Many details in the model are our interpretation of other people's research, not our own measurements. So far, the model has been checked by AIs, not by human experts. This is a proposal, not a final answer.

Who is behind this, and what comes next?

Sebastian Bohnen is an independent thinker. He does not work at a university or a research institute. He built this model together with AIs, which searched for research, argued against his ideas and translated. That made it possible to bring together ideas from many fields that usually stay apart. But it also means: so far, only AIs have checked the model, not human experts.

How can a model like this become accepted? The same way a wall proves it is real: by pushing back. We plan five steps: publish the model on an official research server; ask human experts to try to prove it wrong; test one of our predictions with real data; write down what we expect before we look at a new case; and send it to a scientific journal.

AI is getting better very fast. The length of tasks AIs can do on their own has doubled about every seven months since 2019. Since September 2026, the US government even officially calls such systems “Super Intelligence” – a sign of how seriously it takes them, even if they are not truly superintelligent yet. Our model shows one result of this: new theories can now come from outside universities. Universities, teachers and students will have to find a new role – less handing out knowledge, more checking it. They will have to change faster than they are used to. But they are still needed. AIs sometimes present made-up things as facts. That is why a model made with AI needs to be checked by people who take responsibility for their judgment.