A field guide to your own brain

Somewhere behind your eyes, about 1.4 kilograms of soft folded tissue is turning these black marks into meaning. It's doing that without any effort on your part, while also holding your balance, keeping your heart going, and running a background hum of half-formed thoughts you didn't ask for.

Most of us carry one around for a lifetime without ever being properly introduced. So here's a short field guide — what the parts are, how the signals move, what the evidence for any of it actually is, and where it all still runs out.

Nothing here is advice. It's a description.

What's actually in there

The brain isn't one thing. It's a committee of specialised regions that have learned to work together, and it's easiest to read from the top down.

The cerebrum is the great wrinkled dome filling most of the skull. Its folds aren't decoration, they're a packing problem solved badly but effectively: the outer layer, the cerebral cortex, is only a few millimetres thick, and flattened out it'd cover far more area than a skull can hold. So it's crumpled. The ridges are gyri, the valleys sulci, and the pattern is broadly similar between people without being identical — your folds are yours.

Underneath and behind sit the older, quieter structures. The cerebellum — "little brain" — handles the fine timing of movement and balance, and it's the reason you can bring a cup to your mouth without watching your hand. The brainstem, where brain meets spinal cord, runs the things you never think about: breathing, heart rate, blood pressure, the switch between sleep and waking.

The asymmetry there is stark. People have lost startling amounts of cerebrum and carried on living recognisable lives. Damage the brainstem and there's no version of that.

Eighty-six billion, and not where you'd guess

The figure people used to quote was a hundred billion neurons, and nobody seems to have known where it came from. It got repeated for decades because it was round and plausible.

Suzana Herculano-Houzel eventually went and counted, using a method that dissolves brain tissue into a soup of intact cell nuclei that can be stained and tallied. The answer came out at roughly 86 billion neurons. She also demolished a second textbook claim on the way through: glial cells, the supporting cast, had long been said to outnumber neurons ten to one. They don't. It's closer to one to one.

The genuinely surprising result was the distribution. About 69 billion of those neurons — around four in five — are packed into the cerebellum, which is a fraction of the brain by volume. The cerebral cortex, the part that gets all the attention and does the thinking you're aware of, holds something like 16 billion.

Neurons aren't the unit that matters anyway. Each one can connect to thousands of others, and it's in the pattern of connections rather than the cells themselves that anything interesting appears to live.

How a signal actually travels

A neuron at rest isn't idle. It's holding a voltage across its membrane — the inside sits about 70 millivolts negative relative to the outside — by constantly pumping ions in one direction against their inclination to go the other way. That stored difference is the charge in the gun.

When a neuron fires, gates open, ions rush across, and the voltage flips and travels down the fibre as a wave. It's all-or-nothing: a neuron can't fire gently. Everything about intensity is carried in how often it fires and how many neighbours fire with it, which is a surprisingly crude code for what it manages to do.

Speed comes from myelin, a fatty sheath wrapped round the long fibres exactly like insulation on a wire, with small gaps left along its length. The signal skips between the gaps instead of crawling along the whole membrane. A bare fibre carries a signal at walking pace, roughly a metre a second. The fastest myelinated ones manage well over a hundred metres a second.

At the end of the fibre the electricity stops. Neurons don't touch — there's a gap, the synapse, and the signal crosses it chemically, as a puff of neurotransmitter released from one side and caught by receptors on the other. That two-stage arrangement looks inefficient and it's the whole point: a chemical junction can be turned up, turned down or silenced without any rewiring, and that adjustability is what learning appears to be made of.

The chemistry has been badly oversold

Popular writing has flattened neurotransmitters into moods. Dopamine as pleasure, serotonin as happiness. Neither survives contact with the research.

Dopamine tracks anticipation and motivation far more closely than enjoyment — it's busiest in the wanting rather than the having, and it responds particularly to rewards that arrive unexpectedly. Serotonin is involved in digestion, sleep, appetite and temperature regulation as much as anything emotional, and the overwhelming majority of the body's supply isn't in the brain at all. It's in the gut.

The honest version is duller. These are signalling molecules doing different jobs in different circuits, and the same molecule means different things depending on where it arrives and which receptor catches it. Any account that hands one feeling to one chemical has simplified past the point of being useful.

Two halves and a cable

The cerebrum comes in two hemispheres, wired to the body in a way that surprises nearly everyone: each half controls the opposite side. Your left hemisphere moves your right hand. The two are kept in constant conversation by the corpus callosum, a thick bundle of something like 200 million fibres running between them.

In the 1960s a small number of patients had that cable surgically cut to stop severe seizures spreading, and the results gave Roger Sperry and Michael Gazzaniga something no experiment could ethically have produced. With the hemispheres disconnected, information could be fed to one side alone. Flash a word to the left visual field — right hemisphere — and the patient would say they'd seen nothing, while their left hand reached out and picked the matching object.

The strangest finding came next. Asked why the hand had done that, patients didn't say they didn't know. They produced a confident, fluent, entirely invented reason. Gazzaniga called the machinery responsible the interpreter, sitting in the left hemisphere, and its job seems to be building a plausible story about behaviour it had no part in causing. That's worth remembering the next time you're sure you know why you did something.

Penfield's map, and how tidy it isn't

The brain has no pain receptors of its own. Headaches come from the tissues around it, which means surgery can be performed on an awake patient under local anaesthetic to the scalp, with the patient talking throughout.

Wilder Penfield made use of that at the Montreal Neurological Institute from the 1930s onwards. Operating on people with severe epilepsy, he stimulated points on the exposed cortex with a small electrode and asked what they felt. A touch here produced a tingle in the thumb. A touch a centimetre away, the lower lip. Working across many patients, he built a map of which strip of cortex corresponded to which piece of body.

The famous picture that came out of it is the homunculus, a distorted little figure laid along the cortex with enormous hands and lips and a shrivelled trunk. The distortion is the information: the amount of cortex given over to a body part reflects how finely it's controlled and how densely it's sensed, not how big it is.

Recent imaging work suggests the strip is less tidy than the drawing implies, with regions concerned with whole-body control interleaved between the effector areas. Which is roughly what you'd expect. The neat picture was always a summary.

The most expensive organ you own

Two per cent of your body weight. Around twenty per cent of your energy budget. It runs on that constantly, day and night, whether you're sitting an exam or fast asleep.

The part that upsets people's intuitions is how little of that cost is task-related. Marcus Raichle and others have pointed out that the change in energy use when you start doing something demanding is small — a few per cent on top of an enormous baseline. Most of the brain's expenditure goes on activity that's always happening, whatever you're doing. When you're apparently at rest and mind-wandering, a particular set of regions known as the default mode network is more active, not less.

Keeping that running takes a serious blood supply, delivered through an arterial ring at the base of the brain, and the vessels there are unusually selective about what they let through. The blood-brain barrier is a tight seal that keeps most large molecules and most pathogens out of the tissue. It's also the single biggest obstacle in neuropharmacology, because a molecule that would work beautifully on a neuron is useless if it can't get to one.

Where memories are made

If you still have any of this tomorrow, a small seahorse-shaped structure buried in each temporal lobe will have had a hand in it.

The hippocampus is essential for turning recent experience into lasting memory, and the case that made that clear is the most studied in neuroscience. In 1953 a young man called Henry Molaison had tissue including both hippocampi removed in an attempt to control severe epilepsy. The seizures improved. He also lost the ability to form new long-term memories of events, permanently, and spent the following fifty-five years greeting the same researchers as strangers.

What he could still do turned out to matter as much. He learned new motor skills — tracing a shape in a mirror, improving steadily day after day — while insisting each time that he'd never tried it before. Memory clearly wasn't one system. Knowing that something happened and knowing how to do something are handled by different machinery, and only one of them went through his hippocampus.

Molaison agreed during his lifetime that his brain should be studied after his death. When he died in 2008 it was frozen, sectioned into more than two thousand slices and photographed, and the whole thing was streamed as it happened.

It changes shape, within limits

The brain was long assumed to be fixed once development finished. It isn't. Connections strengthen with use and fade without it, and the change can be large enough to see on a scan.

The best-known demonstration involves London taxi drivers, who spend years memorising the city's street layout for the qualifying exam. Eleanor Maguire's group found their posterior hippocampus — a region tied to spatial memory — measurably larger than average, with the difference scaling with years on the job, and it appears to be a consequence of the training rather than a precondition for it, since it grows over the course of qualifying.

Plasticity is real. It's also the most over-claimed idea in popular neuroscience, and it's routinely used to sell training programmes that produce improvement on the trained task and nothing else. Getting better at a brain-training game makes you better at that game.

One related question is genuinely unsettled rather than merely oversold. Whether the adult human brain grows new neurons in the hippocampus was argued fiercely in 2018, when two well-conducted studies published within months of each other reached flatly opposite conclusions using different tissue-preparation methods. It matters, and it isn't resolved.

How anybody learned any of this

Fair question: how do you study an organ you can't open up and prod while its owner goes about their day?

For most of the history of the subject, the answer was to wait for accidents. Lesion studies — careful observation of what somebody could no longer do after a specific region was destroyed by injury, stroke or surgery — built the foundations of neuroscience. In 1861 Paul Broca examined a patient, Louis Victor Leborgne, who understood speech perfectly but could produce only one syllable, "tan", and found damage in the left frontal lobe. That region carries Broca's name now.

The other famous case has been embroidered. Phineas Gage survived a tamping iron passing through his skull in 1848 and is usually presented as a man whose personality was destroyed. The careful historical work suggests something less tidy: the dramatic account rests on a thin contemporary record, and Gage went on to work for years as a coach driver in Chile, a job requiring considerable competence. Something changed. It probably wasn't the permanent ruin of the retellings.

It's an uncomfortable inheritance either way. A great deal of what's known came from people who lost something irreplaceable and were then studied.

Modern imaging changed the picture, though less completely than headlines imply. Functional MRI doesn't photograph thought. It measures changes in blood oxygenation as an indirect proxy for neural activity, averaged over seconds and across many repetitions, and the coloured blobs are statistical maps rather than pictures. How much care that demands was made memorably clear in 2009, when Craig Bennett's group ran a dead salmon through a scanner, showed it photographs of people, and found apparent brain activity — because with enough voxels and no correction for multiple comparisons, you'll find a signal in a dead fish. It was a methodological joke with a serious point, and standards tightened afterwards.

Three myths worth dropping

The brain attracts more folklore than any other organ, and three claims have proved unkillable.

The first is that we use ten per cent of it. There's no evidence for this and plenty against: imaging shows activity throughout the brain over the course of a day, damage to almost any region produces some deficit, and evolution doesn't maintain an organ this expensive to run in order to leave nine-tenths of it idle.

The second is the split between a logical left brain and a creative right brain. Real hemispheric specialisations exist — language leans heavily left in most people, which is why a stroke on that side can take away words while leaving someone otherwise sharp. But personality types don't map onto hemispheres, and virtually any task of real complexity recruits both.

The third is learning styles: the idea that people are visual or auditory or kinaesthetic learners and ought to be taught accordingly. It's intuitive, it's believed by a large majority of teachers, and it's repeatedly failed when tested properly. Studies that match teaching method to stated preference generally find no benefit to what's actually learned. People do have preferences. The preferences just don't predict what works. What matters far more is the material — you learn geography with maps and music with sound, whoever you are.

Being wrong about the brain isn't a moral failing. These ideas persist because they're appealing and tell a flattering story about untapped potential and hidden types.

What's still missing

A field guide ought to be honest about where the map stops, and here it stops early.

The roundworm C. elegans has 302 neurons and every connection between them has been mapped since the 1980s. Even with that complete wiring diagram in hand, nobody can fully predict the animal's behaviour from it. That's 302. You're carrying 86 billion.

And the largest question isn't close to answered. Nobody can explain why any of this electrochemical activity is accompanied by experience — why there's something it's like to read a sentence, rather than the processing simply happening in the dark. There are theories. There's no consensus, and no agreed way to test between them.

Which is a strange position to be in. The organ doing the reading right now is the least understood object we know of, and it's the one asking the questions.