Lesson 4 — The Neuron: How a Thought Travels

Understanding the Complex: How Does the Brain Actually Work?

---

A thought is not a thing. It's a process.

More specifically: a thought is a pattern of electrical activity distributed across millions of neurons, arriving, cascading, and settling — like a wave moving through water. At no point does the "thought" exist in one place. At no point is it written down in a single cell. It lives in the pattern, not the parts.

But to understand the pattern, you have to understand the parts. And the most fundamental part is the neuron.

---

The shape of a neuron

Neurons are among the most morphologically diverse cells in the body — they come in hundreds of shapes and sizes — but most share a common architecture.

The cell body, or soma, contains the nucleus and most of the metabolic machinery. Branching off the soma are dendrites — tree-like extensions that receive incoming signals from other neurons. Extending from the soma in the opposite direction is the axon: a long, thin fiber that carries the neuron's output signal away to other cells.

Axons can be remarkably long. The motor neurons that run from your spinal cord to your toes have axons over a meter in length. The sciatic nerve — the longest nerve in the body — is essentially a bundle of such axons.

Many axons are wrapped in a fatty insulating sheath called myelin, produced by specialized glial cells. Myelin allows the electrical signal to "jump" between exposed nodes rather than traveling continuously, dramatically increasing conduction speed — from about 1 m/s in unmyelinated fibers to up to 120 m/s in myelinated ones. Multiple sclerosis is a disease of myelin degradation, which is why it produces such varied neurological symptoms.

---

Firing: the action potential in detail

At rest, the neuron maintains a membrane potential of around −70 mV. Incoming signals from other neurons — arriving via dendrites at synapses — nudge this potential up or down. Excitatory signals push it toward zero and beyond; inhibitory signals push it further negative.

If enough excitatory input arrives within a short enough window, the membrane potential reaches a threshold — typically around −55 mV. At this point, voltage-gated sodium channels snap open: positively charged sodium ions flood in, driving the potential to +40 mV in less than a millisecond. This reversal propagates along the axon: each section depolarizes and triggers the next, producing the traveling wave called the action potential.

The action potential is all-or-nothing. The neuron either fires completely — producing a full-sized spike — or doesn't fire at all. There's no half-strength action potential. This is different from an analog signal: you can't turn the volume up or down.

But here's where it gets interesting: neurons encode information in firing rate, not spike size. A neuron stimulated gently might fire 5 times per second. The same neuron stimulated intensely might fire 200 times per second. The magnitude of each spike is identical — but the frequency carries the information.

This is what neuroscientists mean when they say the nervous system uses analog rate coding. The spikes are digital (on/off), but the message is analog (how many per second).

---

The synapse: where neurons talk

An action potential travels down the axon and reaches its tip — the axon terminal. But neurons don't touch each other directly. Between the terminal of one neuron and the dendrite of the next lies a tiny gap: the synapse, typically 20–40 nanometers wide.

When the action potential reaches the terminal, it triggers calcium channels to open. Calcium ions rush in, causing vesicles — tiny membrane-bound packets — to fuse with the terminal membrane and release their contents into the synaptic cleft: molecules called neurotransmitters.

Neurotransmitters drift across the gap and bind to receptors on the receiving neuron's dendrite. Depending on the neurotransmitter and the receptor type, this binding either excites or inhibits the receiving neuron.

The main excitatory neurotransmitter in the brain is glutamate. The main inhibitory one is GABA. But there are dozens of others — dopamine, serotonin, acetylcholine, norepinephrine — each with specific roles in modulating activity across different brain circuits.

---

Hebbian plasticity: the learning rule

The synapse isn't fixed. Its strength — how much effect a given signal has — can be modified based on activity history.

Donald Hebb proposed the principle in 1949: "Neurons that fire together, wire together." More formally: if a presynaptic neuron repeatedly activates a postsynaptic neuron, the synapse between them strengthens. This is called long-term potentiation (LTP), and it's the cellular basis of learning.

The reverse — synapses weakening when neurons are rarely co-active — is called long-term depression (LTD). Together, LTP and LTD allow the brain to encode information by adjusting the relative strengths of millions of synapses over time. A memory is not a stored file; it's a pattern of synaptic weights distributed across a network.

This idea — that connection strength encodes experience — is directly analogous to how artificial neural networks work. The perceptron, invented in 1958, and the modern deep learning architectures used in AI today, are all built on this same principle: adjust the weights of connections based on whether the output was correct.

The brain was the original inspiration. But as we'll see in Lesson 6, the analogy has important limits.

---

Next lesson: Map of the Brain — regions, networks, and the fascinating case of the man who couldn't form new memories.

---

Reading time: approx. 10–11 minutes