Lesson 5 — Map of the Brain

Understanding the Complex: How Does the Brain Actually Work?

---

In 1953, a 27-year-old man named Henry Molaison — known in the literature simply as H.M. until his death in 2008 — underwent surgery to treat severe epilepsy. His surgeon, William Beecher Scoville, removed large portions of both hippocampi from his temporal lobes.

The surgery stopped the seizures. It also destroyed Henry's ability to form new memories.

He could remember his childhood, his parents, the years before the operation. He could learn new motor skills — if you gave him a task that required practice, he'd improve over time. But every morning was like waking up for the first time. He couldn't remember the nurse who'd just left the room. He couldn't remember that his parents had died. He read the same magazine articles repeatedly, each time as if for the first time.

H.M. was studied intensively for 55 years. He participated in hundreds of experiments and never fully understood that he was a research subject. His case remains one of the most important in the history of neuroscience — because it revealed, with unusual precision, what the hippocampus does.

---

The brain is not one thing

The brain is often talked about as a unified organ. In some ways it is. But it's more accurately understood as a collection of interacting systems, each with distinct evolutionary origins, distinct anatomical structures, and distinct functions — all working together so seamlessly that the seams are invisible in everyday experience.

Broadly, the brain is divided into three major sections:

The brainstem — the oldest and most conserved structure, shared with virtually all vertebrates — controls basic life functions: breathing, heart rate, blood pressure, sleep-wake cycles, and basic reflexes. It connects the brain to the spinal cord.

The cerebellum ("little brain") sits at the back and bottom. It contains roughly half of all neurons in the brain and is primarily involved in movement coordination, balance, and the fine-tuning of motor actions. Damage produces a characteristic clumsiness — not paralysis, but imprecision.

The cerebral cortex — the wrinkled outer layer you probably picture when you think "brain" — is dramatically enlarged in primates and especially in humans. It's divided into four lobes (frontal, parietal, temporal, occipital) with distinct specialized functions, and it's the seat of most of what we think of as higher cognition.

---

Key regions and what they do

The hippocampus, as H.M.'s case demonstrated, is essential for forming new declarative memories — facts and episodes. It consolidates new information into long-term storage during sleep. It's also involved in spatial navigation (the map the London taxi drivers grow). Crucially, H.M.'s case showed that the hippocampus is not where memories are ultimately stored — those are distributed across the cortex — but where they are initially encoded and organized.

The amygdala — two almond-shaped structures flanking the hippocampus — is involved in emotional processing, particularly fear and threat detection. It assigns emotional valence to experiences, which is why emotionally charged memories (your first kiss, the moment of bad news) are often more vivid than neutral ones. The amygdala can trigger a fear response before the cortex has finished analyzing the situation — the reason you jump at a shadow before recognizing it's harmless.

The prefrontal cortex is the most recently evolved region, occupying the front of the frontal lobe. It's involved in planning, decision-making, inhibition of impulses, working memory, and social cognition. It's also one of the slowest to mature — the prefrontal cortex isn't fully developed until the mid-20s, which neuroscientists sometimes cite to explain the risk-taking behavior of adolescents.

The visual cortex (occipital lobe) processes visual information in hierarchical stages — simple edges and orientations first, then shapes, then objects, then faces and scenes. This hierarchical processing principle was a direct inspiration for convolutional neural networks in deep learning.

---

The connectome: brain as network

Describing the brain as a collection of regions is useful but incomplete. The regions don't work in isolation — they're embedded in a dense web of long-range connections. This network of connections — the connectome — shapes which regions talk to which, when, and how strongly.

Advances in diffusion tensor imaging (DTI) and connectomics have revealed that the brain is organized into interconnected modules — clusters of regions that preferentially communicate with each other — connected by hub regions that coordinate across modules. This "small-world" network architecture is similar to the internet or social networks: densely connected locally, with a few highly connected hubs that allow rapid global communication.

Damage to a hub region (like the thalamus, which relays signals between cortical areas) can produce widespread deficits. Damage to a peripheral module may produce only specific, local impairments.

The field of connectomics — mapping the full wiring diagram of the brain — is one of the most ambitious projects in modern biology. A complete connectome of the human brain at synaptic resolution would require storing roughly a zettabyte of data. The mouse brain connectome, far more tractable, is currently being completed by teams at the Allen Institute.

---

Next lesson: Plasticity — How the Brain Rewires Itself. Learning, long-term potentiation, and why AI borrowed from neuroscience (and where the analogy breaks down).

---

Reading time: approx. 10–11 minutes