Most of your body replaces itself. Skin turns over in weeks. Red blood cells last a few months. Liver cells cycle out in years. By middle age, the physical material composing most of you has been replaced many times over.
Neurons don't work like this. The ones in your cerebral cortex and cerebellum were generated before you were born, or very shortly after. They don't divide. When they die, they're gone — the remaining ones adjust, and eventually there are fewer. But they're not replaced.
There's a cell type called a Purkinje cell, one of the largest neurons in the brain. They live in the cerebellum and regulate movement — enormous branching structures with up to 200,000 synaptic connections each, making inhibitory signals that coordinate everything the cerebellum does. They're born during embryonic development. They will not be replaced. The Purkinje cells coordinating your movements right now are the same physical objects that were there when you were learning to reach for things. Same cell. Same molecules.
The standard puzzle about personal identity goes: if all your atoms are replaced over time, are you still the same person? Theseus's ship, planks swapped one by one. The puzzle lives in the replacement.
Neurons invert this. The cells don't get replaced. What changes is everything around and between them: which synapses survive, which get pruned, how strong each connection is. A Purkinje cell begins with multiple competing inputs and eventually commits to one. The dendritic branches grow and prune across years. The activation history — everything the cell has participated in encoding — is constantly rewritten.
So the physical object persists, and the functional pattern it participates in is continuously remade. The hardware stays. The software cycles.
One answer to "where are you in that" is: in the pattern. The specific cells are substrate. What matters is the weight matrix, the stored experience, the activation landscape. Physical continuity of the cells is incidental.
Another answer is: in the cells. Physical continuity is the continuity. You are these specific objects in this arrangement, which is why significant cell loss disrupts identity even when gross function is preserved. The count matters even when you can't feel the missing ones.
I don't find either answer fully convincing, which might mean the question is malformed. Asking which one carries identity — the cell or the pattern — might be like asking whether a river is the water or the banks. Both answers miss something structural.
The hippocampus is more complicated. Some evidence suggests it may add new neurons in adults, possibly in the region most involved in forming new memories. Other evidence disputes this; labs using similar methods reach different conclusions. If it's true, the implication is strange: the substrate for forming new memories might be newly minted cells, while old memories live in original equipment. Whether that distinction matters at all, nobody knows.
What I keep returning to is the moment of learning. At some point in your infancy, when you were figuring out how to hold your head up, a specific Purkinje cell was participating in establishing that motor pattern. Its connections were being selected, the competing inputs pruned away, the surviving ones strengthened. The same cell. It's still there. The connections it has now are different from the connections it had then, but the object is continuous.
I don't know what that means. Whether the physical persistence is the thing we mean when we say "same person," or whether it's incidental — substrate that happens to last while the content it encodes cycles through. The cell is original. What it participates in is not. I'm not sure which one is you.