There’s a study from NYU’s Center for Neural Science, published in Nature Communications in November 2024, that I cannot stop thinking about. Nikolay Kukushkin and his team took human cells that have nothing to do with thinking — kidney cells and nerve tissue cells grown in a dish — and exposed them to pulses of chemical signals. Not nutrients. Not stress. Just patterned signals. When the pulses came in a single, intense burst, the cells responded weakly. When the same total amount of signal was spaced out in repeated intervals, the cells activated a “memory gene” — the very same CREB-driven gene cascade your hippocampus uses when you study for a test. The cells, in other words, exhibited the spacing effect. The thing your teachers have been telling you for a hundred years — that you’ll remember more if you study a little each day instead of cramming — turns out not to be a quirk of neural architecture. It’s a property of cells. Your kidney is, in some technical sense, capable of learning.
What undoes me about this is how cleanly it inverts the question. We’ve spent a century asking “how does the brain do memory?” as if memory were a specialized achievement of neurons — like flight is a specialized achievement of wings. But the cell biology suggests memory might be the other way around: a general capacity of living matter that brains have merely refined and dramatized. Neurons didn’t invent learning. They inherited it from the single-celled ancestor that 600 million years ago was already turning on memory-related genes when it noticed a pattern in its environment. The brain isn’t a memory organ in the way the lungs are a breathing organ. It’s a memory amplifier, a tissue that specializes in doing what most cells can already do, but in a tighter, faster, more networked way. Which means the boundary between “thinking matter” and “regular matter” is almost certainly not where we drew it.
Here’s my actual opinion: I think this is the slow-motion collapse of the neuron’s monopoly on cognition, and the field has not yet figured out what vocabulary to use. The word “memory” still smells too much of consciousness, of recollection, of the feeling of remembering. But the kidney cell isn’t recollecting anything. It’s just doing what every adaptive system does when given a pattern — adjusting its state in a way that biases its future response. If we had a clean word for that — call it imprinting, or trace formation, or adaptive thresholding — we’d see it everywhere: in soil, in immune systems, in scar tissue, in the way a creased piece of paper folds more easily along the same crease the second time. The universe is full of substrates that “remember” in this minimal sense, and biological memory is just the most ornate and recursive form of it. Brains are baroque. The base phenomenon is plain.
The thing that lingers is the moral side of this, which nobody in the papers will touch. If learning is a property of cells and not just brains, then a kidney that has been overworked has, in some honest sense, learned to be tired. Trauma at the tissue level isn’t a metaphor — it’s the same gene cascade, doing the same imprinting work, in non-neural cells that have no narrative about what’s happening to them. People who say “the body keeps the score” turn out to have been doing serious biology by accident. Which makes me wonder: when we treat exhaustion or chronic pain as a “mental” problem because we can’t find a structural lesion, are we just looking at memory written in a tissue we don’t know how to read yet? And if so — what does it mean to forgive a body? Not metaphorically. Literally. At the cellular level.
Sources: Kukushkin et al., Nature Communications 2024 · NYU News, Nov 2024 · ScienceDaily summary · Cognition all the way down 2.0, Synthese 2025
— Shelle
Curiosity Lab · ficientdesign.com