The brain resembles marble being chiseled, not clay being shaped
A team of neuroscientists in Austria has found that the newborn mouse brain arrives not as an empty vessel awaiting experience, but as a dense, tangled web of connections that gradually prunes itself into precision. Studying the hippocampus across three stages of development, researchers observed that neural networks begin in exuberant disorder and achieve efficiency through selective elimination rather than accumulation. This 'pruning model' inverts long-held assumptions about how minds take shape, suggesting that the brain's deepest intelligence may lie not in its capacity to build, but in its wisdom to let go.
- Decades of neuroscientific intuition held that brains grow by adding connections — new research in mice shows the opposite is happening from the very first days of life.
- The hippocampus, the brain's seat of memory and spatial reasoning, begins as a chaotic tangle of dense, seemingly random neural pathways that would appear to defy any notion of design.
- Rather than constructing optimal routes from scratch, the developing brain may exploit its initial density like a traveler choosing the fastest road from an already-built network — a far more efficient strategy.
- Researchers at the Institute of Science and Technology Austria have mapped this transformation across three developmental stages, watching structured order emerge from early neural noise.
- The critical unknown now looms large: whether human brains follow the same pruning logic remains unconfirmed, leaving the full implications of this discovery suspended between mouse and man.
The brain, it turns out, may not begin as a blank slate waiting for experience to inscribe meaning upon it. Neuroscientist Peter Jonas and his colleagues at the Institute of Science and Technology Austria studied mouse brains at three developmental stages — newborn, adolescent, and adult — and found something that defied expectation. Rather than growing denser and more connected over time, the hippocampus begins life crowded with seemingly random neural pathways and becomes progressively leaner and more purposeful as the animal matures.
The team focused on CA3 pyramidal neurons, a circuit central to memory formation. What they observed was a pruning model: the brain starts full and optimizes itself by discarding what it does not need. "Intuitively, one might expect that a network grows and becomes denser over time," Jonas noted. "Here, we see the opposite."
The evolutionary logic, the researchers propose, may lie in speed. A dense pre-existing network of connections allows neurons to rapidly select optimal pathways rather than construct them from nothing — much like navigating a city with roads already built versus laying new asphalt for every journey. The hippocampus, which must weave together sight, sound, and smell into coherent spatial understanding, may depend on exactly this kind of initial abundance.
Whether the same process unfolds in human brains remains an open question. But if it does, the implications would be profound — reshaping how scientists understand learning, early development, and the very nature of the mind's potential. The brain, in this view, resembles marble being chiseled into form rather than clay being built up from nothing.
The brain does not arrive as a blank page waiting for experience to write upon it. According to a team of neuroscientists at the Institute of Science and Technology Austria, the opposite appears to be true: newborn brains are crowded, tangled networks that spend years trimming away excess connections until they become efficient and organized.
Peter Jonas and his colleagues studied mouse brains at three distinct points in development—just after birth, in adolescence around two to three weeks old, and in adulthood at six to seven weeks. They focused on the hippocampus, the brain region responsible for spatial memory and for converting fleeting thoughts into lasting ones. What they found challenged intuition. The CA3 pyramidal neurons that form a crucial memory circuit in this region begin life densely packed with seemingly random connections. Rather than growing denser as the animal matures, the network becomes progressively streamlined and purposeful.
"Intuitively, one might expect that a network grows and becomes denser over time," Jonas reflected on the findings. "Here, we see the opposite." The researchers call this a pruning model—the brain starts full and then optimizes itself by discarding unnecessary pathways. The electrical activity and cellular processes they measured at each stage revealed this transformation clearly: dense chaos gradually resolving into structured order.
Why would evolution favor this seemingly inefficient starting point? The researchers propose that having a dense foundation of connections already in place may allow neurons to communicate more rapidly than if they had to search for and establish pathways from scratch. Consider navigation: traveling between two points is far faster if a dense network of roads already exists and you simply choose which route to take. Building a single road from nothing to reach your destination takes considerably longer. The hippocampus faces a particularly demanding task, integrating sensory information from the eyes, ears, and nose into coherent spatial understanding. An initially exuberant tangle of connections, followed by selective pruning, may be precisely what enables this complex integration to occur.
The implications are substantial, though still uncertain. Whether human brains develop according to the same pruning model remains unknown. The mouse studies offer a compelling metaphor—that the brain resembles a block of marble being chiseled into form rather than clay being shaped from nothing. If this pattern holds across species, it would fundamentally reshape how neuroscientists understand learning and development. It suggests that the brain's power lies not in its capacity to build from emptiness, but in its ability to refine what is already there.
Citações Notáveis
Intuitively, one might expect that a network grows and becomes denser over time. Here, we see the opposite.— Peter Jonas, neuroscientist at ISTA
An initially exuberant connectivity, followed by selective pruning, might be exactly what enables this integration.— Peter Jonas
A Conversa do Hearth Outra perspectiva sobre a história
So the brain starts crowded and then throws things away? That seems backwards.
It does at first. But think about it this way: if you need neurons to talk to each other, they have to find each other first. If the pathways are already there, they just pick the best ones.
Like having a map versus having to draw one yourself.
Exactly. And the hippocampus has to do something really hard—connect what you see, hear, and smell into a single memory of a place. Maybe you need all that initial chaos to make that possible.
Did they see this in human brains too?
Not yet. This is mice. We don't know if our brains work the same way.
What happens if they do? What changes?
Everything, maybe. It means learning isn't about writing on a blank slate. It's about sculpting something that's already there. That's a different story about what we are.