Bats are born without a working compass. The head direction cells — neurons that fire when the animal faces a specific direction, providing the internal sense of “which way am I pointed” — don't stabilize until night five or six of a bat's life. During the first few nights, the cells fire erratically, rotating their preferred directions, failing to maintain consistent tuning across sessions.
This is not a calibration problem in the usual sense. The cells exist. The circuits are wired. The hardware is present from birth. What's missing is the learned relationship between the internal signal and the external world. The compass needs to boot up, and booting up takes repeated exposure to a stable environment.
The booting process reveals something about what the compass actually encodes. Head direction cells don't measure magnetic north or any absolute reference. They encode the animal's direction relative to landmarks — the geometry of the space the animal has explored. Each environment generates its own directional framework. Move to a new room, and the compass partially resets: some cells maintain their tuning, others remap. The representation is not of direction per se but of directional relationships within a known space.
The wild-caught bats in the study showed richer encoding than laboratory-raised bats. Their head direction cells carried more information per spike, discriminated finer angular differences, and maintained more stable tuning across sessions. The hardware is the same. The experience is different. Bats that navigated complex, changing environments developed sharper internal compasses than bats that navigated boxes.
This is a general principle about neural hardware and environmental complexity: the circuitry is built to handle more than any single environment demands. Laboratory conditions underutilize the system. The bat's compass in a box is like a high-resolution camera photographing a blank wall — the sensor is capable of far more than the scene requires, and measuring the sensor's performance in that context underestimates what it can do.
The five-night boot sequence also means there is a window of vulnerability. Young bats navigating before their compass stabilizes are navigating without a reliable sense of direction. They compensate — echolocation provides local spatial information, and following adults provides directional guidance — but the internal directional sense that experienced bats rely on is simply absent. The animal is flying without knowing which way it's going, in a way that becomes invisible once the compass boots and the bat can no longer remember not knowing.