Soup vs Sparks
Companion to Coherence Dynamics Circa 1890
Physics does the work. Just add coupling.
No field = entropy wins. Add a hidden field →
Can you manually coordinate B while A syncs automatically? Click Start to find out.
A and B are independent. Add cross-coupling to link them →
A self-organizes through coupling.B decays to entropy.Try adding a hidden field to B →
Left: The “Soup”
Coupled oscillators (Kuramoto model). Each oscillator has its own natural frequency, but through coupling, they spontaneously synchronize.
No central controller tells them what to do. The synchronization emerges from the physics of coupling. The high-dimensional dynamics are native to the substrate.
Right: The “Sparks”
Independent binary switches. Each cell flips randomly with no connection to its neighbors.
To achieve coordination, you'd need an external program that explicitly reads each cell, computes the desired state, and writes it back. The coordination must be simulated, not instantiated.
The Key Difference
Increase coupling on the left and watch synchronization emerge. The oscillators find each other through physics—no step-by-step calculation required.
The “sparks” (right) could in principle be programmed to simulate the same pattern, but they would need to explicitly calculate each state transition. That's the difference between faking the manifold and being the manifold.
Block's Question Answered
Why did evolution favor the slow, messy “soup” over fast, reliable “sparks”? Because electrochemical systems provide a high-dimensional manifold that can be steered. Electrical gap junctions are fast but offer limited controllability. Chemical synapses are slow but expose many tunable parameters—a rich control surface for navigating dynamical regimes.