**Question**
<br>Can neural population dynamics be well captured by methods developed to describe population dynamics? Are population spike times “incompressible”, such that energy in an incompressible fluid is conserved, so too are population spike by redistributing spiking, not adding them. Draws on ideas of plasticity, criticality, E/I balance, and population coupling.
**Hypotheses:**
<br>Action potential increases in coherent populations are balanced by decreases, such that total population activity follows a Bernoulli incompressible flow-type equation:
<br>v/2 + gz + P/p = constant
<br>where the analogy is (something like):
<br>v = instantaneous spiking
<br>g = mean rate of population
<br>z = p - g
<br>P = difference [?]
<br>p = density of spiking in time (mean rate, single unit)
Extension: add a time function for compensatory “pressure” (population rate) changes to lag behind speed changes.
**Null Hypothesis:**
<br>Driven increases in spiking from single neurons add spikes to the population.
**Method:**
<br>Record from populations of neurons, expected to be related to each other. Targeting approach: dense recording from a “network”, either complete local sampling or precise across-area sampling.
**Result:**
<br>??
