Nature tends toward heterogeneity of physical channels for signalling, as opposed to engineered systems which normally dedicate one channel (electricity) at a time for all signals (and when they accidentally do not we call it a "side-channel").

The heterogeneity provides noise resistance from the environment and a physical priority/filter system for signals necessary for an organism's survival.

These physical channels provide upper bounds for the richness, latency and diffusion of the information; they also incur different costs for maintenance and usage. In turn, evolutionary selection of organisms will yield signalling channels for pain, navigation and other signals necessary for survival within a short time-frame towards fast channels like an electric charge emitted and relayed / filtered by neurons (cells appear to also be affected by charge distributions throughout bodies of certain organisms, which affects limb / organ regeneration; but that's not a charge signified by rapid state changes as one would find in a neuron). Faster responses are made-up for by predictive / reflexive circuits. Slower signals tend to be chemical or mechanical in nature. Chemical messengers diffuse throughout the body better because they can travel with the bloodstream, and trigger all sorts of metabolic machinery for energy availability in /anticipation/ of danger (in exchange for internal damage and reduced immune response).

There are also a multitude of non neural cells that support neural cells in the brain, structurally and by providing resources. Neurons move about a lot, so there is certainly a lot of mechanical activity which might create a small amount of sound in some localized area of brain matter. If that is the case, and if that sound can be isolated as a signal event which in turn becomes useful for constructing neural circuitry, then yeah I would not be surprised if there was an acoustic communication channel as well.

mRNA has been suspected as an information channel between neurons for a while, I think.

The principle is that life is information's way of folding itself into the appropriate substrates. The "appropriate substrates" are physical mediums that survive the environment longer and better than lesser appropriate substrates.

They was some debate and research over mechanical pulses in axons a few years ago. Not sure what's the status of it.

I would like to find the bounds on the "bits" communicated between neurons, so I could then calculate the throughput of the eye for example. Helps to put bounds/identify the algorithms/processing for CV. However, I kept running into increasing complexity/unknowns over the general (and increasingly simplistic) theory of neuro communication and just general unknowns about the dendrite to axon terminal process and all the post signaling cascades.

Sigh, it's a fascinating area.

> what's next? Neurons communicate via sound?

Certainly, but the neurons communicating via sound is to neurons in external brains.

You forgot hormones.