Our Deep but Still Incomplete Understanding of Neural Signaling

Core Concepts

What we understand well: the signaling hardware

The mechanisms in Topics 1–4 are among the most solid knowledge in all of biology: the resting potential, the action potential's phases and channels, synaptic transmission, summation, and saltatory conduction. These were worked out with directly repeatable experiments (voltage clamping, patch clamping, channel blockers, fluorescent voltage indicators). When a textbook says a neuron fires an all-or-none spike driven by voltage-gated Na⁺ channels, that is established, reproducible physiology — not speculation.

The neural code problem: what does a spike mean?

The unresolved question is not how a spike happens but what information it carries. Competing hypotheses include:

• Rate coding: the meaningful variable is how many spikes per second a neuron fires — more spikes, stronger signal.
• Temporal coding: the precise timing of spikes carries information (e.g., some neurons lock their spikes to the phase of brain rhythms).
• Population coding: the meaning lives in the pattern across many neurons at once, not in any single neuron — like the way a crowd's direction of motion is clear even though no individual person matters.

Current evidence suggests real brains use all three, in different proportions in different circuits — but the field does not have a complete, agreed-upon "dictionary" translating spike patterns into percepts.

Glia: the under-appreciated participants

For decades, glial cells were cast as "support" — scaffolding and janitors. That view is now known to be incomplete. Astrocytes regulate ion and neurotransmitter concentrations at synapses, respond to neural activity with calcium signals, and can influence transmission. Microglia prune synapses during development and in disease. Oligodendrocytes and Schwann cells make myelin — and recent work suggests activity can even influence myelination. The honest summary: glia matter for signaling in ways we are only beginning to map.

Synaptic plasticity: mechanism understood, meaning debated

We know the molecular machinery of long-term potentiation (LTP) and long-term depression (LTD) in detail — receptor insertion/removal, second messengers, gene expression. What remains debated is how strongly these phenomena explain memory. LTP is a strong candidate mechanism for certain kinds of learning, but a single synapse's potentiation is not itself a memory; memories clearly involve distributed changes across many synapses and circuits. The chain from "synapse strengthened" to "I remember Tuesday's exam" has major unresolved links.

The limits of our tools shape our knowledge

Every technique trades resolution for scale:

• Single-unit recording (electrodes): millisecond precision on single neurons — but you can only sample a tiny fraction of the billions of neurons, and insertion may disturb the tissue.
• EEG: non-invasive, millisecond timing, whole-brain coverage — but it sums activity through the skull and cannot resolve individual neurons or deep structures.
• fMRI: whole-brain spatial maps — but it measures blood-oxygen changes (an indirect correlate of activity) on a seconds timescale, not neural spikes.
• Optogenetics and calcium imaging: can stimulate or watch defined cell populations — but mostly in animal models, and light penetration limits depth.

What we "know" about the brain is partly an artifact of what these tools can see. That is not a flaw to be embarrassed about; it is the normal, healthy state of a young science.

Model systems have limits

Much of what we know comes from the squid giant axon (huge and easy to study), the sea slug Aplysia (few, identifiable neurons), flies, worms, and mice. These models reveal fundamental mechanisms that apply broadly — that is their power. But a squid axon is not a human cortex. Findings in model organisms transfer to humans selectively, and assuming they transfer everywhere has led to repeated disappointments (e.g., drug candidates that worked in mice and failed in humans). Generalizing across species requires evidence, not assumption.

Common Confusions