JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Yes and No: EPSPs, IPSPs, and Adding Them Up

When one neuron releases its chemical message, the receiving cell does not simply obey. It feels a small push — sometimes a "yes" that nudges it toward firing, sometimes a "no" that pulls it back. These little pushes are EPSPs and IPSPs, and crucially they are graded, not all-or-none: they come in every size. A real neuron hears thousands of yeses and noes at once, and its job is to add them all up across space and time and decide, moment by moment, whether to fire. That tally is where a synapse stops being a wire and becomes a computation. This guide walks you, in plain pictures, from a single nudge to the whole vote.

Two flavors of nudge

In the last guides a presynaptic terminal fired, calcium rushed in, vesicles dumped neurotransmitter into the cleft, and those molecules drifted across to the other side. Now we stand on that far shore — inside the receiving cell — and ask the only question that matters there: when the message lands, which *way* does it push? The answer is always one of two. Either it makes the receiving neuron a little more eager to fire, or a little more reluctant. A "yes" or a "no."

The "yes" push is an excitatory postsynaptic potential, or EPSP. It is a small depolarization — the inside of the cell becomes a touch less negative, sliding *toward* the threshold at which the cell would fire. The "no" push is an inhibitory postsynaptic potential, or IPSP: a small hyperpolarization (or a clamping in place) that drags the cell *away* from threshold, making it harder to fire. Same machinery — a chemical landing on the membrane — but opposite directions of travel.

Graded, not all-or-none

Here is the single most important thing to hold onto, and it surprises almost everyone. An action potential — the spike a neuron sends down its axon — is all-or-none: it either fires at full size or not at all, like a gunshot. EPSPs and IPSPs are the opposite. They are graded potentials: they come in every possible size. A faint EPSP might nudge the voltage by half a millivolt; a strong one by several. More neurotransmitter, or more synapses active at once, means a bigger push. Think of a dimmer knob, not a light switch.

Graded signals also *fade*. An EPSP born out at the tip of a dendrite is loudest where it starts and grows quieter as it spreads inward toward the cell body, like a ripple flattening as it travels across a pond. So a single faraway nudge, on its own, almost never reaches threshold. This fading is not a flaw — it is exactly what lets a neuron weigh near and far, recent and old, against one another instead of obeying the first signal to arrive.

  voltage
     |        spike! (all-or-none, fixed height)
 threshold ---/\----------------------------------
     |       /  \
     |   ___/    \___      <- EPSP (graded: any height)
     |  /             \___
 rest|-/-------------------\___---  <- IPSP dips below rest
     |                         \_/
     +--------------------------------> time
EPSPs and IPSPs are small, smooth, any-size bumps and dips. Only when the climb crosses threshold does the cell answer with a single, fixed-height spike.

Adding them up: space and time

A real neuron is not listening to one synapse. A single neuron can carry thousands to tens of thousands of synapses spread across its dendrites, and at any instant a shifting crowd of them are saying yes or no. Because the postsynaptic potentials are graded, they simply *add together* as they spread toward the cell body. This running total is called synaptic integration, and it happens in two directions at once.

  1. Spatial summation — adding across place. Two EPSPs arriving at the same moment from *different* synapses stack on top of each other; together they may reach threshold where neither could alone. An IPSP arriving alongside them subtracts from the sum.
  2. Temporal summation — adding across time. One synapse firing rapidly, again and again, sends EPSPs so close together that each new one piles onto the leftover tail of the last before it can fade. A fast drumbeat from a single input can climb to threshold all by itself.

Picture a bathtub. EPSPs pour water in; IPSPs pull the plug a little. Spatial summation is many taps running at once; temporal summation is one tap left open so the level keeps rising before it can drain. The cell fires when the water reaches a marked line — and inhibition is not just "less yes," it is an *active* drain that can cancel a flood of excitation if it is placed well.

Where the vote is counted

All this adding has to be tallied *somewhere*, and the neuron has a designated counting room: the axon hillock, the little mound where the cell body narrows into the axon. It is packed with voltage-gated channels and has the lowest threshold of anywhere on the cell. The grand sum of every EPSP and IPSP — after all the fading and stacking — arrives here. If the running total at the hillock crosses threshold at this instant, the cell fires one action potential. If it falls short, nothing leaves. The vote is yes only if yes wins *right there, right now*.

Now the deep point. Because the tally weighs *where* a synapse sits (close to the hillock counts more than far out), *when* it arrives (recent more than fading), and *which sign* it carries (yes minus no), the neuron is not relaying a message — it is deciding one. Two cells with the very same inputs can fire differently if those inputs land in different places or at different times. The synapse, multiplied across a dendritic tree and summed at the hillock, has quietly turned into a small machine that computes an answer.