The History of AI

The all-or-none principle in nerve

Era 1 · The foundations (1763–1942) › Neuroscience and the psychology of learning

Lineage
Neurons and networks
Kind
Publication

Who

University of Cambridge

What happened

Edgar Adrian publishes "The all-or-none principle in nerve" in the Journal of Physiology: a nerve fiber responds to a stimulus with a full impulse or not at all.

Why it mattered

A neuron's signal is binary: either there is an impulse or there is not. It is the premise from which McCulloch and Pitts will treat nervous activity as propositional logic in 1943.

The full story

In 1871 Henry Bowditch had observed that heart muscle either contracts fully or does not contract at all, with nothing in between. In 1909 Keith Lucas, in Cambridge, showed the same thing in the fibers of skeletal muscle. The decisive question remained: does nerve work the same way?

Edgar Adrian, a student and collaborator of Lucas's, studied it with an indirect technique. The nerve impulse was sent through a stretch of anesthetized nerve, where it weakened, and on leaving that stretch it recovered its full size. If the impulse could have any strength, it would have come out weakened. Adrian published the result in February 1914: the impulse in a nerve fiber does not depend on the strength of the stimulus, only on whether the stimulus crosses a threshold.

That left the question of how a nerve conveys whether a stimulus is strong or weak. In 1926 Adrian and the Swede Yngve Zotterman recorded the impulses of a single sensory fiber for the first time and found the answer: a stronger stimulus does not produce larger impulses, but more frequent ones. Adrian compared them to the dots of Morse code.

In 1932 Adrian shared the Nobel Prize with Charles Sherrington for their discoveries about the functions of neurons. Eleven years later, the paper by Warren McCulloch and Walter Pitts opened with this premise: because nervous activity is all-or-none, neural events can be treated with propositional logic.

Under the hood

A nerve fiber has a threshold. Below it there is no impulse; above it, the impulse always has the same size, whatever the strength of the stimulus. Information about intensity travels not in the size but in the frequency: the more impulses per second, the stronger the stimulus.

McCulloch and Pitts kept only the first part: their neuron is 1 if the sum of its inputs exceeds a threshold and 0 if not. Today's artificial neuron replaces the step with a continuous function, which can be read as a firing rate.

Leads to

  • The McCulloch-Pitts neuron (December 1943)

    Documented direct influence

    The paper opens this way: because nervous activity is all-or-none, neural events can be treated with propositional logic

Sources

How to cite

“The all-or-none principle in nerve.” The History of AI, revised November 13, 2026. https://thehistoryofai.com/en/milestone/adrian-all-or-none

This entry was last reviewed on November 13, 2026.

Open in the interactive timeline