The Current That Remembers
The Earth Shattering Planarian Worm Event
Split a planarian in half and both pieces grow back — the severed tail produces a new head, the severed head regenerates a new tail. This is well known. Less well known is what happens when you alter the worm’s bioelectric signature before you cut it. Michael Levin’s lab at Tufts has spent years demonstrating that the pattern of voltage differences across a body’s cells — what researchers have begun calling the electrome — does not merely respond to anatomy. It encodes it. Every cell participates in a tissue-wide electrical conversation, and that conversation carries a map: this is the head-end, this is where the eye belongs, this is where growth should stop. The voltage landscape runs in parallel with the genome, not as a readout of it, but as a separate governance layer — determining which genetic programs activate where.
In a series of experiments that became quietly famous among researchers and internet addicts, Levin’s team used gap-junction inhibitors to reset this electrical script in living planarians. The worms looked entirely normal. Their genomes were untouched. But when cut, one in four grew back with two heads instead of a head and a tail.
The arresting part is what came next. These two-headed worms were cut again — this time without any chemical intervention. The pieces regenerated as two-headed worms. And again. The electrical rewrite had become permanent, passed from tissue to tissue through every subsequent regeneration, persisting without ongoing instruction. “We have permanently rewritten the target morphology,” Levin wrote. The genome had not changed. The shape-memory had.
What, exactly, was remembered? Not a sequence of bases. Not a neural circuit — the planarian has no brain. The worm remembered in voltage: in the standing electrical geometry of its flesh, maintained by ion channels and gap junctions acting collectively, without any center issuing instructions. No single cell holds the record. The record is the relationship between all of them.
We carry an entrenched picture of memory as storage: engrams in neurons, code in chromosomes, words on a page. Memory, on this picture, is always a record of the past made by a recording device at a discrete moment, available for retrieval. But this is not how the two-headed planarian works. There was no write-head, no bitstring. The memory is the field itself — the ongoing pattern of voltage differentials, continuously maintained and self-reinforcing. What persists is a present-tense commitment: a shape the tissue is perpetually arguing toward.
If memory can live in a field — without a brain, without genes to specify it, without any governing center — then memory precedes, by a very long evolutionary distance, the kind of creature we usually imagine as having one. The planarian is not conscious of its own strangeness. But it has something more ancient than consciousness: a persistence. A shape it will not relinquish even after everything else about it has been cut away.
We tend to locate our memories in the past, as if they were photographs taken at a receding moment. But the worm suggests a different topology: memories as forward commitments written into present-tense electrical patterns, constantly re-enacted by tissues that have no access to the original moment of inscription. What was written: two heads. What persists: two heads. What could undo it: nothing short of another electrical intervention.
Perhaps all memory operates this way. Perhaps what we call remembering is not retrieval but re-enactment — a visit to a site that never stopped happening, a standing waveform insisting on itself each time we approach it. If so, then the self is less a library than a voltage: something maintained by the ongoing activity of the system, not stored within it. Something that would vanish if the current stopped, and resume the moment it began again.

