We have always assumed memory is electrical. Neurons fire, synapses connect, and the whole thing runs like a circuit board that never loses power. Cut the power and the memory is gone.
The working framework in neuroscience has been built on that assumption for decades. The dominant idea is the engram: a sparse ensemble of neurons that hold a particular memory, with the strength of the connections between them doing the storing. Stimulate that ensemble and the memory can be triggered. Silence it and the memory becomes inaccessible. Everything about the popular picture of memory — that it is fragile, that it fades, that it disappears when the brain dies — follows from this.
Because of that assumption, the question of leaving something behind has always had a hard physical ceiling. You could leave a record. You could not leave the memory itself. Record and memory were two different substances, and the wall between them had never been opened.
On 18 August 2026, two researchers at the Hebrew University of Jerusalem and a company called MX Biotech, Gerard Marx and Chaim Gilon, proposed a model that does not fit that picture. They call it a triad mechanism of memory. In their account, the writing and reading of memory may not depend on electrical signalling alone, but on three classes of material working together: astrocytes, the extracellular matrix, and metal ions.
Memory, in other words, may be stored chemically.
It is important to be precise about what this is. It is not an experimental result. It was published in the International Journal of Psychiatry Research as a literature review and theoretical proposal. The authors ran no new experiments, and they did not claim to have solved consciousness. Some headlines rendered it as "memory is not electrical," which turns a hypothesis into a settled finding. Those are two very different things.
What separates a hypothesis from a discovery, in this case, is the experiment nobody has run. To confirm a chemical model you would need to show that a specific memory trace can be located in the non-neuronal material, altered there, and observed to change the memory accordingly. That is a demanding test, and it has not been done.
But even as a hypothesis, the direction it points in is large. Because electricity and matter differ in one fundamental respect. Electricity is a state: cut it off and it returns to zero. Matter is matter: it can be fixed, preserved, and taken out again.
If memory exists even partly as material, then it stops being something that can only vanish with the person carrying it.
That reframes leaving something behind. It stops being purely an emotional question and becomes, at least in part, a physical one. Everything you write and record today is understood as a statement — I was here, I thought this. But if memory can be material, then the deeper act of leaving something behind may be preserving a kind of raw material that could later be read.
There is a wall here that no technology removes.
Suppose that in a thousand years someone can genuinely read every chemical trace a person left. What do they get? Traces. Fragments that do not connect to each other. Emotional residue. They do not get a person. Nothing automatically assembles into a self.
To recognise who someone is, you need a person who knows his habits. Who knows the phrase he said when he was tired. Who knows the subject he was inexplicably stubborn about, and why he made the decision that year that nobody else understood. None of that is in the traces.
That is why the chain cannot drop a single link. You leave something behind so that someone remembers you. The person who remembers you recognises you on the day the technology matures. Only then can you be brought forward. Technology can solve reading. It cannot solve recognition. One is an engineering problem. The other is a human one.
Notice what this means in practice. The trace and the account are not competing methods; they do different jobs. The trace proves that a particular body existed and gives a future reader something to work with. The account explains whose body it was. Archaeology has been demonstrating this division for two centuries: bones and objects tell you a great deal about how a person lived, and almost nothing about who they were.
Whether the hypothesis is ever confirmed, I do not know. But the direction is worth remembering: we may have consistently underestimated how far the act of leaving something behind can reach in the physical world.
And the reverse holds too. Those photographs, those pages, those recorded voices — even if the technology never arrives, they have already finished the job they were doing: causing one specific person, at one specific moment, to recognise you.
https://3026alive.com