What Is A Memory Made Of? Researchers Look Past Neurons For Answers

A New Theory Ties Memory to Emotion, and Leaves AI Out in the Cold

A familiar song can bring back a memory from decades ago. A certain smell can call up a person or place. A frightening memory can trigger a physical reaction almost instantly, heart racing before the mind even catches up. Computers can store and retrieve information too, but two researchers argue biological memory works on a fundamentally different level, one built from chemistry, not just circuitry.

In a review paper published in the International Journal of Psychiatry Research, Gerard Marx of MX Biotech Ltd. in Jerusalem and Chaim Gilon of the Hebrew University of Jerusalem argue that neuroscience has been examining memory with an incomplete picture of the brain. Their critique is blunt: most previous attempts to explain consciousness left out key biological players, and in doing so, left the real mystery untouched.

Marx and Gilon reviewed more than a dozen major academic works on consciousness, from philosophical treatises to neuroscience texts, and found a recurring problem they describe as the “naked neuron” assumption: researchers were studying neurons in isolation, ignoring the web of other cells and biological scaffolding wrapped around them.

Neuroscience Has Been Missing a Cell

One major oversight, they argue, is the dismissal of astrocytes, star-shaped brain cells physically intertwined with neurons. Another is the neglect of a dense mesh of proteins and sugars that coats all brain cells, what the authors call the brain’s surrounding molecular scaffold. Rounding out the ignored elements are chemical messengers, including neurotransmitters and signaling molecules released by non-neuron brain cells.

“A formula or equation that ignores 3 parameters out of a 6 parametric model cannot be solved,” the authors write, comparing incomplete brain theories to a math problem missing half its variables. Leave out astrocytes, the surrounding scaffold, and the chemical messengers, they argue, and any theory of consciousness is built on a broken foundation.

A New Theory of How Memories Get Made

To fill the gap, Marx and Gilon propose what they call a “Tripartite Mechanism” of memory. Rather than picturing memory as something stored inside neurons the way a computer stores data on a chip, they argue that memory is physically encoded in the protein-and-sugar mesh surrounding brain cells. The idea remains a hypothesis rather than a demonstrated mechanism of memory: no laboratory experiment in this paper measures the process directly.

Storage alone, though, does not create memory in their model. It requires three cooperating parts: astrocytes paired with neurons, the surrounding molecular mesh, and chemical signals that include both metals and neurotransmitters. Together, this trio encodes experiences into the mesh in a way that carries emotional weight alongside information.

That emotional dimension is central to the authors’ argument. A computer can store and retrieve data, but it cannot encode grief, joy, or fear alongside that data. Biological memory does exactly that, they contend, because the chemical messengers involved in encoding are the same ones tied to emotion. Marx and Gilon describe this as cognitive information with “emotive context,” and argue it separates true memory from mere data storage.

In this model, astrocytes are far from passive bystanders. Marx and Gilon position them as the primary drivers of thought and memory, with neurons playing a supporting role as the brain’s wiring for input and output, carrying signals in from the senses and commands out to muscles and glands. The astrocyte clusters, the authors argue, are what actually read, write, and interpret the memory code.

A Machine, the Authors Argue, Cannot Feel a Memory

Artificial intelligence gets a clear verdict from Marx and Gilon. No matter how sophisticated its circuits or algorithms, an AI system is, in their words, inherently “demotive,” meaning it cannot experience the emotional states baked into biological memory. In their view, it cannot be conscious and cannot be genuinely intelligent in the biological sense, regardless of how advanced its computing power becomes.

Marx and Gilon extend their thinking to intelligence more broadly. Consciousness, they suggest, appears even in simple organisms like bacterial colonies that can remember favorable and unfavorable environments. Intelligence, by contrast, requires the more elaborate machinery of neural networks working alongside astrocytes to draw on past experience. In humans, that shows up as musical talent or mathematical ability. In animals, it shows up as tool use, group coordination, or communication.

The Mystery Remains

Marx and Gilon are careful not to overclaim. Their tripartite model, they acknowledge, may represent only “a fragment from the entangled mysteries of consciousness, memory, and intelligence.” What drives the actual felt experience of being alive remains beyond the reach of their mechanism or anyone else’s. Still, by identifying specific physical components that can be measured and tested, the authors argue the field needs to move beyond what they see as largely philosophical descriptions and toward testable biochemical mechanisms. Consciousness may remain an enigma, but they contend it is one that has to be solved with biochemistry, not metaphor.

---