Everyone knows Freud the psychoanalyst. Fewer know Freud the neurologist.
Before the Interpretation of Dreams (1900), before the Oedipus complex, before the talking cure — Freud was one of the best neuroanatomists of his generation. He published papers on the structure of nerve cells. He traced neural pathways in the medulla. He knew the brain.
In 1895, he wrote a text that most of his readers have never seen: the "Project for a Scientific Psychology." In 42 dense pages, he attempted to ground psychology entirely in neuroscience. Neurons, he proposed, have "contact barriers" between them — the word "synapse" did not yet exist, but the concept did. Memory, he argued, is a permanent alteration of these barriers. The mind operates on a "principle of neuronic inertia" — it seeks to minimize stimulation.
This last idea is striking. In 1895, Freud proposed that the fundamental law of mental life is the minimization of surprise — 115 years before Friston's free-energy principle. He had the right intuition.
He abandoned the Project in 1897. Not because he lost interest — because he had no way to test it. You cannot observe a synapse in 1895. You cannot measure the firing rate of a neuron. You cannot run an experiment on a "principle of neuronic inertia" when your only tool is a scalpel.
Freud spent the next 40 years building a clinical framework — psychoanalysis — that was testable in a different sense. You can test whether free association works. You cannot test whether the "contact barrier" theory of memory is correct with 19th-century technology.
GEME is, in a very literal sense, what Freud's 1895 Project looks like when given modern tools. A minimal set of axioms (not neurons, but frames). An economic principle (not neuronic inertia, but competitive pressure). A self-referential loop (not the ego, but L6 judgment). And, crucially: code that runs, data that can be measured, experiments that can be reproduced.
Freud had the right architecture. He was just 130 years too early for the implementation.