The Julian Experiment | What Makes You the Same Person?

Meet Julian

What would have to survive before we could honestly say that the same person had returned?

Imagine that historians, neuroscientists and artificial-intelligence engineers decide to reconstruct the Roman emperor Julian. They gather everything that survives: his letters, speeches, philosophical writings, laws, military decisions, accounts written by admirers, accusations written by enemies, descriptions of his education, religious convictions, habits, friendships and political ambitions. They reconstruct the books he is likely to have read, the Greek philosophical traditions that shaped him, the court in which members of his family were killed after Constantine’s death, the years of constrained upbringing that followed, the military success in Gaul and finally the brief reign during which he attempted to redirect the religious future of the Roman Empire. The archive is imperfect, but suppose technology becomes so good that the resulting system speaks with unsettling plausibility. Ask it about Christianity and it answers with arguments consistent with Julian’s surviving writings. Ask about Homer, Iamblichus, Persia or imperial responsibility and it responds in ways historians find difficult to fault.

Then someone presses the obvious button.

“Julian?”

The system answers yes.

What exactly has been demonstrated?

Perhaps an extraordinary historical reconstruction. Perhaps one of the most sophisticated educational instruments ever made. Perhaps a model capable of predicting what Julian might have said under circumstances he never encountered. But none of those achievements establishes that Julian himself has returned.

The distinction between those claims is the entire experiment.

The Archive Is Not the Man

Historical figures tempt us because they appear unusually reconstructable. A person such as Julian left behind words in his own voice as well as administrative records and hostile and sympathetic accounts. In comparison with the majority of people who lived in antiquity, he casts a long informational shadow. That makes it possible to imagine collecting enough fragments that the absent person slowly emerges from them.

But an archive is not a life recorded at full resolution. Julian did not write down every hesitation, private association, forgotten childhood experience, bodily sensation or thought he chose not to share. Accounts written by others were shaped by political position, religious conflict, literary convention and memory. Even Julian’s own works reveal the Julian who decided to write, not an unrestricted recording of every process that constituted him.

This matters because information about a person is not obviously identical with the information that made the person.

A biography can explain why someone acted.

That does not mean the biography can become the actor.

The distinction becomes obvious when we apply it to ourselves. A future historian could possess every email you sent, photograph you took, purchase you made and conversation you recorded and still encounter an enormous silence surrounding the subjective life that connected those traces. The archive would contain evidence of you. It would not automatically contain the place from which you experienced being you.

Julian is therefore useful precisely because he is dead. He removes the emotional convenience of assuming that reconstruction and continuity are the same thing.

DNA Would Not Bring Him Back

The first candidate for identity is biology. Suppose a future civilization recovered enough genetic material to reconstruct Julian’s genome with remarkable accuracy. Would a genetically identical child born in the twenty-fourth century be Julian?

Almost certainly we would say no.

The genome would influence development, physiology and perhaps some behavioral dispositions, but Julian was not produced by DNA alone. He was produced inside a particular body undergoing a particular developmental history, surrounded by particular people, languages, dangers and institutions. The murder of relatives after Constantine’s death belonged to his biography. So did his education, his immersion in Greek culture, his encounters with Neoplatonic religious thought, his military command and the political world of the fourth-century empire.

Give the same genome a childhood in modern Paris and we do not get Julian with Wi-Fi.

We get another person.

This immediately reveals that biological similarity and personal identity are different questions. Identical twins already demonstrate the point in a less speculative form: extremely similar genomes do not create one shared person.

Yet moving from DNA to experience does not solve the problem either.

The Perfect Biography Still Produces a Copy

Suppose we become more ambitious. We reconstruct not merely Julian’s genes but his entire developmental environment. The child hears the same languages, reads the same texts, encounters reconstructed versions of the same teachers, experiences events arranged to resemble the decisive events of Julian’s life and reaches adulthood possessing memories almost indistinguishable from those attributed to the emperor.

Would he now be Julian?

The thought experiment becomes ethically grotesque long before it becomes philosophically successful. More importantly, similarity of history still does not establish identity.

Two people can undergo similar events without becoming one person. Even if the reconstruction were implausibly precise, the new individual’s experiences would occur centuries later, inside another causal chain. He might sincerely believe himself to be Julian because his memory architecture had been engineered that way. His sincerity would tell us something about his psychological state. It would not tell us that the consciousness that once looked through Julian’s eyes resumed.

The crucial variable appears to be missing from the archive.

Continuity.

Not merely the same information, but the question of how one conscious state becomes the next one.

The Copy Problem Arrives in Rome

This becomes clearer if we remove history from the experiment.

Imagine a machine capable of copying a living person perfectly. It reproduces the brain at whatever level turns out to be necessary, including memory, personality, emotional dispositions and every relevant pattern of neural organization. The duplicate wakes and remembers walking into the machine. It recognizes your family. It knows your secrets. It remembers the same childhood.

Now leave the original alive.

There are immediately two people who remember being the person who entered.

Neither informational similarity nor sincerity can make them numerically identical, because from the moment of duplication they begin accumulating different experiences. One turns left. The other turns right. One hears a joke the other never hears. Their common past branches into two futures.

This is why the Julian Experiment cannot be solved simply by adding more historical data.

Even a perfect model confronts the copy problem.

If Julian himself were somehow standing beside the reconstruction, the question would become easy. One is Julian; the other is an extraordinary likeness.

Death does not logically change the criterion simply because the original is no longer present to object.

FAQ

Who was Julian?

Julian, commonly called Julian the Apostate, was Roman emperor from 361 to 363 CE. He was educated in Greek philosophy and is particularly remembered for attempting to restore and reorganize traditional non-Christian religious life during an empire increasingly shaped by Christianity.

Why use Julian for this experiment?

Because a substantial body of his own writings survives alongside extensive historical testimony, giving us unusually rich material for imagining a reconstruction while still leaving enormous parts of the private person inaccessible.

Could AI recreate Julian’s personality?

AI could increasingly approximate aspects of his language, ideas and historically documented behavior. That would be a model of Julian, not demonstrated return of the original person.

Would Julian’s DNA be enough?

No. Genetic information contributes to biological development but does not contain the complete biography, memories or continuous conscious history of an individual.

Would perfect memories make the reconstruction Julian?

Not obviously. A duplicate could possess identical memories while existing simultaneously with the original. That shows that memory similarity and numerical identity are different concepts.

What is the real question behind the Julian Experiment?

It is not whether technology can eventually build something indistinguishable from Julian.

The harder question is what must remain continuous for the word same to remain justified.

If we solve that question, we may learn something about resurrecting the dead.

But long before that, we will have learned something much more important about the people who are still alive.

Because the Julian Experiment was never really about Julian.

It asks why the person reading this sentence is still the person who began the article.