Oracles, Dice, and Talking Heads

Two and a half centuries ago, a wooden marvel toured the palaces of Europe: a turbaned figure seated at a chessboard, beating everyone who sat across from it. Philosophers. Generals. Monarchs. The Mechanical Turk seemed to think, chose its moves, pushed the pieces with a wooden hand that clicked softly on every turn — and nobody could explain how a machine played like that.

When the truth finally came out, decades later, it turned out there was a living chess player hidden inside the cabinet. He watched the board through a system of mirrors, chose the move, and worked the puppet's arm from within. No intelligence at all. Just a man in a box.

But here's the detail that really lands: the automaton toured for almost ninety years. Three generations. A grandfather saw it and told the story. A father heard it and passed it on. And nobody opened the right little door — not because they couldn't, but because they didn't want to. People were happy to take the shape of an answer for the answer itself.

That's a very old principle. Far older than the eighteenth century. Let's start at the beginning.

The mountain air on Parnassus has its own smell: pine resin, sun-warmed limestone, and something else — laurel, maybe, or just the stone itself. Here, at Delphi, an oracle operated from the eighth century BCE — and it didn't run for a few generations, it ran for over a thousand years, all the way into the fourth century CE. That number matters. Archaic Greece, the rise of the city-states, the Hellenistic age, Rome, the first Christians — Delphi outlasted all of it as a working institution. Not because real magic was happening there. Because the system was built with remarkable intelligence.

The Pythia — the priestess of Apollo — didn't prepare for a session like some storybook fortune-teller. She fasted. She drank from the Castalian spring: cold mountain water with a faint mineral edge. She chewed laurel leaves — bitter, resinous, mildly intoxicating. Then she sat on a bronze tripod set over a crack in the rock, from which, ancient writers claimed, some kind of vapor rose. The modern theory: ethylene gas seeping up through geological faults beneath the temple may have caused a mild trance — her speech becoming broken, oddly compressed, perfectly suited to interpretation. That doesn't explain prophecy. But it does explain why her words were so easy to read in any direction afterward.

Because the Pythia herself didn't own the answer. Only the raw material. Priests translated her ecstatic fragments into coherent verse. Between the mountain gas and the political decision stood a whole chain of intermediaries, each one adding their own reading. What came out the other end wasn't the priestess's speech anymore — it was the shape of an answer.

Now, about access. Consultations weren't available on demand. Only on certain days — and never in winter. To ask a question, you had to travel, make a sacrifice, wait for the right day. The whole ritual wasn't just ceremony — it was a discipline of questioning. The system forced you to think before you asked. It slowed the impulse down. And that's exactly where the real work began.

If the question was vague, the answer came back just as vague, wrapped in beautiful language. The classic case: King Croesus asked whether he'd defeat the Persians. The answer came: a great empire will fall. Croesus heard a promise of victory, marched his army — and a great empire did fall. His own. The oracle's answer was flawless. Croesus had simply asked the question in a way that let any outcome fit — and the mirror obligingly reflected his own wish back at him.

But when a city-state came with a concrete request — should we found a colony on this particular coast, what ritual will reconcile these two feuding clans — the answer turned out to be genuinely practical. Delphi could give sound advice. It just needed a sound question.

The oracle at Dodona, tied to Zeus, worked completely differently, but the logic was the same. There, priests read meaning into the rustle of leaves in the sacred trees — quiet, wordless, demanding translation. And priests who read the entrails of sacrificed animals only got their raw material after the sacrifice itself — an act that produced the very thing they'd then interpret. The question shaped not just the answer, but how that answer was even obtained: one god was consulted through leaves, another through sacrifice. The form of the question and the form of the answer were inseparable.

A completely different climate, a different language, a different mechanism — but the same core idea.

The I Ching — the Book of Changes — is the oldest Chinese text still in active use, not as a museum piece but as a living practice. You take coins, or a handful of yarrow stalks, throw them, record the result — six times over. The pattern of outcomes builds a hexagram: six lines, each solid or broken. Sixty-four possible combinations, each with its own text. Not a verdict, not a prophecy — more like a map of the present moment: here's the shape of the situation, here's the direction things are moving, here's what's worth paying attention to.

Notice the difference from Delphi? There, people asked about the future — what will happen. Here, it's about what's happening now and where it's heading. A completely different frame: not prophecy, but diagnosis.

And here too, everything hinges on the question. Too broad — "how should I live?" — and the answer fits anyone and no one in particular. That's Croesus again, minus the army: a question so roomy that any hexagram will seem to fit. Too narrow — "will I win this specific deal today?" — and the system breaks down: a hexagram describes movement, not a single point. It can't predict a specific outcome, because that was never its job. The right kind of question frames a situation as a process: "I'm caught between two obligations that pull against each other — where's the real ground here?" That's a question a hexagram can actually work with.

What's elegant about the mechanism is that randomness acts as a catalyst. You generate the uncertainty yourself, by throwing coins, and then decode it through a text written millennia ago. A random result triggers a commentary that's anything but random — carefully composed, deliberate. And that commentary forces you to think about your own situation in terms you wouldn't have chosen on your own. That shift in perspective is where the value actually lives — not in the answer, but in the change of angle. Delphi did the same thing through priests. Same principle, different go-betweens.

Back to the cabinet with the turban.

Wolfgang von Kempelen built his automaton in seventeen seventy — a wooden figure, a cabinet full of brass gears, a chessboard. The construction was clever: before every match, the audience was shown the open drawers full of machinery and convinced there was no one hiding inside. While the drawers were open, the operator crouched behind a partition. Once they were closed, he'd unfold himself, settle in, and get to work.

The Mechanical Turk toured for nearly ninety years. By some accounts, Napoleon played it — and lost. The audience had no doubt: the machine was thinking. Doubting it would have been uncomfortable — it would have spoiled the show.

Here's the detail people usually skip past. The chess automaton didn't win because a chess master happened to be inside. It won because the audience never asked the right question. Nobody said, "what happens if we take the whole cabinet away?" The questions that would have exposed the mechanism simply never came up — because the audience had already accepted the story of a thinking machine, and fit everything they saw into it. Same trap as Croesus: accept the convenient answer, and you stop checking it.

Line all three stories up, and they point to one diagnosis. Not about deception, and not about mysticism. About what happens in the space between a question and an answer. Any answer-machine — a priestess in the fumes, a text thousands of years old, a wooden chess player — works like a mirror. It hands you back your own question, dressed up in symbols. Ask it something precise, and the mirror is useful. Ask it something vague, and it just flatters you.

And here's the tool that falls straight out of these three thousand years of history: before you ask anything of a system that promises answers, check how many ways your question could be read. If there are many — the oracle isn't the problem. It's just handing your own uncertainty back to you, nicely wrapped.

The priests at Delphi understood this instinctively. The whole ritual — the long road up the mountain, the sacrifice, waiting for the right day — wasn't magic. It was discipline. A system that forced a person to sit with their question before they ever opened their mouth. The I Ching does the same thing differently: the act itself — counting stalks, throwing coins, recording each line — is a pause long enough for a vague question to sharpen into a real one. And the Mechanical Turk simply proved the flip side: when the intermediary is invisible, people stop checking the mechanism and start mistaking the shape of an answer for its substance.

Ask yourself — what would your last question have sounded like if you'd had to earn the right to ask it: a trek into the mountains, a fast, a sacrifice, a wait for the right day? What would you have changed in how you phrased it — and would it even still be the same question?

Probably not. And that "probably not" is exactly why people spent thousands of years building these devices in the first place.

The world is enormous and one question is enough to start. moygid dot online.