Bibliomancy

What it is

Opening a book at random and taking the passage found there as an answer, an instruction or a comment on your situation.

The book is usually one already regarded as authoritative — scripture, or a poet of canonical standing. The randomising step is the same as in casting lots; what differs is that the outcome is a text, which supplies far more material for interpretation than a yes or a no.

Where it came from

The Romans used it with poets. Sortes Vergilianae — consulting Virgil — is attested widely enough that the Historia Augusta records several emperors doing it, and the practice extended to Homer among Greek speakers.

Christians adopted the method with scripture, as Sortes Sanctorum, and the most famous instance is also the most consequential.

In Book VIII of the Confessions, Augustine describes sitting weeping in a garden in Milan, paralysed between his old life and conversion, when he hears a child's voice from a neighbouring house chanting tolle lege — take up and read. He takes it as a divine instruction, opens Paul's letters at random, and reads Romans 13:13–14: not in rioting and drunkenness, not in chambering and wantonness, but put on the Lord Jesus Christ. He writes that the light of certainty flooded in and all the shadows of doubt fled.

Western Christianity's most influential theologian was converted by a randomly opened book.

The Church's position was consistently hostile and consistently ignored. The Council of Vannes in 465 condemned the practice, as did Agde in 506 and others afterwards — and the repetition of the prohibitions across centuries is the clearest possible evidence that clergy kept doing it. Bibles were used for divination throughout the medieval period, and the practice persists in various Christian communities today under names like Bible dipping or, less reverently, Bible roulette.

How it is done

Frame a question or simply a need. Open the book without looking. Read what your finger lands on, or the first passage on the page. Some traditions specify three attempts and a synthesis.

What we can and cannot say

We can say there is no evidence of anything guiding the page. This is the least controversial claim in the section.

We can say the mechanism that makes it feel accurate is unusually clear here, and it is not mainly about the reader.

Sacred and canonical texts have extraordinarily high base rates of applicable content. They are about suffering, doubt, betrayal, hope, endurance, appetite, mortality and the desire to change — because those are what people write books about and what makes books last. Open any page of the Psalms and you will find something about distress and something about deliverance. Open Virgil and you will find something about fate, journey and loss. The odds of landing on a passage that can be made to bear on a human predicament are close to certain, because the whole book is about human predicaments.

This is worth stating plainly because it is the strongest version of a point that runs through the entire section: the impressiveness of a hit depends on how large the target is, and here the target is the whole page.

We can say the practice nonetheless does something real, and Augustine is the demonstration. He was not undecided in the sense of lacking information; he had been arguing himself in circles for years. What the random verse supplied was not knowledge but permission to stop deliberating — an external authority that ended a debate he could not end himself.

That is the same function identified under what divination is, and it explains why the practice survives condemnation. A person who needs a decision more than they need the right decision has been well served.

We cannot say what would have happened in that garden without the book, which is precisely the kind of counterfactual history cannot supply.

Further reading

• Augustine, Confessions, Book VIII — short, and worth reading whatever you make of it.
• Pieter van der Horst, "Sortes: Sacred Books as Instant Oracles in Late Antiquity" (1998).