Is there a hidden code in the Bible? We scanned 289,515 lines

The question

An acrostic is a word spelt by the first letters of lines that follow one another. Some ancient ones are certain, some are argued over, and it is easy to find more, because in a long poem short words turn up by accident.

So this test asks one narrow thing: do hidden words appear in the first letters of ancient verses and lines more often than luck would put them there? It was run by an AI model (Claude) on public texts, at the direction of the author Paul L. G. Harris.

The test

The rules were written down on 4 October 2026 and fingerprinted before any scan. One program then read the first letters of 289,515 verses and lines, downwards and upwards, looking for dictionary words. The six collections were the Hebrew Bible, the Septuagint, the Greek text of the Christian scriptures, Greek verse from Homer to Nonnus, Latin verse and the King James Bible. The King James text is the control: a translation cannot carry a planned acrostic, so it shows what chance does alone.

How a shuffle test works. Put one book's lines in random order: the letters stay, but any plan in the order is gone. Search the shuffled book the same way, 1,000 times, and you see what chance writes from those letters. If the real book holds about as many hidden words as the shuffles, no planner is needed. If far more, something shaped the order. Psalm 119 runs through the Hebrew alphabet, eight verses to a letter: its opening letter goes backwards 0 times in 175 steps, while no shuffle of its verses in 100,000 managed fewer than 67 (a plain count added after the main run, so labelled post hoc).

First the scan had to find agreed answers, and stay quiet about a chapter with 22 verses and no alphabet (Lamentations 5). It did both. A find "survives" only if fewer than 5% of shuffles produce anything as rare anywhere in the same collection.

What came out

Words of five letters or more that occur at least five times, found against expected from shuffling:

The real texts hold what chance predicts. One result is far outside it: 12 alphabet runs in the Hebrew Bible against 0.8 expected. The scan found the known alphabet poems (Psalm 119, Lamentations 1 to 4, Proverbs 31 and others) without being told where they were. Lines 345 to 353 of Nicander's Theriaca begin N, I, K, A, N, D, R, O, S, the poet's name. No shuffle of any Greek poem produced a word that rare: the study puts it at about 3 in 100 million.

The famous short ones cannot be settled by counting. Allowing for every word and place searched, Aratus's ΛΕΠΤΗ has a corrected chance of 0.60, Homer's ΛΕΥΚΗ (generally thought accidental) 0.18, and Virgil's MARS turns up somewhere in shuffled Latin verse in 62% of shuffles. The numbers rank the accident above the accepted case. Scholars trust Aratus for context, and one piece can be counted: in all 289,515 lines and verses, his is the only hidden word of five letters or more that is also the first word of its first line. That count was thought of after looking, so it is a lead, not a proof.

A striking hit fell apart. Luke 3:32 to 36 spells a Greek word for "joy" from its last letters, but every verse ends in a Hebrew name and verse numbers date from 1551.

Esther. Esther has four verses where the first (or last) letters of four consecutive words spell YHWH, the divine name, forwards or backwards. But the Hebrew Bible has 125 such patterns in 235,874 four-word windows, 121 of them outside Esther. Shuffling Esther's words predicts 1.31 against four found (p = 0.043), the one comparison fixed in advance. By rate Esther ranks second of 39 books, behind Joel. About three times its share, at odds of roughly 1 in 20 to 1 in 25: suggestive, no more.

The "Bible code". A short demonstration, not a full test, searched Genesis for words spelt by evenly spaced letters, at every gap from 2 to 1,000, and did the same in 100 scrambled copies. The counts match: HITLER appears 33 times in King James Genesis, against 29.1 in the scrambled copies.

What this does not show

• It does not test meaning or faith. It tests whole dictionary words in the first letters of consecutive lines, and misses anything else.
• The texts are modern printed editions, and one extra line defeats an exact search. The scan missed the signatures of the Greek poet Dionysius Periegetes this way.
• The shuffle is blunt. Hebrew stories open verse after verse with "and", so real Hebrew has fewer hidden words than its shuffles (25 against 50).
• Nothing under about six letters can survive on numbers alone.
• Esther is borderline. Three medieval manuscripts enlarge the letters at 5:4 (per Manguno, 2014): readers noticed the pattern, which does not show the author planned it.

No new acrostic was found. A null result here says nothing against any faith or text.

Where it goes in the books

This is Study 1 behind Written in the Open, Book 3 of Hidden in Plain Sight. The alphabet poems and Esther are in Chapter 1, Nicander, Aratus and Virgil in Chapter 3, and the full account in the Interlude. Chapter 10 covers the Bible code with a separate, fuller test.

You are invited to check it. The plan, code, results and a checksummed list of all 219 source files are kept together, and one command reruns every number. If you find a slip, say so.