Methodology

1 · How the texts are prepared

Every text is stored as plain UTF-8 Hebrew with a four-line provenance header: English name, Hebrew name, edition & license, and source URL. On load, each verse is reduced to its consonantal letter stream — vowel points (nikkud), cantillation (te'amim), maqaf, spaces, and punctuation are stripped and carry zero gematria value, exactly as classical practice requires. The Torah is treated as one continuous scroll for ELS; other books are indexed independently.

Torah Dimensions ships more than one base text on purpose. The default Torah, the Westminster Leningrad Codex, and the Miqra-according-to-the-Masorah (Aleppo tradition) edition have different letter counts — 305,859, 304,740, and 304,801 respectively. That variance is not an error; it is the reality of manuscript transmission, and letter-exact work should be able to see it. Every text's license is displayed inside the software and listed in the third-party notices.

2 · Gematria & letter transforms

The calculator computes every classical method from the same consonantal stream: Standard (Mispar Hechrachi), Gadol (finals extended 500–900), Katan (reduced), Ordinal (Siduri), Triangular (Kidmi), Squared (Perati), and the integral-reduced Katan Mispari. The At-Bash cipher (א↔ת, ב↔ש …) and single-step letter elevation/descent are provided as transforms. Final forms share their base letter's value except under Gadol.

Tagin Gematria follows the Baal HaTurim: the crown-count of each letter (three for שעטנז ג"ץ, one for בד"ק חי"ה) multiplied through its value — a standalone calculation over any word you type.

3 · ELS — equidistant letter sequences

An ELS search walks the letter stream and reports every place a target word appears at a fixed skip d, forward or backward, within a chosen skip range. Options include ignore-finals, anagram (any ordering of the target letters), and a verse-range limiter so a passage can be searched instead of the whole scroll. Each hit reports its exact letters, skip, direction, and location.

Stacked conjunctions — the unit that matters

A single short word found as an ELS is statistically meaningless; short sequences occur at some skip in any long text. The pattern with real teeth is a conjunction: several target words all appearing at one shared skip, same direction, and — optionally — in the listed order along the stream. Torah Dimensions scores the whole conjunction as one unit. Every control text is required to try the same feat, so what gets measured is the rarity of the group, not one lucky word.

4 · The Control Lab — honest significance

This is the heart of the instrument. Any analysis — value search, phrase match, reishei/sofei, letter count, ELS, or a stacked ELS conjunction — can be run in parallel on the Torah and on a pool of comparison texts, with identical parameters applied to every text (a fixed skip range, never a per-text auto value). Counts are normalized per 100,000 letters, and the primary text is scored against the pool.

The comparison pool

• Secular Hebrew literature — public-domain translations (Tolstoy, Hugo, Dostoevsky, Cervantes, Sienkiewicz) and original Hebrew (Mapu). Real language, no claim of encoded meaning.
• Two synthetic null models, generated and reproducible: Uniform — letters drawn independently and uniformly over the 22-letter alphabet; and Torah-Frequency — letters drawn independently from the Torah's own empirical letter-frequency distribution (all 27 forms, matched to within 0.08%). Both are built to the Torah's own word-length and words-per-verse distributions, so only the letter content is random. A finding that beats the Torah-Frequency null cannot be dismissed as "the Torah simply has common letters" or "typical word lengths."

The statistics reported

For each run the Control Lab reports the primary's rate, the control mean and standard deviation, a z-score (how many standard deviations the primary sits above the pool), the percentile (share of controls it exceeds), and an empirical p-value — the fraction of texts (primary included) that matched or beat it.

Unit Convergence — do several names really cluster?

A newer Control-Lab instrument for the most common "hidden name" claim. Give it several whole units — names, words, or phrases, each kept intact as an anagram — and it finds the tightest passage of a chosen range whose letters can spell every unit, then asks how often a pool of secular texts converges the same units at least as tightly. It reports the primary's convergence, a z-score, the selection-free count (how many controls pack the units as closely), and the expected number of such clusters in a same-length control. The lesson it teaches is the product's whole thesis: any one name can hide in almost any text by chance; only the conjunction of several, tested against controls, separates a discovery from a coincidence — and more often than not, ordinary prose clusters common Hebrew letters just as tightly.

5 · Seen & Unseen

Over any verse or range, the 22 letters split into those present ("seen") and those absent ("unseen"); each set's gematria is summed. The classical anchor: Genesis 1:1 uses eleven letters; the absent eleven sum to 411; the 411^th word of Bereishit is נפש — the soul, the unseen world. The whole Torah is premapped verse-by-verse under both the common chapter-and-verse divisions and the masoretic te'amim divisions (which differ by a handful of verses — 5,846 versus 5,853), so you can ask which verses conceal every letter of a given word.

6 · The rabbinic lexicon — every word from the oral tradition

Alongside the letter-level tools, Torah Dimensions carries a word-by-word lexicon of the entire Torah — all 187 chapters — built strictly from the oral tradition: Chazal, Targum Onkelos, Rashi, Ramban, Ibn Ezra and Radak, never modern comparative-Semitics glosses. Each word carries its literal rabbinic meaning, its shoresh, part of speech, and full grammar — gender, number, person, state, tense/aspect, and binyan — disambiguated by the ta'amim and masoretic nikkud.

Two disciplines set it apart. First, honest morphology: the map reports the form a word actually carries, never one forced to fit the sentence — a feminine form in a masculine-context verse is marked feminine, a dual is called dual, and a plural-form word taking singular agreement (like אֱלֹהִים) is described exactly so. Second, multiple connotations: where the mefarshim read a word more than one way, every legitimate reading is listed with its source. Hover any word in a verse for its entry, or look a word up directly.

7 · What this instrument does and does not claim

Torah Dimensions is a measurement instrument, not a doctrine. ELS and "Torah code" research is genuinely fascinating and genuinely contested — and you should read both sides in the original. For: Witztum, Rips & Rosenberg, "Equidistant Letter Sequences in the Book of Genesis," Statistical Science 9:3 (1994). Against: McKay, Bar-Natan, Bar-Hillel & Kalai, "Solving the Bible Code Puzzle," Statistical Science 14:2 (1999) — with Brendan McKay's research archive preserving the full exchange. Both papers are open-access. This software takes no side. It shows you the pattern, shows you the same pattern in text with no meaning, and reports the numbers. The judgment is yours.

Hebrew is treated as a language of letters, not of coincidences. The tool measures; it does not interpret. Meaning is left to the one who studies.

8 · The Modern Methods — uncharted territory

An honest distinction: gematria, acrostics, and tagin are classical techniques with centuries of tradition behind them. Even ELS is younger than most assume — pioneered barely a century ago by R. Michael Dov Weissmandl, who worked his searches by hand, with a single medieval precedent in Rabbeinu Bachya's commentary (1291). The software's Modern Methods — figurate arrangements, Chaos-Game fractal portraits, entropy & compression profiles, an 8×8×8 letter-lattice, and Unit Convergence — have no tradition at all. They are new computational lenses over the same letters: no tradition vouches for them, and no literature has mapped their results. Each still answers to the Control Lab, so an honest verdict rides on every one.

9 · The scholars behind the newest methods

Torah Dimensions' newest analytical dimensions stand on the published work of two scholars, credited here with gratitude — and their books belong on your shelf:

• Rabbi Yitzchak Ginsburgh — 913: The Secret Wisdom of Genesis (Gal Einai). The figurate teachings — Genesis 1:1 as the triangle of wisdom: 28 letters = T₇, total value 2701 = T₇₃, with ואת הארץ = 703 = T₃₇ nested inside. More at inner.org.
• Prof. Haim Shore (Ben-Gurion University) — Coincidences in the Bible and in Biblical Hebrew. The regression methodology relating a Hebrew word's gematria to the logarithm of the physical quantity it names — nineteen analyses, honestly presented with their statistics. More at haimshore.blog.

As an Amazon Associate, we earn from qualifying purchases.